NEWS

* NEWS --- Changes in the Daisy program and libraries -*- outline -*-

This file contains an overview of changes between Daisy releases. When
you upgrade Daisy, it is recommended that you read at least the
headlines of the changes since the last version.  For more detailed
information on changes, read the 'ChangeLog' file from the source
release.

* 5.26

** Duplicate definitions are now an error instead of a warning.

** New 'fractiles' summary model.

Calculate fractiles for specific log variables.

(defsummary NYield fractiles
  (tags "Matrix-Leaching" sorg_N sorg_DM)
  (fractiles 10 50 90 [%]))

** New harvest log variable.

This allows you to add details from the harvest log into normal log files.

(deflog MyHarvest crop
  (where "myharvest.dlf")
  (when true)
  (entries (number (path column "${column}" Vegetation crops
                         harvest "${crop}" sorg_DM)
                   (handle content_sum)
                   (spec fixed Harvest sorg_DM))))

(output (MyHarvest (crop Grass) (when monthly)))

** New 'store_SOM' and 'restore_SOM' management actions.

This will store / restore the state of the SOM pools. Intended use for
long run scenarios to test short term effects of weather without long
term effects of changes to SOM pools.

** Atmospheric 'CO2' can now be specified in weather file.

Currently this is only used in conjunction with the Farquhar
photosynthesis model, where it is required.

* 5.25

** Made duplicate definitions errors instead of warnings.

** Better logging of biopores.

Biopore leaching is now included in the Soil/Field Nitrogen/Chemical
logs, the same way biopore percolation has been included in the
Soil/Field water logs. You can now also seperate drain flow from soil
matrix or biopores, and seperate input and output from biopores,
rather than just the net exchange.
 
You no longer need to include the bottom of the biopore zone in order
to balance Field nitrogen and Field chemical. You still need to
include the entire root zone though.

* 5.24

** FAO_PM now assume dry surface by default.

This what most people would expect. Use (pet (FAO_PM (use_wet true)))
to get higher potential evapotranspiration for wet surfaces.

** Bug fixes.

* 5.23

** No slurry volatilization by default.

Specify 'volatilization" manually.
 
** You now need to specify Soil border when logging to a specific depth.

E.g. you need 

(Soil (border -75 -150 [cm]))

in the column definition, for

(output ("Field water" (from -75 [cm]) (to -150 [cm])))

to work.

** You can now specify initial values of soil variables with a PLF.

E.g. to specify initial value of h that is at field capacy near the
soil surface and two meter below ground, but at -50 cm one meter
below, write this:

(defcolumn Test Askov
   (SoilWater (initial_h_plf (-2 [m] -100 [cm]) 
                             (-100 [cm] -50 [cm]) 
                             (0 [m] 2 [pF]))))

As all PLF's, values between the specified point will be interpolated,
so the pressure 1.5 meter below ground is -75 cm. The numeric cells
will be initialized so the center of the cell has the specified value.

** New 'handle_z' and 'handle_x' parameters in the 'post-process' program.

These allows you to turn a 2D log file into a 1D or 0D log file. The
valid values are 'sum' and 'average' to either add or average the
values for the dimension you want to collapse, wighted by the length
of the numeric cells in the same dimension. The default value for both
is 'all', which will not collapse the dimension.

E.g.

(defprogram test post-process
  (top -0 [cm])
  (bottom -100 [cm])
  (file "soil_water_content.dlf")
  (handle_z sum)
  (handle_x average)
  (dimension [mm])
  (where "average_swc.dlf"))

(run test)

will give you to average water content in mm
in the top meter of the soil
for each timestep.

* 5.22

** Fix compability problems with OSX.

** New 'pipe_outlet' parameter for 2D description of drains.

  (Movement rectangle
            (pipe_outlet PLF 
                         ;; ((YEAR MONTH DAY) WATER_LEVEL)
                         ((1987  9 1) -200 [cm]) ;; Interpolates between values.
                         ((1987 10 1)  -50 [cm]) ;; So mid-September is -125 cm.
                         ((1988  2 1)  -50 [cm])
                         ((1988  3 1) -200 [cm])))

Any drain points specified with 'drainpoints' that are located below
the pipe outlet will have the specified pressure. This means that the
drain points may act as a subsoil irrigation system. The intention is
to simulate drain outlets that are located below the water level in
the recepient.

Note that this is different from the similarly named 'pipe_outlet' in
the 'lateral' drain model, where submerged drains will simply cease to
be active.

** New 'crop' log parameter.

All the relevant log parameterizations in 'log.dai' now support a
'crop' parameter, with the name of the crop to log.

* 5.21 

** You can now specify seasonal variation in aquifer pressure.

  (Drain lateral)
  (Groundwater aquitard
               (pressure_table season
                               ((1 [d] -10 [cm])
                                (100 [d] -100 [cm])
                                (361 [d] -10 [cm])))))

** 'minimal_timestep' changed to 1 microsecond.

Use 

  (minimal_timestep (hours 1))

to go back to 1 hour timesteps.

* 5.20

** Theta_res now defaults to 0.01 in the hypres hydraulic model.

Old value was 0.

** The 'pipe' Grounwater model removed.

Use 

  (Drain lateral)
  (Groundwater aquitard)

instead.

** The 'default' horizon model removed.

Please be explicit about the texture classification system.
'USDA3' is what Daisy used to assume.

** You can now specify chalk content in horizons.

It is currently mostly ignored by the model, but might make it easier
to make the soil fractions add up to 1.

** Removed "N balance" and "Root Zone Water Balance".

Use "Field nitrogen", "Soil nitrogen", "Field water", eller "Soil
water" i stedet for.

** "Soil water" log now includes biopores.
** "Field water" log now include biopore percolation.

** New 'PenWaterFac' root parameter.

Allow water content to affect root penetration. 

  (sow ("Spring Barley" (Root (PenWaterFac (0 1) (0.9 1) (1 0)))))

In this example, root penetration will occur at full speed when the
soil water content is root depth is less than 90% saturated, and stop
when it fully saturated.

* 5.19 

** Volatilization is back in Field nitrogen.

** Bug fixes for surface detension.

* 5.18

** You can now specify pH of soil.

The new 'SoilpH' parameter in the default column model allows you to
specify a 'soilph' model. Two models are implemented. The default is
'neutral' which specify a soil pH of seven. The other is named 'year',
and allows you to specify pH as a function of calendar year and soil
depth.

** Organic matter turnover can now depend on soil pH.

The new 'pH_factor' parameter in the default organim matter model
allows you to specify how turnover is affected by pH.

** New 'afforestation' vegetation model.

This is a forced growth model, i.e. not based on any growth factors
beside time. An example can be found in the 'sample' folder, as
'afforestation.dai'.

** You can now log the highest or lowest value in a volume.

The new 'space' parameter of the 'volume' select model has three
values, 'min', max', and 'sum'. The later is default and enable the
current behaviour.

** The "Field nitrogen" log parameterization has changed.

Dead leaves are now included as part of the crop nitrogen, instead of
seperately. The parameterization now works for permanent vegetation as
well.

* 5.17

** New 'post-process' program to convert daisy log files.

The program converts a Daisy log file into a Daisy log file. The main
function is to select a subset of the data, but it may also convert
between units.

  (defprogram convert post-process
    ;; Select a subset of a ldf file.
    (file "Estrup-content-Bromide.dlf")   ;Input file.
    (where "post.dlf")                    ;Output file.
    ;; Limit x position.
    (left 300 [cm])
    (right 400 [cm])
    ;; Limit depth.
    (top -50 [cm])
    (bottom -100 [cm])
    ;; Limit in time.
    (after 2000 1 1)
    (before 2001 1 1)
    ;; Convert unit.
    (dimension [kg/m^3]))
  
  (run convert)

* 5.16 

** New action 'spray_surface'.

Apply chemical directly to surface, ignoring canopy. The
'irrigate_surface' and 'fertilize' actions now also ignore canopy.

** The 'Freundlich' sorption model changed.

The 'K_OC', 'K_clay', and 'm' parameters now all assumed they have
been calibrated for 'C' having the unit mg/L, rather g/cm^3 as before.

* 5.14

** Fix calculation of 'runoff'.

* 5.13

** New 'pipe_outlet' parameter for the 'lateral' drain model.

This allows (partial) simulation of high water level in drain outlet.

  (Drain lateral
         (pipe_outlet PLF 
                      ;; ((YEAR MONTH DAY) WATER_LEVEL)
                      ((1987  9 1) -200 [cm]) ;; Interpolates between values.
                      ((1987 10 1)  -50 [cm]) ;; So mid-September is -125 cm.
                      ((1988  2 1)  -50 [cm])
                      ((1988  3 1) -200 [cm]))))

** New "Drain lateral" log to keep track of above.

** New 'RsqrW' model for weighted opservations.

** The 'B_C' and 'M_C' hydraulic models now takes an 'l' parameter.

** Some fixes in weather support, might have broken stuff.

** Some prototype work on the GUI.

* 5.12

** "AOM Pools" and "Organic Matter" log models should work again.

Individual pools (SOM1, SOM2, SOM3) is refered to by name (SOM-SLOW,
SOM-FAST, SOM-INERT) rather than numbers.

** New "polygon" soil zone model.

When using 2D geomerties you can now specify soil zones with arbitrary
polygons. Specifically, the "zones" Soil parameter now accept "zone"
models. The "box" model works as before, but the "polygon" model
allows you to specify an arbitrary polygon. E.g. 

    (defcolumn Andeby default
      (Soil (horizons (-20 [cm] Ap) (-2.5 [m] C))
            (zones ((polygon (outer (0 0) (0 10) (-10 0))) HorX)
                   ((box  (top finite -3 [cm]) (bottom finite -24 [cm])
                          (left finite 0 [cm]) (right finite 25 [cm]))
                    HorY))

** You can now plot 2D soil type profile.

To plot the soil type profile of the column Andeby: 

    (defgnuplot PlotMe profile
      (where "andeby-soil.pdf")
      (column Andeby))
  
    (defprogram RunMe gnuplot
      (graph PlotMe))
  
    (run RunMe)

** New "growth" root density model.

Use it like this:

    (sow ("Spring Barley" (Root (rootdens growth))))

Or for row crops, like this:

    (sow ("Spring Barley" (Root (rootdens growth (row_distance 80 [cm]))))
         ("row_width" 80 [cm]))

The models works like this:

When the root mass is growing, new roots are distributed equally in
the root zone up to DensRtTip (default 0.1 [cm/cm^3]), and any
additional roots are distributed according to the existing root mass.

When the root mass is shrinking, roots are lost proportional to the
existing root mass.

Unlike the existing models, this model is dynamic, that is, the
spatial distribution of the roots will depend on the history. If most
root mass is added while the root zone is small, the roots will be
concentrated in the initial root zone. If the root mass is mainly
added when the root zone is fully developed, the roots will be more
equally distributed within the root zone.

** Bug fixes.

* 5.08

** Bug fixes, mostly affecting variable timesteps.

* 5.07

** New "vector" gnuplot model.

This model will plot 2D flux values as a vector field.

  (defgnuplot MyVector vector
    (when 2000 1 1) ; At point
    ;; (begin 2000 1 1) (end 2001 1 1) ; Sum in interval.
    (factor 0.4) ; Adjust lenght of vectors.
    (legend none)
    (file "matrix_water_flux.dlf"))
  
* 5.06 

** Numeric adjustment of water movement and solute transport.

* 5.05 

** New 'stabilizing_method' in the Mollerup transport method.

** New 'max_rain' weather parameter.

Set this to decrease timestep under heavy rain.

* 5.04 

** Bug fix in secondary domain sorption.

* 5.03 

** Stabilizing secondary domain solute transport.

* 5.02 

** Numeric adjustment of water movement.

* 5.01

** New conditions 'daily_precipitation_above' and 'daily_air_temperature_above'.

* 5.00

** Bug fixes.

** New 'solubility' and 'solubility_infiltration_factor' parameters.

In the default chemical model. These control the maximum concentration
in water that enters the soil.

** New 'Rsqr' summary model.

Will compare simulation results with a data set and calculate R^2.

** New 'minimize' program.

Allow you to find the parameters that constitute the best fit.

** New 'r_pore_min' horizon parameter.

This parameter controls the division of sorption sites between primary
and secondary domain.

** New 'B2M' and 'M2B' biopore log variables.

These will allow you to log water entering and leaving the biopores
seperately.

** New units [fraction dry soil] and [dry soil ppm] for logging.

To log content as fraction of dry soil.  Just using [ppm] only works
for mediums with the same mass/volume ratio as water.

** New 'profile' xysource model.

** New 'valid' parameter in the 'expr' xysource model.

** Restore model sample to reference manual.

* 4.99

Bug fixes.

* 4.98

Bug fixes.

* 4.97

Bug fixes.

* 4.96

** New TempOffset weather parameter.

Works like PrecipScale.

* 4.95

** Restored 'TempOffset'.

* 4.94

** Bug fixes.

* 4.93

** More flexible weather data handling.

*** The 'simple' and 'none' weather models are gone.

The new 'const' model (see later) can sometimes be used as
replacement.

*** The 'default' weather model is rewritten.

There are probably many new bugs, and the error messages are likely
less informative.  However, it is more flexible as described below.

**** More fine grained weather data available weather files.

You can now specify "Minute", "Second", and even "Microsecond" in the
weather files.  If you have set 'minimal_timestep', Daisy will try to
make the timesteps match the weather data.

You can also specify time with a "Date" (or "time") column.  The
format is one of 

  yyyy-mm-dd
  yyyy-mm-ddThh
  yyyy-mm-ddThh:mm
  yyyy-mm-ddThh:mm:ss
  yyyy-mm-ddThh:mm:ss.uuuuuu

depending on the desired resolution.

All weather data describe the preceding time step, except for daily
data which describe the specified date.  So

  1999-02-05

will specify the weather for the fifth of February, 1999, while 

  1999-02-05T04

will specify the weather for the the hour between 3 and 4 the same day
(assuming hourly data).

**** Keywords, columns, and parameters interchangeable.

All numeric data about the weather files can be specified as keywords,
columns, or parameters.  For example, to run with a constant vapour
pressure of 1000 Pascal, you can specify

  VapPres: 1000 Pa

in the weather file header.  Or alternatively, as a parameter:

  (weather default "dk-taastrup.dwf"
    (VapPres 1000 [Pa]))

Doing the later will overwrite any value specified in dk-taastrup.dwf.

Data that are not simple numbers (like Surface or PrecipCorrect) can
be specified as keywords or parameters.

**** Single number allowed for PrecipCorrect and PrecipScale.

If you specify a single number, it is assumed to represent the whole
year.

*** New 'table' model.

The new 'table' model differs from the 'default' model on three points.

1. The file name is specified with an explicit 'file' parameter.  E.g. 

  (weather table (file "dk-taastrup.dwf"))

2. The 'missing_years' parameter is not supported.

3. The 'table' model can be used with the 'combine' and 'time' models,
explained below.

*** New 'const' model.

This is identical to the 'table' model, except that the 'file'
parameter is not accepted.  This means weather data must be specified
as parameters.  The value will therefore be constant, except the
radiation which will still follow a day cycle based on relative
extraterrestrial radiation.

*** New 'combine' model.

The 'combine' model will allow you to combine multiple sources of
weather data.  It has a single parameter 'entry', which is a list of
weather sources entries.  Each entry have 'begin', 'end', 'use', and
'source' parameters.  The 'begin' and 'end' parameters specify the
time interval where we should use this particular source.  The 'use'
parameter specified the data we should extract from this particular
source.  The particular value 'Any' specifies that we should use all
the data provided by the source.  Finally, 'source' specifies the
source.  This can be any weather model but 'default'.  For each type
of weather data, Daisy will examine each entry in sequence, and use
the first match.  If the 'begin' and 'end' parameters are unspecified,
the entry will cover from the start to the end of the simulation.  The
default value for 'use' is 'Any'.

Fo example, we might want to use weather data from the file
"dk-taastrup.dwf", except for precipitation during the summer, where
we instead use data from file "dk-taastrup-hourly.dwf".   

  (weather combine
           (entry ((begin 2006 6 1 0)
                   (end   2006 9 1 0)
                   (use Precip)
                   (source table (file "dk-taastrup-hourly.dwf")))
                  ((source table (file "dk-taastrup.dwf")))))

*** New 'time' model.

The time model allows you to use weather data from a different period.
E.g.

   (weather time (from 2002) (to 2006)
            (source table (file "dk-taastrup.dwf")))

will use four years old weather data in the simulation.  Note that all
weather data is mapped, so 2001 data will be used for 2005, and 2003
data for 2007.  It can be combined with the 'combine' model to map
only a selected period.  For example

  (weather combine
           (entry ((begin 2006 1 1 0)
                   (end   2007 1 1 0)
                   (source time (from 2002) (to 2006)
                           (source table (file "dk-taastrup.dwf"))))
                  ((source table (file "dk-taastrup.dwf")))))

will use data from "dk-taastrup.dwf", with the exception that 2002
data will be used for 2006.

Note that while dates until the 28 of February will be an exact match
for the two specified years, other dates may be one day off due to
leap years.

If you don't want to map year to year, you can specify 'offset'
instead of 'from' and 'to'.

  (weather time (source table (file "dk-taastrup.dwf"))
           (offset (days 1) (hours 0) (minutes 0) (seconds 0) (microseconds 0)))

This will use yesterdays weather data.  So will this

  (weather time (source table (file "dk-taastrup.dwf"))
           (offset (days 1)))

as all the 'offset' parameters default to zero.  There is no 'year'
parameter to offset, since that is not a well defined length of time.

* 4.92 

** Bug fixes and code cleanup.

* 4.91

** Bug fixes and code cleanup.

* 4.90

** New "log_prefix" top level parameter.

All output files will have this added to the front of their names.
Can be used as an alternative to the "directory" parameter, but
remember the directory seperator.  E.g. 

  (log_prefix "C:/foo/")

to put all log files in "C:\foo".

** Fixed OpenMI bug.

** New 'reset_offset' column source parameter.

This allows you to force time plots to start at zero.

** New 'sequence' action.

This is just like activity, except that all the action list is within
a 'do' parameter.

** Some work on ice.

It doesn't really work yet, though.

** Allow plot of boundaries with xysource.

** Prevent rain from being below 0.1 degrees.

** New "source" groundwater model.

Reads data from a ".ddf" file, unlike the old "file" model.

** Don't infiltrate chemicals when negligible water enter the soil.

* 4.89

Bug fix affecting OpenMI.

* 4.88 

Bug fixes.

* 4.87

** FAO_PM pet model: 'use_wet' now default true.

Should lead to more evapotranspiration from wet surfaces.

** Reduced memory use during initialization by orders of magnitude.

** Dimension change in log files.

The dimension of most flux variables in the standard log files has
changed so it no longer contain the log timestep.  For example,
logging "Field water" weekly would have the unit for fluxes be mm/w,
while logging monthly would have the same units be mm/m.  Now it will
be mm.  The actual values will be the same, the interpretation is
"amount of water since last we logged".

The advantages of this new interpretation are that 1) We avoid weird
units, 2) it works with intervals that have no nice symbol (like
biweekly), and 3) it works with variable timesteps.

Technically, internally in Daisy most flux unit are per time
(independent of the actual timestep used by Daisy).  When a log file
parametrization specifies (handle sum) this value will be multiplied
with the length of the timestep.  What happens now is that the unit is
also multiplied be the unit of the timestep [h].

If you really want the dimension of flux variables to be per time,
specify (handle average) instead, which will divide the value and unit
by the total time between log entries in hours, returning to the
original unit.  You can then tell Daisy to convert this to another
timestep by specifying 'dimension' explicitly.

** Variable timesteps.

Yo can now allow Daisy to temporarily decrease timesteps in order to
achieve more stable results.  You do this by setting the parameter
'minimal_timesteps', like this:

  (minimal_timestep (microseconds 1))

One microsecond is the smallest possible timestep.  The 'timestep'
parameter specifies the largest timestep.  There is no hard upper limit
here, but many models will likely fail to work with timesteps larger
than one hour. 

By default, 'minimal_timestep' is set to the same value as 'timestep',
thus forcing the models to use fixed timesteps.

This variable timestep affects all Daisy processes, unlike the already
existing variable timesteps internal to the water movement and solute
transport processes.

** Drain component and biopore tertiary model may limit timestep

If you specify drain pipes with the 'Drain' column parameter, or
specify macropores by setting the Tertiary parameter to the 'biopores'
model, the timestep may be limited by the source/sink terms.  Daisy
will calculate an initial estimate for the source/sink terms, and try
to limit the timesteps so that the water change in each numerical cell
stay "reasonable" within the timestep.

What is reasonable is determined by the 'max_sink_change' parameter in
SoilWater.  It is given as percentage of the difference in water
content between wilting and saturation point.

  (SoilWater (max_sink_change 1 [%]))

The default value is 10%.

** Chemistry may also limit timestep.

This is again based on the matrix water sink term from the drain
component and biopore tertiary model.

You can specify acceptable loss relative to the total content
(including sorbed forms), the solute content alone, or the content of
the secondary domain (if applicable).  The parameters, together with
their default values, are specified like this.

  (Chemistry original
             (max_sink_total 50 [%])
             (max_sink_solute 90 [%])
             (max_sink_secondary 150 [%])
             (min_sink_total 1 [%]))

The last parameter, 'min_sink_total' overwrites the other, meaning
that 1% of the total content of a cell will always be allowed to
escape through the sinks.           

** New 'microsecond' resolution.

Daisy can now track time on microsecond basis if so desired.  You can
specify microseconds all the same places where you used to be able to
specify seconds.  For example, to get microseconds in log files, specify 

  (log_time_columns year month mday hour minute second microsecond)

** New 'Time' log.

This log will provide information about the timestep.

** The 'none' chemistry is now build in.

So there is no longer any reason to load 'chemistry.dai' just in order
to disable chemistry.

** Biopore wall resistance change.

The R_primary and R_secondary parameters are gone, instead we have
K_wall_relative, which denote the conductivity of the biopore wall
relative to the surrounding matrix.

** Water movement on the x surface with 2D simulations.

The water movement on surface within the system being simulated is now
controlled by a new Surface parameter LocalDetentionCapacity.  If the
average ponding on the surface is above this threshold, all part of
the surface will have the same ponding.  Otherwise, if the ponding
locally is above this threshold, excess water will be distributed
evenly to the rest of the surface (limited by the threshold).

** Tertiary flux below surface now approximated.

The 'biopores' tertiary model didn't use to calculate a flux below
surface.  It still doesn't but it find an approximate value based on a
mass balance that ignores the storage capacity of the biopores.  The
values will likely be very much of for individual timesteps, but
averaged over a long period (long enough that the storage capacity
become an insignificant part of the mass balance), the approximation
should work well.

** New 2D plot facilities.

*** Content.

Assuming 'soil_water_content.dlf' list water content in all cells, the
following small setup file will generate a 'daisy.gnuplot' file, which
will show a 2D representation on the screen when feed into the gnuplot
program.

  (defgnuplot sample soil
    (file "soil_water_content.dlf")
    (when 2000 5 28))
  
  (defprogram showme gnuplot
    (graph sample))
  
  (run showme)

The 'type' parameter allows you to select between 'block', 'smooth',
'surface', and 'contour' plots, and the 'top', 'bottom', 'left' and
'right' parameters allows you to zoom into a rectangular subset of the
soil..  More information about generating plots can be found in the
tutorial, and more details on the 'soil' model can be found in the
reference manual.

*** Flux

You can now plot horizontal flux as a function of depth, or vertical
flux as a function of the x-position.  Here is an example that plot
horizontal flux at two different distances from the drain.

(defxysource hor50 flux
  "Horizontal flux at x=50."
  (file "matrix-water-flux.dlf")
  (x 50 [cm])
  (when 1999 9 28))

(defxysource hor150 flux
  "Horizontal flux at x=150."
  (file "matrix-water-flux.dlf")
  (x 150 [cm])
  (when 1999 9 28))

(defgnuplot plothor xy
  "Horizontal flux at x=50 and x=150."
  (source hor50 hor150))

(defprogram showit gnuplot
  (graph plothor))

(run showit)

* 4.86

** No externally visible changes.

* 4.85

** Made albedo logable.

* 4.84 

** Surface decompose rate now defaults to canopy dissipation rate.

It use to default to soil decompose rate.

** Drain moved from Movement to column.

(defcolumn Example default
  (Drain lateral
         (L 18 [m])
         (pipe_position -110.0 [cm]))
  (Groundwater aquitard
               (K_aquitard 0.050 [cm/h])
               (Z_aquitard 200 [cm])
               (pressure_table (file "Rorrende_88.gwt"))))
* 4.83

** Made 'K_at_h' with with the 'B_C_inverse' hydraulic model.

* 4.82

** New LAIfactor permanent vegetation parameter.

This is used for scaling LAI.  For example

  (Vegetation (Coniferous (LAIfactor 2.0 [])) )

will use the Coniferous parameters, with twice as much LAI.

* 4.81

** Many speedups.

** The default weather model will now distribute daily rain better.

Rain intensity below the value of 'minimum_precipitation' parameter
will be applied beginning at midnight at the rate specified by the
parameter, followed dry weather for the rest of the day.

The default value (0.83 mm/h) is the average hourly rain intensity at
the Taastrup agrometeorological station.

The effect of is that transpiration and soil evaporation increase, and
evaporation from ponded and intercepted water decreases.

** Most management actions are now instantanious.

They used to take one timestep (1 hour).

For example 

  (wait (at 1987 4 5 1))
  (fertilize (pig_slurry (first_year_utilization 100 [%]))
             (equivalent_weight 100 [kg N/ha]))
  (sow "Grass")
  (sow "Spring Barley")

will now both fertilize and sow the two crops during the same hour, it
used to take three hours to complete all three actions.

This includes irrigation, even when a irrigation duration is
specified.  For example

  (wait (at 1987 6 5 1))
  (irrigate_overhead 2 [mm/h] (hours 3)
                     (solute (NH4 1 [g N/l])(NO3 2 [g N/l])))
  (fertilize (mineral (weight 80.0 [kg N/ha])
                      (NH4_fraction 0.5 [])))

will start the irrigation and then immidiately fertilize in the first
hour.  The irrigation will then continue for two more hours, without
need for further manager involvement.

This affects the 'activity' action model.  The change was made to make
management more independent of the selected timestep, and because the
old semantics could be problematic to get right.

** New EpFacWet canopy parameter.

Specifies K_c for wet surface.  By default it is identical to EpFac.

** The FAO_PM reference evapotranpiration model used actual albedo.

Not it used the reference value.

** The FAO_PM model will now use wet surface reference.

** New biopores tertiary model parameter 'active_msg'

Set this to 'cell' (old style), 'range' (new default) or 'none' to control
message when biopores are activated or deactivated.

** New hydraulic model 'B_C_inverse'.

This is an version of the Burdine/Campbell model where the 'b' and
'h_b' parameters are found from the soil wilting point and field
capacity.

The 'b' and 'h_b' parameter values will be written to the daisy.log
file.

(defhorizon Ap FAO3
  "Andeby top soil."
  (clay 8.0 [%]) (silt 10.5 [%]) (sand 81.5 [%])
  (humus 1.12 [%]) (dry_bulk_density 1.5 [g/cm^3])
  (C_per_N 11.0 [g C/g N])
  (hydraulic B_C_inverse 
             (K_at_h -100.0 [cm] 1e-6 [m/s])       
             (Theta_wp 0.03 [])
             (Theta_fc 0.2 [])
             (Theta_sat 0.4 [])))

** New 'sorption' reaction model.

Faster than 'equilibrium', more flexible than 'adsorption'

** The 'aperture' secondary model take new 'use_secondary' parameter.

Set this to false to make 'aperture' only affect hydraulic
conductivity, without introducing a secondary domain for solute
transport.

** Fix output for gnuplot 4.4.

** Chemicals may now decompose when part of the litter pack.

** Bug fixes.

* 4.80

** Ported to Linux/AMD64/GCC4.4

** Bug fixes.

* 4.79

** Potential evaporation changes.

The default value of the Surface parameter 'EpFactor' has been
adjusted to 0.6 from 1.0.  The default value for the crop parameter
'EpFac' has been adjusted to 1.20, which mean that a full canopy
(LAI=5) will continue to have a Kc close to 1.15.  But for Danish
conditions, spring and autumn evaporation will be signficantly lower.

This reflect measurements in Kjaersgaard et al, 2008 and other places.
Autumn Ep will likely still be too high, unless you activate the new
Litter parameter described below.

** New 'Litter' parameter in the default colmun model.

The 'litter' component has three possible values.

*** 'none' (default)

Crop residuals and organic fertilizers have no effect on water and
solutes on the surface.

*** 'permanent'

A permanent litter layer is assumed, it covers the entire surface, can
store water and solutes, and may affect albedo and surface
evaporation.  It does not change over time.  This replaces the
'Litter' parameter of the permanent vegetation model.

*** 'residue'

The dry matter weight of organic matter on the surface determine the
fraction of the surface covered by litter, as well as the water
holding capacity of the litter pack.  Albedo and surface evaporation
is affected on the part of the surface covered by litter.

Two build-in parameterizations are available, namely 'Millet' and
'Maize'.

** You can now simulate organic matter without mineral nitrogen.

The default organic matter model will assume unlimted NO3 and NH4 if
the chemistry module is not tracking those.

* 4.78

** Fix answers to exercises.

* 4.77

** Bug fixes and cleanup

* 4.76

** Removed duplicate Mineralization from Field nitrogen.

* 4.75

** New Drain parameter

You can now specify drainage independent on groundwater model.

(defcolumn Example default
  (Movement original 
            (Drain lateral
                   (L 18 [m])
                   (pipe_position -110.0 [cm])))
  (Groundwater aquitard
               (K_aquitard 0.050 [cm/h])
               (Z_aquitard 200 [cm])
               (pressure_table (file "Rorrende_88.gwt"))))

This will eventually replace the old 'pipe' groundwater model.  The
parameters have the same names, but those pertaining to the lower
boundary are in 'Groundwater aquitard', and those pertaining to the
drain pipes are in 'Drain lateral'.

** The 'Field nitrogen' log model simplified.

NO3 and NH4 are now combined as "mineral nitrogen".  Also,
documentation has improved.

** The 'Jarvis99' colloid generation reaction can now depend on tillage.

The new 'Mmax_tillage_factor' allows tillage age to affect Mmax.  The
'tillage_replenish_all' parameter set Ms to Mmax after tillage.

** New 'pressure' model for finding average conductivity.

In the Mollerup uzrect model, the K_average parameter can now take a
'pressure' value, which allows fast transport away from saturated and
near soil, even if the surrounding soil is very dry.  The
K_average_horizontal parameter has been removed.

** Changes to stomatacon models.

Water stress effect is now part of the stomatacon models, rather than
a seperate parameter of the Farquhar photosynthesis model.  And the
water stress parameters have been renamed to beta and delta (from k
and l), to match convetional names.

** Improvements to the reference manual and updates to the tutorial.

** Bug fixes.

* 4.74

** New 'area' column parameter.

When you log multiple columns to the same log file, the values will be
weighted with this parameter.  Example: If you have one column
representing 1 hectar with a percolation of 6 mm/d, and another column
representing an area of 4 hectars with a percolation of 1 mm/d, the
total percolation will be 

  1 [ha] * 6 [mm/d] + 4 [ha] * 1 [mm/d]
 -------------------------------------- = 2 [mm/d]
            1 [ha] + 4 [ha]

The parameter does not affect simulations with only a single columns,
nor log files where you have specified the column (with the "column"
paramater).

** Allow using base names for logged chemicals more places.

Specifically, biopores.

** More units and expressions

** New 'accumulate' plot parameter

** dk-soil.dai

The "Jyndevad Ap" and "Jyndevad C" now parameterised for Brooks and
Corey hydraulic model.

** New 'pipes' tertiary model.

This works like 'pipe' groundwater model, but can be combined with any
lower boundary.

** New 'K_average_horizontal' parameter to the Mollerup uzrect model.

* 4.73

** More work on solving SVAT loops

The 'MaxTdiff' and 'MaxEdiff' parameters moved form the SSOC svat
model to the default bioclimate model.

** FAO_PM pet model can now calculate evapotranspiration for wet surface.

Set 'use_wet' to true, and 'rb' to desired boundary layer resistance.
The 'wet' surface value will be used as a maximal evapotranpiration,
while the 'dry' surface value will be used as a maximal transpiration.

* 4.72

** More Farquhar and SSOC work.

* 4.71

** More Farquhar and SSOC work.

* 4.70

** More Farquhar photosynthesis logging.

* 4.69

** Better logging for Farquhar photosynthesis.

You can now log photosynthesis for shadow, sunlit and resereved leaves
seperately.

** Fixed bug in Farquhar photosynthesis.

* 4.68

** New 'K' parameter in the 'pressure' secondary domain model.

The value of this paramater will be used as the soils hydraulic
conductivity, whenever the soil water pressure is above the limit
specified by the 'h_lim' parameter and the hydraulic conductivity
specified by the hydraulic model is lower.  Using this parameter is an
alternative to using a conventional bimodal hydraulic conductivity
model.  The default value of 0.0 is appropriate if you have already
specified a bimodal hydraulic conductivity model.

** New 'cracks' secondary domain model.

This is similar to the 'pressure' model, except that both 'h_lim' and
'K' are calculated from two new parameters, 'aperture' and 'density',
describing the median aperture and density of the soil fissures.

  (defhorizon C Ap
    "Andeby C horizon with cracks."
    (secondary_domain cracks
                      (aperture 100 [um])
                      (density 10 [m^-1])
                      (alpha 0.001 [h^-1]))
    (humus 0.12 [%]))

** New 'extern_subsoil' action model.

This allows an exteral model to control subsoil irrigation containing
arbitrary solutes.

* 4.67 

** Bug fix affecting OpenMI interface.

* 4.66

** New 'z_0b' parameter in SSOC SVAT model.

Increase surface roughness to force more turbulence and decrease canopy
temperature, especially useful for very low wind situations.

** Cleanup and bug fixes.

* 4.65

** Finished mixing layer.

Most people will see this as NH4 infiltration becomming slower.  You
can now have an equilibrium reaction on the surface.  The
'test-colloids.dai' file in the sample folder shows how to use this to
generate colloid bound chemicals on the surface.

* 4.64

** More functionality to the surface mixing layer.

Sorbing chemicals spread on the surface will now sorb to the soil in
the mixing layer.

** New 'bound_release' reaction.

This allows sorbed chemicals to be released with colloids from the
mixing layer.  See 'sample/test-colloids.dai' for an example.

** Code cleanup.

** Bug fix: Permanent vegetation should now work again.

* 4.63

** Bug fixes.

* 4.61

** More log models take 'unit' parameter.

"Field chemical", "Soil chemical", "Field nitrogen", and "Soil
nitrogen" now accept a "unit" parameter.  For exammple

  ("Field chemical" (unit [kg/ha]) (chemical colloid))

to get a summary of colloids in the field using [kg/ha] is the base unit.

** New 'Styczen88' and 'Morgan98' colloid generation models.

The 'Styczen88' model use momentum rather than kinetic energy for
colloid generation.  'Morgan98' is similar to 'Jarvis99' except that
it takes vegetation into account, but does not have a pool of readily
available colloids.

** New 'rainergy' and 'ponddamp' components.

These are used by the colloid generation models.  'rainergy' represent
models for kinetic energy in rain, and 'ponddamp' represent the
dampening effect of ponding on colloid generation.

** Bug fixes and cleanups.

* 4.60

** Cleanups and bug fixes.

* 4.59

** Cleanups and bug fixes.

* 4.58

** Major internal restructuring.

A lot of code has been modified, which potentially could mean many new
bugs.  

** The individual 'AOM', 'SMB' and 'SOM' pools are now components.

If you specify individual pools, you need to precede the parameters
with the keyword 'component, like this:

   (defam slurry organic
     (description "Average based on numbers provided by Torben Bonde, Danish
   Environmental Protection Agency, approximately 1991.
   Added by <sha@kvl.dk>, 2000.")
     (volatilization 0.15 [])
     (om (component (initial_fraction 0.72 [])
                    (C_per_N 100 [(g C/cm^3)/(g N/cm^3)])
                    (turnover_rate 2.0e-4 [h^-1])
                    (efficiency 0.60 0.60 [])
                    (fractions 0.0 1.0 0.0 []))
         (component (initial_fraction 0.18 [])
                    (turnover_rate 2.0e-3 [h^-1])
                    (efficiency 0.60 0.60 [])
                    (fractions 0.0 1.0 0.0 []))
         (component (C_per_N 11 [(g C/cm^3)/(g N/cm^3)])
                    (turnover_rate 1.0 [h^-1])
                    (efficiency 1.0 [])
                    (fractions 0.0 0.0 1.0 []))))

You may want to use some of the build-in parameterizations, which can
be found in the reference manual.

** The vernalization crop submodule has been made into a component.

This means you have to specify 'default' before the vernalization
parameters in the crop parameterisation, like this:

   (Vernal default (DSLim 0.33) (TaLim 5.00) (TaSum -50.0))
  
The crop parameterizations distributed with Daisy has been updated.

** Water transport model failures more discrete.

The number of times the water (or solute) models failed to find a
solutions is now reported only at the end of the simulation.  You can
still find them in the daisy.log file, and the new "Failure" log
paramterization will also include information about when the failures
occured in an easy parsable format.

** Sample taastrup.dwf extendended and renamed to dk-taastrup.dwf

In now contains weather data until 2008-10-31.  Note that the level of
quality control of the newer data has been reduced.

** New dk-taastrup-hourly.dwf sample file.

This file contains hourly weather data from Taastrup for the period
1995-03-24 to 2008-10-31.  The data is detailed enough to feed the
more advanced photosynthesis and svat models in Daisy, but the quality
control performed on the data is not impressive.

** New genweather.dai sample setup file.

This file demostrates how to aggregate and distribute weather data.

** The 'Field nitrogen' and 'Field chemical' logs now includes tertiary domain.

The content of the biopores is now included in the balance for these
two log parameterizations.

** New 'secondary' parameter to the 'equilibrium' reaction.

The 'secondary' parameter is a flag indicating that the equilibrium
reaction takes place in the secondar domain rather than the primary domain. 

** New 'colloid' parameter to the 'equilibrium' reaction.

The value of the 'colloid' parameter should be a name of chemical to
use instead of the soil for the 'rho_b' scope variable.  The idea is
to use it in situation where a solute may bind to colloids instead of
to imoobile soil.  See 'test-colloids.dai' for an example.

** Colloids work.

The 'colloid' chemical and the 'colloids' chemistry have been removed
from the 'chemistry.dai' file, and move to the 'test-colloids.dai'
file instead.  The later now contains an example with both colloids
and a colloid bound pesticide

** Bug fixes.

* 4.57

** Bug fixes.

* 4.56

** OpenMI now exports multiple columns as one element set.

* 4.55

** New GP2D program.

A new 'GP2D' program for calculating root density using a has been
added.  See 'test-GP2D.dai' for information about how to use it.

* 4.54

** New 'CO2' parameter for the default weather model.

This affects the FCC3 and FCC4 photosynthesis models.

** Added vegetables.

** New WP_ET in harvest log.

This is the weight of the storage organ divided by the total
evapotranspiration since emergence.

* 4.53

** Checkpoints are now defined as programs.

* 4.52

** New paramters for the 'sow' action.

*** 'plant_distance' is now an alias for 'row_width'

The intention for this is for PRD irrigation of potatoes, where the
drip line is withing the ridge, where we are more interested in the
soil gradient within the row, than between the rows.

*** New 'row_position' and 'plant_position' parameters.

These avoid the need to set them explicitly in the rood density submodel.

*** New 'seed' parameter.

This will overwrite the seed initial weight parameter

** New svat model 'SSOC'.

* 4.51

** Bug fixes and SVAT work.

* 4.50

** Proper logging of "direct to drain" biopores.

* 4.49

** More work on biopores and SVAT.

* 4.48

** Workaround for bug in FAO_PM.

Simulations results produced with any version of Daisy since 4.34
using FAO_PM for potential evopotranspiration are wrong.

** Ongoing work on tertiary domain, colloids, and SVAT.

* 4.47

** Fixed bug affecting 'rectangle' movement.

* 4.46

** More work on the tertiary domain.

See <http://code.google.com/p/daisy-model/wiki/TertiaryBiopores>.

* 4.45

** OpenMI 1.4 supported.

** Code cleanups and bug fixes.

** More work on the tertiary domain.

* 4.43

** User definable units.

You can now define your own units.  The most useful way to do this is
to derive from 'SIfactor', like this:

  (defunit [mph] SIfactor
    "Miles per hour.
  A speed unit in common use in the USA."
    (length 1)
    (time -1)   
    (factor 0.44704))

The factor is the number you should multiply with to convert to SI base
units, in this case [m/s].

The new unit can be used for any parameter that accepts the same SI
base.  For example, since irrigation is usually specified in [mm/h],
which has the same SI base, you can use the newly defined [mph] instead.

  (irrigate_overhead 1e-6 [mph])

You can also derive from an existing unit, like this:

  (defunit [mph] [mm/h]
    "Miles per hour.
  A speed unit in common use in USA."
    (factor 0.44704))

This saves you from specifying the dimensions (length and time).
However, note that the factor remains the same.  It should still
specify how to convert to the SI base units [m/s].

See the reference manual for a full list of SI dimensions and base units.

You can also define your own base units, independently of the eight SI
defined dimensions.

  (defunit EUR base "European currency.")
  
  (defunit EUcent factor
    "European cents."
    (base EUR)
    (factor 0.01))
  
However, there is little point in doing this, as none of the existing
Daisy code expects these units.

** More work in macropores.

* 4.42

** More work on units and macropores.

* 4.40

** More work an ABA, colloids, and macropores.

** More work on units.

* 4.39

** Start work on colloids.

** More work on defineable units.

** More work on PRD setups.

* 4.38 

** More work on tertiary transport.

* 4.37

** More work on tertiary transport.

* 4.36

** The 'SeedN' CrpN parameter removed.

It has been marked obsolete for some time.  Use the 'NCrop' Production
parameter instead.

** New 'Seed' parameter in the 'default' crop model.

The 'Seed' parameter govern initial growth.  The default model for
initial growth is based on a forced LAI function as before, and thus
take over some parameters from the Canopy submodel.  This means crop
parameterizations needs a small change.

Before:

 (Canopy (SpLAI  0.020 [(m^2/m^2)/(g DM/m^2)])
         (SpLAIfac (0.0 100) (1.0 100))
         (DSLAI05 0.5)
         (LeafAIMod (1.5 1.1)(2.0 0.3))))

Now:

 (Seed LAI
       (DSLAI05 0.5)
       (SpLAIfac (0.0 100) (1.0 100)))
 (Canopy (SpLAI  0.020 [(m^2/m^2)/(g DM/m^2)])
         (LeafAIMod (1.5 1.1)(2.0 0.3)))


Note that two additional changes has been made to this model:

1) The 'LeafAIMod' parameter is taken into account when deciding when
to switch from the forced LAI model to the leaf mass based model,
which will avoid mysterious jumps in LAI when switching model.  

2) The forced LAI is now treated like other kinds of forced LAI, which
mean the initial LAI can be found in the 'ForcedCAI' log variable, and
the mass based CAI can be found in the 'SimCAI' log variable.  The LAI
actually used can, as always, be found in the 'CAI' log variable.

** New 'release' seed model.

The 'LAI' seed model suffers from two drawbacks:

1) It does not take the amound of seed applied when sowing into
account, except indirectly through the value of the 'DSLAI5' parameter.  

2) The model becomes very sensitive to low radiation or stress during
the initial growth phase, as not enough dry matter may be produced to
be self sutainable.

The alternative 'release' model is based on a first order release of
assimilate from the seeds, which naturally scale with the amount of
seeds applied, and ensures a minimum amount of dry matter for leaves
and roots that doesn't depend on the weather and soil conditions.

  (defcrop "Spring Barley; Seed" "Spring Barley" 
    "A spring barley with initial growth governed by seeds."
    ;; We remove the old initial guess on seed N.  A negative number for
    ;; NCrop will force Daisy to calculate a new number based on seed
    ;;amount and N concentration.
    (Prod (NCrop -1.0))
    ;; Here we select the 
    (Seed release
          ;; We release 30% of the carbon per day.  It is first order,
          ;; so the total amount will be decreasing over time.
          (rate 0.3 [d^-1])
          ;; 15% of the seeds are water.
          (DM_fraction 0.85 [])
          ;; Almost half of the seeds is available as carbon assimilate.
          (C_fraction 0.4676 [])
          ;; The seeds also contain some nitrogen to get the crop started.
          (N_fraction 1.8 [%]))
    ;; The initial leaves are very thin.  Note that LeafAIMod is often
    ;; used for modifying old leaves, if so, you need to append the
    ;; original value for LeafAIMod here. 
    (Canopy (LeafAIMod (0.0 2.0) (0.4 1.0))))

  ;; When we sow the modified crop, we need to specify the amount applied.
  (sow ("Spring Barley; Seed" (weight 180 [kg w.w./ha])))

* 4.35 

** Now 'SoilWater' parameter 'max_exfiltration_gradient'.

This allows you to limit exfiltration.  It will be multiplied to the
hydraulic conductivity in the top node, to find the maximal
exfiltration.

* 4.34

** Bug fixes.

* 4.33

** Work started on different stomate conductance models.

** Bug fixes

* 4.32

** The "Tertiary" parameter moved to "Movement".

It now default to "old" for 1D simulations, which mean that macropores
will be enabled if clay + humus is higher than 5%.

The default is "none" for 2D simulations.

* 4.31

** New 'MaxWidth' root system parameter.

The parameter is similar to the MaxPen parameter, but denote the
maximal horizontal distance of root from the tuber, rather than the
maximal vertical distance.  The parameter is ignored unless you also
select a 2D root model.

* 4.30

** Primary, secondary and tertiary transport

New nomenclature. Intra-aggregate pores constitute the primary domain,
and inter-aggregate pores the secondary domain.  Both the primary and
secondary domain are dominated by cabalitory forces.  Pores large
enough for the cabalitory forces to become insignificant constitute
the tertiary domain.  The primary and secondary domain together also
constitute the soil matrix.

In practice, this means we can use Richard's Equation for the soil
matrix.  The convection-dispertion equation assumes equilibrium
between pore sizes, which we doesn't believe is a good assumption
betweeb the primary and secondary domains.  Therefore, transport in
each of those domains are considered seperate, with an exchange rate
between them.

For the tertiary domain we cannot even use Richard's Equation.

** Secondary

A new horizon parameter "secondary_domain" is used for enabling solute
transport in the secondary domain.  The default is that no secondary
domain exists.

** Tertiary

The old macropore transport parameters have been moved to 

  (Tertiary old)

This must be specified to get what used to be "default" behavior, that
is macropore flow for horizons with more than 5% clay.  You can also
specify the old macropore model directly, like this:

  (Tertiary old (macro none))

which has the same effect as 

  (Tertiary none)

The 'macro' parameter used to be located under SoilWater under the
same name, and was then moved to 'Movement'.

The 'old' tertiary model only works with the 'vertical' movement
model, therefore 'none' is the default for now.

* 4.29

** Bug fix: Soil MaxRootingDepth was mostly ignored.

* 4.28

** New 'row_width' parameter to the 'sow' action.

Specifying this will determine which rootdens model will be used,
unless specified in the crop parameterization.

* 4.27

** New 'every' conditions.

This one match the simulation at a user specified interval.  It takes
the same parameters as the 'timestep' parameter top level Daisy paramater.  

(every (years 0) (days 1) (hours 2) (minutes 3) (seconds 4))

** The optional 'step' parameter removed from the ..ly conditions.

This affects condition 'hourly', 'daily' 'weekly', 'monthly' and
'yearly' conditions.  They now always matches once at the last
timestep of the interval they are named after.

If you need other intervals, use the new 'every' condition.  Thus,
instead of (daily 2) you can write (every (days 2)).

** New 'minutely' and 'secondly' conditions.

These are true at the end of each minute and each second, respectively.

** Surface chemicals now decompose.

The two new paramaters 'surface_decompose_rate' and
'surface_decompose_halftime' determine how fast they decompose.  If
they are both unspecified, the 'decompose_rate' (or
'decompose_halftime') parameters are used instead.  No abiotic factors
currently affects the surface docomposition, not is any decompose
products produced (yet).

Set 'surface_decompose_rate' to 0 in order to get the old behaviour.

* 4 26

** New 'pluck' and 'cut' actions.

The 'cut' actions is identical to 'harvest', except that it will warn
if the crop dies.  The existing 'harvest' action will now warn if the
crop doesn't die.

The 'pluck' action allows you to selectively harvest storage organs
without cutting the crop down.

** Less harvest of dead leaves.

Dead leaves will no longer be harvested, unless you also have asked to
harvest live leaves with the 'leaf' harvest parameter.

** The 'crop_dm_over' condition now takes 'sorg_height' into account.

Thus, root fruits will not be considered part of the shoot dry mass.

** Bug fix: 'pipe' groundwater gave 50% too much deep percolation.

If you have a calibrated setup, multiply K_aquitard with 1.5 to get
the same results as before.

** Bug fix: No longer ptint toplevel description in log files.

* 4.25

** Bug fixes.

* 4.24

** New 'aquitard' groundwater model.

It is similar to the 'pipe' groundwater model, except that it does not
actually include drain pipes.  Like the 'pipe' groundwater model, the
'aquitard' model will create a dynamic groundwater table.  

The intention is the that the 'aquitard' groundwater model should be
used with the 'rectangle' movement model which supports explicit drain
pipes, but it should work fine with 1D simulations as well.

The parameters 'K_aquitard', 'Z_aquitard', 'h_aquifer', and
'pressure_table' are supported with the same meaning as for the 'pipe'
groundwater model.

** Bug fixes.

* 4.23

** New log models "Soil chemical" and "Field chemical" in "log-std.dai".

Use the "chemical" parameter to specify which chemical to log.  The
default file name will be "soil-" or "field-" followed by the name of
the chemical, followed by ".dlf".

** Name of the column automatically prepended to the log file names.

This happens for all the standard log parameterizations if you specify
the "column" parameter, but do not specify the "where" parameter.

** The 'column', 'crop', and 'chemical' parameters now appear in log headers.

The new 'parameter_names' table log model parameter list parameters
whose value should be included in the preamble.

** You can now log soil content as [ppm] with the 'array' select model.

That is, content of chemical as part per million of soil dry bulk density.

** You can now irrigate at a specific x interval.

The 'irrigate_subsoil' action model now takes a 'volume' parameter,
that allows you to specify a rectangle to irrigate.

  (irrigate_subsoil 5 [mm/h] (hours 3)
                    (volume box
                            (top finite -10 [cm]) (bottom finite -20 [cm])
                            (left finite 3 [cm]) (right finite  6 [cm])))

** You can now incorporate fertilizer in a specifuc x interval.

The new 'incorporate_fertilizer' action model allows you to specify
exactly where to put the fertilizer.

  (incorporate_fertilizer (mineral (weight 100.0 [kg N/ha])
                                   (NH4_fraction 0.5 []))
                          (volume box 
                                  (top finite -5 [cm])
                                  (bottom finite -15 [cm])
                                  (left finite 2 [cm])
                                  (right finite 8 [cm])))

** You can now specify different soil types in specific rectangle.

The new 'zones' Soil parameter contain a list of (volume horizon)
pairs, where the volume is specified as for the preious two news
items, and the horizon is a soil type as specified in the 'horizons'
parameter.  If multiple zones apply to a specific volume of soil, the
soil type of the first one listed will be used.  If no zone apply, the
horizons specified by the 'horizons' parameter will be used.

If you have specified a geometry where a cell overlaps multiple zones,
the soil type of zone containing the center of the cell will be used
for the entire cell.

(defcolumn Andeby default
  "The B.And farm, Andeby, 2002."
  (Soil (horizons (-20 [cm] Ap) (-2.5 [m] C))
        (MaxRootingDepth 60.0 [cm])
        (zones ((box (top finite -5 [cm]) (bottom finite -15 [cm])
                     (left finite 2 [cm]) (right finite 8 [cm]))
                Drain_padding)
               ((box  (bottom finite -5 [cm])
                     (left finite 2 [cm]) (right finite 8 [cm]))
                Drain_canyon)))

** Bug fixes.

* 4.22

** 2D Daisy

Daisy can now calculate two dimensional flow.  To enable, you must
specify the new Movement column parameter, like this:

(defcolumn Andeby2D Andeby
  "Two dimensional setup for Andeby."
  (Movement rectangle
	    (Geometry (zplus -2.5 -5 -10 -20 -40 -60 -80 -100  -120  -150 [cm])
		      (xplus 2 4 6 8 10 [cm]))))

The 2D movement model is called "rectangle", because it is called
"rectangle", because it is based on a rectangular grid, where you
specify the grid endpoints in both the z and x dimension.  

The same equations (Richard's, convection-dispertion, heat) is solved
for two dimensions, as is solved for one dimension.

Most stuff, including drain pipes and macropores, will work with no
changes for two dimensions.  

The one dimensional description is still available (and default) as
the "vertical" model.

** Inorganic nitrogen merged with chemistry.

The 'SoilNO3' and 'SoilNH4' column parameters have been removed.

New build-in chemical parameterizations for 'nutrient', 'NO3', and
'NH4' have been added.

The 'Denitrification' column parameter has been removed., and new
reaction models 'nitrification' and 'denitrification' have been added.

New build-in chemistry parameterizations for 'N' has been added,
compromising the chemicals 'NO3' and 'NH4', and the reactions
'nirification' and 'denitrification.

A new chemistry model 'multi' has been added, the main paramater
'combine' specifies a list of chemistry parameterizations to activate.

The default value of the 'Chemistry' column parameter is now a 'multi'
model, containing 'N' in its 'combine' parameter.

A new 'pesticides' chemistry model has been added to 'chemistry.dai',
it enables all the pesticides defined in that file.

To enable a single pesticide like 'DDT', you should contact a lawyer,
and specify in the column description:
 
  (Chemistry multi (combine (N (trace &old DDT))))

in the column setup.  To enable all pesticides (somewhat slower),
instead specify:

  (Chemistry multi (combine N pesticides))

You can now specify any chemicals in the 'solute' argument to the
irrigation actions.

The 'log.dai' and 'log-std.dai' libraries have been rewritten to
accomodate the changes.  Old log parameterizations involving inorganic
chemicals, including NO3- and NH4+, will need to be rewritten.

** You can now specify decompose products in the default chemical model.

(defchemical "Pesticide A" insecticide
  (decompose_halftime 20 [d])
  (decompose_products (7 [%] "Metabolit 1")
                      (5 [%] "Metabolit 2"))
  (adsorption linear (K_OC 2e6 [cm^3/g])))

Here 7% (weight relative) of the pesticide named "Pesticide A" will
become "Metabolit 1" when decomposed, and 5% will become "Metabolit
2".  The decompose products can be defined with 'defchemical' like any
other chemicals, and their fate followed by lsiting them in the
"trace" parameter in the default chemistry model.

  (Chemistry multi (combine (N (trace &old "Metabolit 1" "Metabolit 2"))))

** New 'GP2D' rooddens model for row crops.

** New 'device' gnuplot parameter for specifying output devise.

** More work on the 'rectangle' movement model.

* 4.21

** log_time_columns

New top level parameter.  Use it like

  (log_time_columns year month mday hour)

to avoid minute and second in the log files.

* 4.20

** More work on OpenMI.

* 4.19

** water_volume and water_interval:  New select models.

These models will log how much water corresponds to a specific
pressure in the specified volume or interval.

** min_root_density and min_root_crop: New log parameters

When logging a volume or interval, you can specify a min_root_density.
If the root density is below this thresshold, the contribution from
the cell will be scaled accordingly.  This can be used to log a
dynamic root zone.  If the new min_root_crop parameter is specified (and
different from "*"), the root zone of that crop only will be used.
Otherwise, all roots will be used.

** New 'location' column parameter.

This replaces the old 'x' and 'y' paramaters.  It is a list of (x y)
points, intended to identify the location of the column on an
externally defined map.  This is most relevant when coupled with a GIS
system, for example through the OpenMI interface.

** New 'extern' depth model.

This is used for specifying an externally defined pressure table with
tillage drainage.

** New OpenMI.dai library file

Together with the also new OpenMI_simple.dai and OpenMI_management.dai
files in the sample directory, they are intended to demonstrate how to
use the OpenMI interface to Daisy.

** daisy_dotnet.dll and daisy_OpenMI.dll included in distribution.

The daisy_dotnet.dll is a .NET wrapper to the C interface of daisy.dll.

The daisy_OpenMI.dll defines Daisy as an OpenMI component.

* 4.18

Reenabled GUI, more work on OpenMI.

* 4.17

More work on Farquhar photosynthesis and 2D transportation.

* 4.15

** Better support for Drag & Drop

When called from MS Windows with a single absolute file path as an
argument, Daisy will change the currect directory to the directory
where that file is located.  It provides a better default when using
drag and drop.

** Made default maximum crop uptake of nitrate faster.

* 4.14

** Setup program now adds links to the tutorial and exercises to Start menu.

** The path will now point to the installed library files by default.

The new parameter 'install_directory' will by default points to the
directory where you installed Daisy.

* 4.13

** The Daisy distribution is now packaged as a normal Windows installer.

For now, the installer will unpack the windows executable, the
reference manual, the parameter library, the sample files and the
source code.  

The installer will add several Daisy related entries to the start
menu, as well as a shortcut to the desctop.

More might be added later.  Currently everything will be installed
unconditionally.  Since the binaries takes almost all the space,
making some components optional does not seem worthwhile.  But it is
easy to do, so if you want it, let us know.

** Daisy now has a (minimal) GUI.

Basically, if you drag a file to the daisy.exe icon, output will be
shown in a window, with a bit of auxiliary information.  No real
functionality yet.

** The -nw command line option will disable the GUI. 

Adding (ui progress) to the setup file will do the same.

If you call daisy.exe from TextPad, or another programmers editor),
you should probably change the invocation to add the "-nw" command
line flag.  

In TextPad you do this with Configure -> Preferences -> Tools -> Daisy

Add -nw followed by a space (but without the quotes) before $File in
the "Parameters:" field.

** The -q command line flag will shut up Daisy.

Adding (ui none) to the setup file will do the same.

Daisy will then no longer report progress to the screen, but it will
still generate a daisy.log file.

** The daisy.dll is now part of the distribution.

And the daisy.exe file is a small wrapper around the DLL.  The
daisy.dll file, and several other DLL's (mostly GUI related) needed
for the minimalistic daisy.exe to run, are all included in the
distribution.

You can use daisy.dll from your own application using either a C
interface, defined in 'cdaisy.h', or a C++ interface, defined in
'toplevel.h'.  Both of those files can be found in the 'src'
directory (Daisy -> Explore -> src from the Start menu).

A C#/.NET interface will be added soon, as well as an OpenMI interface
for coupling with other OpenMI compliant hydrological models, such as
Mike/SHE.

** GNU/Linux users are screwed.

Well, not really, but a GNU/Linux binary is no longer autmatically
build as part of the release process.

* 4.12

** Reorganized pesticide handling

*** Tracing changes

In order for Daisy to track pesticides, they must now be explicitly
listed in the new Chemistry parameter in the column.  For example, 

  (defcolumn Andeby default
    "The B.And farm, Andeby, 2002."
    (Chemistry default 
               (trace Ioxynil DDT)))

will trace any Ioxynil and DDT sprayed on the field.  Earlier, this
happened automatically.  You can still spray other pesticides on the
field, they will just not be traced.

*** Definition changes

Parameters earlier placed in the 'solute' submodel is now placed
directly under chemical.  For example, instead of

  (defchemical DDT insecticide
    (decompose_halftime 2000 [d])
    (solute (adsorption linear (K_OC 2e6 [cm^3/g]))))

use this

  (defchemical DDT insecticide
    (decompose_halftime 2000 [d])
    (adsorption linear (K_OC 2e6 [cm^3/g])))

The paramterizations found in 'chemistry.dai' have been updated.

*** Loging changes

The standard log files in log.dai have been updated, look there for
inspiration for your own log files.

** More work on 2D transport, and the C, C# and  OpenMI interfaces.

* 4.11

More work on 2D transport and Farquhar photosynthesis, as well as C DLL,
C#, and OpenMI interfaces.  Code cleanup and bug fixes.

* 4.10

** New photosynthesis model: Farquhar

Implement a more advanced system than the default GL model.

** New radiation distribution model: sun-shade

Divides radiation input to leaves into sunlit and shaded.  Works well
with the new Farquhar photosynthesis model.

** Diffuse radiation now supported.

The global radiation can now be divided into diffuse and direct.  The
diffuse part can either be specified in the weather file, or estimated
by the difrad component in the bioclimate.  The sun-shade radiation
distribution model makes use of this new information.

** log-std.dai is now included from log.dai

** New way to specify column, crop, chemical, aom and pool to log.

Now use 

  (crop "Spring Barley")  

instead of 

  (set "$crop" "Spring Barley")  

This was already the case for the log parameterizations in
log-std.dai, but now also work with "log.dai" and "log-old.dai".

Old log parameterizations supporting the "set" attribute must be
updated to use the new system.  For most parameterizations it will be
enough to derive from "column" or "crop" instead of "table", and use
"${column}" or "${crop}" instead of "$col" or "$crop" in the paths.
If you never actually use the "set" attribute, it may be enough to
replace all instances of "$col" and "$crop" with "*".

** New examples from Czechia in the sample directory.

Kindly provided by RISWC, as part of the FertOrgaNic project, with
permission to distribute widely.

** Weather data from Jyndeval and Foulum in the sample directory.
** Soil parameterization from Foulum in dk-soil.dai in sample directory.

Kindly provided by DJF, as part of the FertOrgaNic project, with
permission to distribute widely.

** zplus can no longer be specified in 'Soil'.

It is now (only) a parameter of the 'Geometry' submodel in the
'vertical' movement model.

** Moved SoilHeat back to 'column' from 'movement'.

** "Soil Water Conductivity (log10)" 

Renamed "Soil Hydraulic Conductivity (log10)".

** The new themic logs new use SI units.

* 4.09

** New Folva, Triada, and Agria potato varieties in 'potato.dai'.

** New examples from Italy, Poland and Slovakia in the sample directory.

Kindly provided by the field sites, as part of the FertOrgaNic
project, with permission to distribute widely.

** You can now log actual hydraulic conductivity.

This is the hydraulic conductivity at the end of the time step.

** You can also log thermic conductivity and capacity.

** New 'expr' keyword in the select models.

This allows arbitrary arithmetic expressions in log files.

** New "Soil Water Conductivity (log10)" log model.

This logs log10 of the actual hydraulic conductivity.

** New "Soil Heat Conductivity" and "Soil Heat Capacity" models

** Bug fixes.

* 4.08

** Easier to mix integer and number objects with literals.

* 4.07

** New "emerge" management action.

This will force a crop to emerge the next day.

* 4.06

** You can now declare sequence parameters.

I really need to document user declared parameters someday.

* 4.05

** Bug fixes.

* 4.03

** Lots of restructuring.

In particular, you can no longer choose transport algorithm on a a per
solute basis.  They are now selected per column, through the new
Soltrans attribute.

Some of the changes affect simulation results, in the order of
magnitude 1-2%.

Removed the 'active_groundwater' parameter from 'OrganicMatter'.

** Organic matter is now a component.

In practice, this means that when specifying the OrganicMatter
parameter in the default column model, you need to insert the name of
the model for organic matter you want to use.  In practice, I
suggest you specify "original" in order to get the default parameters
and model.

Instead of 

  (OrganicMatter 
     (init (input 1400 [kg C/ha/y])
           (background_mineralization 30 [kg N/ha/y])))

write

  (OrganicMatter original 
     (init (input 1400 [kg C/ha/y])
     (background_mineralization 30 [kg N/ha/y])))

You can also use named parameterizations,  like

  (deforganic "low input" default
     (init (input 1400 [kg C/ha/y])
     (background_mineralization 30 [kg N/ha/y])))

and use them like

  (OrganicMatter "low input")

* 4.02

** Bug fixes and restructuring.

* 4.01

** New TempScale and TempOffset parameters to the default weather model.

Use

   (weather default "taastrup.dwf" 
            (TempOffset 1.0 0.0 0.0 0.0 0.0 2.0 
                        0.0 0.0 0.0 0.0 0.0 0.0 0.0 [dg C]))

to let the simulation use weather from "taastrup.dwf", except that
temperatures in Januare will be one degree higher, and June two degree
higher.   The argument to TempOffset should be 12 numbers, each
representing an temperature increase for all values for that month.

Use
   (weather default "taastrup.dwf" 
            (TempScale 1.5 0.0 0.0 0.0 0.0 2.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0))

to let the simulation use weather from "taastrup.dwf", except that
temperatures in Januare above zero degree will be 50% higher and below
zero degree 50% lower, and temperatures in June will be 100% higher.

If you use both, scale is applied before offset, as in

  T_used = T_file * scale_m + offset_m

where T_used is the temperature used by Daisy, T_file is the
temperature listed in the file, scale_m is the scale factor for the
current month as specified with TempScale, and offset_m is the offset
for the current month as specified with TempOffset.

** New 'soil_heat_capacity' and 'soil_heat_conductivity' number models.

These extract heat information from a horizon.

** New file 'sample-heat.dai' generates a table of heat properties.

It also shows how to use the 'number' library to extract static
information from a Daisy setup file.

* 4.00

** Bug fixes.

* 3.99

** New program "write".

It simple writes the specified text to the screen or a file.

** More work related to scoped variables and gnuplot.

I really need to write some documentation someday.

** Bug fixes.

* 3.98

** The 'flux_bottom' select model now takes a 'to' parameter.
** The 'flux_top' select model now takes a 'from' parameter.

This allows loging multiple intervals in the same log file.

** New DrainDrying.dai file in the sample directory.

The file demonstrate a setup to test the ability of a simulated soil
to drain water and to dry, which can be used when evaluating the
hydraulic conductivity parameters of the soil.

** New 'and', 'or', 'xor' and 'not' Boolean models.

** Bug fixes.

* 3.97

** Removed the `enable_water_stress' crop parameter.

Use (water_stress_effect none) instead.

** New 'table' action that reads management data from a ddf table.

See the tutorial section about plotting for a description of the ddf
file formar.

** More gnuplot works.

Mostly documented in the reference manual.

** The 'batch' program no longer open all files before starting.

This should allow longer batch runs with more log files.  It also
means that file problems will be discovered later.

* 3.94

** Many more gnuplots options, mostly documented in the reference manual.

** New convertion: [m/s] -> [cm/d]

* 3.93

** New gnuplot syntax, see the tutorial for updated examples.

* 3.92

** You can now turn off the gnuplot legend by specifying "none" as a placement.

* 3.91

** Convert [kPa] -> [pF]

* 3.90

** New "&old" keyword.

When you define a new parameterization, you can use the "&old" keyword
allows you to refer to the original value of a list attribute, so you
can prepend or append elements to the list.  Example:

  (defprogram "Foo Project" Daisy
     "Basic setup file for simulations in the Foo project."       
     ;; We always want a field based nitrogen and water balance. 
     (output "Field nitrogen" "Field water"))

  (defprogram "Foo Soil" "Foo Project"
     "Basic setup files for exmaining soil water and nitrogen."
     ;; We still want the field based balances, but add soil based balances.
     (output &old "Soil nitrogen" "Soil water"))

Here the value of the output parameter in "Foo Soil" will be 

  ("Field nitrogen" "Field water" "Soil nitrogen" "Soil water")

If we had written (output "Soil nitrogen" "Soil water" &old), the
value would have been

  ("Soil nitrogen" "Soil water" "Field nitrogen" "Field water")

The &old keyword can be places anywhere on the list, and even
repeated.  This is not very useful for the 'output' parameter, where
order doesn't matter, and duplicate entries will just confuse Daisy,
but it can be useful for other parameters.

** Batch programs are now more informative.

* 3.89

** gnuplot program now support errorbars.

If there are multiple datapoints for the same time spot, the mean will
plotted with errorbars of one standard deviation to each side.

* 3.88

** More flexible reading of ".ddf" files

*** source filters now match space padded values.

*** ".ddf" files now accept dates in dd/mm/yy format.

* 3.85

** Unstable!

Due to the implementation of the change below, many submodel may not
work correctly.  The symptom is that Daisy will crash! Please report
all such crashes to daisy@dina.kvl.dk, together with the daisy.log
file.

** You can now declare local variables in parameterizations.

For example, here we declare a specialization named 'column', of the
table log model.  The specialization has a new parameter also named
'column', declared with the 'declare' keyword.

  (deflog column table
    "A log table for a specific all for all columns."
    (declare column String
      "Name of column to log.  Use \"*\" to log all columns.")
    (column "*"))

The first argument to the 'declare' keyword is the name of the
parameter, the second the type.  The type can be one of the build-in
types, 'String', 'Number', or 'Integer', or the name of a component
(see the reference manual for a list of components).  For the 'Number'
type, you can specify a dimension right afterwards, in brackets.  Like
"Number [cm]".  The last argument is a documentation string.

After you have declared the parameter, you can set it just like any
other parameter.  Here we set it to the string "*".  We can later
refer to the value within a string, by using the "${" escape
sequence.  Here is an example where we refer to the value of the
column attribute:

  (deflog "Field nitrogen" column
    "Nitrogen input, output, transformation and content for the system."
    (where "field_nitrogen.dlf")
    (when hourly)
    (entries (number (path column "${column}" fertilized_NH4)
                     (spec column default fertilized_NH4)
                     (handle sum)
                     (tag "NH4-Fertilizer")
                     (dimension "kg N/ha/&"))

You can refer to any parameter with a String, Number or Integer type
this way, including the standard build-in parameters.

The effect for the end user is that you can now log "Field nitrogen"
for a specific column, by setting the new attribute:

  (output ("Field nitrogen" (column "Andeby")))

It is the intention that all the log models should eventually use this
new system, and the old (set "$col" "Andeby") system should go away.

WARNING:  Local variables are not saved in checkpoints, so if your
simulation setup depend on local variable, checkpoints may not work.

** You can now ask Daisy to normalize texture numbers.

If you set (normalize true) in any other horizon type than "default",
Daisy will normalize the texture numbers for you.  Otherwise, Daisy
will complain if the sum is not 1.0 [] or 100 [%].  

The "default" horizon always normalize,  but you should avoid using
that horizon model because the texture system is unclear.

** DOS text format input files supported undet Unix.

** New "timestep" condition for use in log files.

Use it like this

  (output ("Field nitrogen" 
           (when (timestep (and (month 3) (mday 31) (hour 23)) "y"))))

to tell daisy that this is yearly data.

Daisy will use this in log files to specify the dimension for flux
variables, e.g. "kg N/ha/y" to indicate leaching with a yearly
timestep.  Note that the actual value of the flux is not affected,
only apparent dimension.  Daisy will alway write the accumulated flux
since the last timestep logged.

** New convertion: [none] -> []

For plotting weather data.

* 3.84

** gnuplot support

The new 'gnuplot' program allows you to create command files for
gnuplot to create plots in encapsulated postscript (*.eps), Acrobat
Reader PDF (*.pdf), LaTeX (*.tex) format or be shown directly in a
window.  The data to be plotted can be read from Daisy log files
(*.dlf), Daisy weather files (*.dwf) or the new Daisy data files
(*.ddf).

See the new section about "plotting" in the Daisy tutorial for a
description. 

** You can now do simple calculation in Daisy input files.

Whenever Daisy expect a number in an input file, you can give it an
expression like those accepted by the 'arithmetic' source model for
gnuplot. 

* 3.79

** Made water and N stress the last entries of 'harvest.dlf'.

This makes it easier to reuse old spreadsheets, but also means their
position will be moved when you change the 'print_N' or 'print_C' flags.

** Fixed sized header for dlf files.

You can now ask for Daisy to use a header with a fixed number of lines
in ".dlf" files.  Just set the 'print_header' attribute to fixed.

  ("Field nitrogen" (print_header fixed))

This is useful for some post processing tools.

** soil_inorganic_N_above: new condition

You can use this new condition to test for the mineral nitrogen in the
soil, useful for simulating precision farming.

* 3.78

** Water stress days estimate was twice as large as it should be.  Fixed.

** harvest.dlf now contain water and nitrogen stress days

These columns should help fast identify why harvest is less than
expected.  

** You can now ask Daisy to combine harvest details.

If you set the 'combine' parameter to 'true', all DM, N, and C content
of the crop will be placed in the "stem" fraction of the harvest.dlf
file.  This is intended for silage, where you don't distriguish
between the various crop parts.

    (harvest "Grass"
             (combine true)               ;Silage.
             (stub 8.0 [cm])              ;Leave 8 cm stub.
             (stem 1.00 []))              ;Harvest everything above stub.


** 1991 version of MARKVAND irrigation implemented for spring crops.

If you want to simulate irrigation of spring crop according to the
MARKVAND decision support system, you need to input the file
"markvand.dai".  

  (input file "markvand.dai")

If you then have an existing management strategy for the crop, say
"maizeman", you can have an irrigated version with

  (while "maizeman"
    (MARKVAND (soil JB1)))

Where "JB1" is soil category using the Danish system.

Current limitations:

*** The code is mostly untested at this point.

*** No changes in MARKVAND since 1991 are implemented.

*** Only spring crops supported.

*** Simulated user input limited to emergence and harvest times.

*** Can't simulate less than perfect weather data.

The markvand submodel use the same weather data as the rest of the system.

*** Only JB soils described conveniently.

However, it is possible to create a description of a specific soil
from the templates in markvand.dai.

* 3.77

** doc fix

** Ep was overstimated for LAI around 1.

* 3.76

** Photosynthesis made a component.

In practice, this means that when specifying the LeafPhot parameter in
the default crop model, you need to insert the name of the model for
leaf photosynthesis you want to use.  In practice, I suggest you
specify "original" in order to get the default parameters and model.

Instead of 

  (LeafPhot (Fm 5.00) (Qeff   0.0500))

write

  (LeafPhot original (Fm 5.00) (Qeff   0.0500))

The standard crop parameterizations in daisy-lib.zip have been
updated.  

Later, it will be possible to use other models for leaf
photosynthesis, such as Farquhar.

** Bug fixes.

* 3.75

** New Soil border parameter.

You can now force Daisy to have numerical borders at some specific
depths, without specifying all such depth with zplus.  

You can use it like this:

  (defcolumn Andeby default 
    (Soil (border -50 -110 [cm])))

Here, there will be numerical borders at 50 and 110 cm depths.

If you try to log at depths that are not specified as numerical
borders (using the 'from' and 'to' parameters to the table log model),
you will get a warning that the results may be inexact.

** Minor fix to hypres.

** No longer warn about use of dk-horizon.dai.

** You can now log geometric average, with (handle geometric).

** Reenabled harvesting of "all".

* 3.74

Bug fixes.

* 3.72

** Assimilate production can now be affected by chronological age.

    (sow ("Spring Barley" (LeafPhot (DAPEff (0.0 1.0) (100.0 0.5)))))

Here, the Spring Barley will have full assimilate production right
after planting, decreasing linearly to half assimilate production 100
days after planting.

* 3.71

** dk-soil.dai move from lib to sample.

Also, minimal Jyndevad and Askov column parameterizations were added.

** Removed the set_subsoil_irrigation and stop_subsoil_irrigation actions.

Use irrigate_subsoil instead.

* 3.70

** Solute in overhead and surface irrigation was wrong unit.

** doc fixes.

* 3.69

** lysimeter bottom may work now.

Use 

  (Groundwater lysimeter)

in the column if you have a lysimeter below the soil (you must define
the soil down to the lysimeter).

** New "hydraulic" program.

This program print the retention curve and conductivity curve for the
specified hydraulic model.

  (run hydraulic (M_vG (l -1.56461)
                       (n 1.36539)
                       (alpha 0.064593)
                       (K_sat 4.04833 [cm/h])
                       (Theta_sat 0.386942)))

Note that you can not do this for the pedotransfer functions (hypres
and Cosby_et_al).  However, when you actually use one of the
pedotransfer models in a sumulation, the parameters of the underlying
model will be written to 'daisy.log'.

** field seperator now obeyed for time columns in table log.

** New log-sample.dai file in sample directory.

This file contain additional log model parameterizations, serving as
examples of how to create your own, geared towards comparing simulated
and measured data.

Since measured data tend to wary a lot, standardized files are less
useful. 

** Daisy will now choose the FAO_PM pet model by default if RelHum is given.

Rather than just VapPres.  

* 3.68

** h_aquifer for drained soil may now vary with time.

There is a new 'pressure_table' parameter that is an alternative way
to specify 'h_aquifer'.  The variable indicate the virtual groundwater
table that corresponds to the pressure in the aquifer.  This table may
be different from the actual table, because of the aquitard.  If you
drilled a hole through the aquitart down to the aquifer,
pressure_table is the height (from the surface) the water would level
in that hole.

There are three different ways to specify the pressure table:

  (Groundwater pipe (pressure_table const -250 [cm]))

This indicates a constant pressure in the aquifer.

  (Groundwater pipe (pressure_table file "example.gwt"))

Here, "example.gwt" should be a file indicating the pressure table in
the same format at used by the "file" groundwater model.  That is,
lines of year, month, day of month and height seperated by
whitespace.  The level between the specified days will be found by
linear interpolation.

  (Groundwater pipe (pressure_table PLF
                                    ((1986  1  1 23) -200   [cm])
                                    ((1987  9  9 23) -100   [cm])
                                    ((1988  1  1 23) -150.5 [cm])
                                    ((1999 10 10 23) -200   [cm])))
  
Here we specify the depth over time directly in the setup, rather than
in an external file.

** The 'batch' program now take a 'directory' paramter.

This is usueful to make the Daisy place log files for different
simulations in a batch run in different directories.  See batch.dai in
daisy-sample.zip for an example.

** Bug fixes.

* 3.67

** You can now control the effect of water stress on crop growth.

In the default crop model, the new "water_stress_effect" controls how
the assimilate production is affected by water stress.  There are thre
different such "wse" models that can be used:

none: No effect, similar to setting the now obsolete
  "enable_water_stress" flag to false.

full: Full linear effect between no stres = full production and no
   water = no production, which is what Daisy has always use.  This is
   the default value.

partial:  When water stress is 0.5, the production effect is specified
  by the y_half parameter, which should be a number between 0 (no
  production) and 1 (full production).

You can the wse model when define or sow the crop, like this:

  (sow (Grass (water_stress_effect partial (y_half 0.67 []))))

To see a table linking the water stress with the effect on assimilate
production without running a simulation, you can add this to your
setup file (at top level):

  (run wse (partial (y_half 0.67)))

** Daisy can now run multiple simulations sequentially.

See the new file "batch.dai" in the sample directory for an example. 

** New top level "run" parameter.

You can now seect among build-in programs to run.

If you specify the "run" parameter, the normal Daisy specific
parameters are disregarded, and the specified program is run.  Daisy
has the following build-in programs:

Daisy: Run a Daisy simulation.  If you use this, you must specify a
  Daisy parameterisation, as top level parameters will be ignored.

batch: Run multiple programs in sequence.

LaTeX: Produce the "components" part of the reference manual in LaTeX
  format. 

AM_table: Produce a table with the know fertilizer types.

wse: Produce a table shoping the water stress effect with for a
  specified model parameterisation.

* 3.66

** Shorter reference manual.

The horizon, select and weather chapters now each have a section
containing the common parameters for all models in that chapter.  This
saved 40 pages of repetitions.

** initial_Theta should work again.

* 3.65 

Bug fix.

* 3.64

** Almost all select models now handle average, min and max values.

The new 'handle' parameter to the select models take five values,
'average', 'min', 'max', 'sum' and 'current'.

The two later corresponds to '(flux true)' and '(flux false)', which mean
the 'flux' parameter is now obsolete.  Use 'handle' instead.

The 'pF' and the (obsolete) date related models do not obey the new
'handle' parameter.

The 'max', 'min' and 'average' select models have been removed, use
'number' instead with the relevant handle.

You need to update 'log.dai' because of this change.

** The interval select model can now log as "fraction of bulk mass".

You need to specify both the 'spec' parameter and the 'dimension'.  If
Daisy knows how to convert the units of the variable found with 'spec'
to 'g/cm^3', and how to convert a fraction into the dimension
specified by 'dimension' (e.g. [%]), it will do so.

** New abiotic_factor log variable in OrganicMatter.

This log the product of the water and heat factors, but ignores the
clay factor. 

** Name of SoilWater Theta and sink terms dimension changed.

From [cm^3/cm^3] to [].  The values are the same, so this should only
affect you if you have initialized Theta explicitly and supplied a
dimension.  You can now use both fractions [], percents [%] and parts
per million [ppm] for initialization.

** You can now specify the Theta_sat and Theta_res hydraulic parameters as %.

** You can now specify C in SoilNO3, SoilNH4 and SoilChemical as [ppm].

The old `solute_ppm' parameter has been removed.

** New Ms parameter in SoilNO3, SoilNH4 and SoilChemical.

This allows you to specify the content as ppm.  The old 'soil_ppm' 

** You can now specify pressure in Pa, hPa, and kPa.

** There is a new 'init-soil.dai' file in daisy-sample.zip.

This illustrates how to initialize soil content.

* 3.63

Bug fixes.

* 3.61

** In the irrigate actions, 'hours' now defaults to zero if 'days' > 0.

** Avoid duplicate references in reference manual.

** Also print version number when there are parse errors.

** You kan now log "a" and "L0" in the "Gerwitz+Page74" rootdens model.

Use the "Root form parameter" log parameterization.

** New experimental "Anders Pedersen" rootdens model.

This is Gerwitz and Page with 'a' as a free parameter.

Use it like this:

    (sow ("Spring Barley" (Root (rootdens "Anders Pedersen"
                                          (a_DS (-1.0 0.08 [cm^-1])
                                                ( 2.0 0.08 [cm^-1]))
                                          (q 1.2 [m])))))
* 3.60

** You can now give specific numbers symbolic names.

Example: 

  (defnumber max_rooting_depth const 60.0 [cm])
  
  (defcolumn Andeby default
    "The B.And farm, Andeby, 2002."
    (Soil (horizons (-20 [cm] Ap) (-2.5 [m] C))
        (MaxRootingDepth max_rooting_depth))
    (Groundwater deep))

This is useful if the same number is used in several places, and all
must be updated at the same time.  Or if you are running a number of
simulation where you always change a few specific parameters, you can
place these at the beginning of the file for convenience.

* 3.59

** More DOM sorption work.

* 3.57

** Reorganised chemical transformations.

** DOM sorption

These are not really documented yet, if you need to do advanced
trasnformations of soil chemicals (actually, anything beyond a
constant decompose rate) or you are working on dissolved origanic
matter, contact us.

* 3.56

** Some planting support.

You can now set the production_delay Harvest parameter without setting
last_cut.  This can be used to make a plant "rest" just after being
planted. 

* 3.55

** The heat related horizon parameters model to new HorHeat submodel.

There are normally not set.

* 3.53

** New Horizon models.

We now have horizons corresponding to the most common texture
classification systems.

(defhorizon AnExample USDA3
  "An horizon defined using the USDA texture classification system."
  (humus 1.6 [%])
  (dry_bulk_density 1.5 [g/cm^3])
  (clay 8.0 [%])
  (silt 10.5 [%])
  (sand 79.9 [%]))

Note that the horizon models all require the sum of the soil
constituents should be 1.0 (or 100%).  This may or may not include the
humus part. 

The old "default" horizon model is still available, and will normalize
the constitituents.  It used the USDA (United States Department of
Agriculture) classification system.  We suggest that you start using
the new models, in order to be explicit about the classification
system.

Other supported texture systems are:

USDA7: Same as USDA3, but the sand is further divided into very fine,
fine, medium, coarse and very coarse sand.

FAO3 and FAO7:  The Food and Agriculture Organization within the UN
use the same classification systems as USDA.

USPRA3 and USPRA4: The United States Public Road Administration are
unique in that their limit for clay is 5 rather than the 2 [um]
everybody else uses.  Sand may be devided into fine and coarse.

ISSS3 and ISSS4: The International Soil Science Society, their
classification is common in Denmark. Sand may be devided into fine and
coarse.

BSI3 and BSI7: The British Standards Institute classification is also
used by Massachusetts Institute of Technologu, (also knows as MIT3 and
MIT7) Both sand and silt may be divided into fine, medium and coarse.

DIN3 and DIN5: German standards, DIN3 is identitical to BSI3, DIN5 is
like BSI7 except that the silt isn't divided.

You can also specify the texture with your own limits, like this.

(defhorizon Ap numeric
  "A horizon with explicit texture classification."
  (limits 2 50 2000 [um])
  (fractions 8 10.5 79.9 [%])
  (humus 1.6 [%])
  (dry_bulk_density 1.5 [g/cm^3]))

The horizon here is identical to the one specified with USDA3 before.

* 3.52

Bug fixes.

* 3.51

** Organic fertilizer water no longer ignored.

The water content of organic fertilizer is no longer discared, but
counted as irrigation, either surface or subsoil.

* 3.50

** All crops give lower yields!

A bug in the code caused an underestimation of the leaf respiration.
You can expect 1 to 5 percent reduction of yields for all existing
parameterizations of the default crop model.

** The file groundwater model give better error message.

This happens if the water table move too high. 

* 3.49

** test.dai moved to daisy-sample.zip

** weather.dwf renamed taastrup.dwf and moved to daisy-sample.zip

** Daisy will now complain if you try to log fluxes not at interval borders.

** Each horizon now gets at least three numeric intervals.

** The checkpoints may now contain the "original" keyword.

A couple of crop variables have also been removed, so old checkpoints
will not run (not that they usually can run on newer version).

* 3.48

** Nitrification parameter moved.

This used to be a parameter of the default Column model.  It is now a
parameter of the default horizon model.  This means you can have
different nitrification parameters for different horizons.

Note that unless you have special knowledge, you should not set the
nitrification parameters explicitly.  The default values are usually
adequate. 

You need to fetch a new log.dai to get correct nitrification logs.  Or
you can change 

  Nitrification "*" 

to just 

  Nitrification 

(i.e. remove the "*") in you log models manually.

** active_underground and active_groundwater only set once.

They have been removed from the nitrification models and the
Denitrification submodel, so now they are only set in the
OrganicMatter submodel.  Setting them in OrganicMatter also affects
nitrification and denitrification.

By default, there is no mineralization, nitrification or
denitrification below the root zone.

** Log summaries can now be redirected to files.

There is a new "where" summary parameter.   It is mostly intended to
help me create a better test suite.

** Daisy will complain if the "stop" time is after end of weather data.

* 3.47

** Description string for activity action allowed.

The form 

  (defaction foo activity
     "Description of the foo activity.")

is now allowed.  It has been used in examples for some time, but
rejected by Daisy. 


* 3.46

** New top level stop command.

If set, this force the simulation to stop when the simulated time
reach (or exceed) the specified stop time.

For example, if you have this

 (time 2002 4 1 0)
 (stop 2003 3 31 23)

in your setup, the simulation will at most run one season.

** New "sample" directory.

The new daisy-sample.zip file contain various sample setups and
parameterizations, for use as inspiration for your own work.

** Bug fixes.

* 3.45

** Fixed bug with apparent mass balance errors after tillage operations. 

* 3.44

** nitrogen_stress moved to CrpN submodel.

You must update log.dai to see nitrogen stress.

** New nitrogen_stress_days CrpN parameter.

This is similar to water_stress_days, giving an accumulate impact
indicater for production loss due to nitrogen stress.

** New nitrogen_stress_limit partition parameter.

Setting this will cause all growth assimilate to go to the storage
organ when there is nitrogen stress in the reproductive phase.

* 3.43

** Preliminary Ryegrass and Wclover crop parameterizations.

Developed by Henning Høgh Jensen.  These should work in a multicrop
environment on clay soil.  

* 3.42

** Bug fixes.

** New carbon summaries: "Crop Carbon Balance" and "Total Carbon Balance".

* 3.41

** New "Vinterraps" parameterization by Christian Thirup.

This is a winter rape that give yields corresponding to Danish
averages. 

** Seed_C is now logged.

** New carbon summaries: "Litter Carbon Balance" and "Soil Carbon Balance".

** New "min_light_fraction" crop parameter.

You can now specify that a certain fraction of the crop in a mixed
crop field get exclusive access to the light.  For example

  (progn
    (sow "Spring Barley" )
    (sow ("Grass to grain" (min_light_fraction 0.2))))

Here, 20% of the light will be reserved for the grass, for the
remaining 80% the grass and barley will compete, with the highest crop
shadowing for the other.

This is intended to emulate small "patches" in the field, where one
crop dominates for some reason.

* 3.40

** Fixed bug when logging with "from" not equal to 0.

** Movement during tillage operations now logged.

The four OrganicMatter log variables, tillage_Org_N_top,
tillage_Org_C_top, tillage_Org_N_soil, tillage_Org_C_soil can now give
information about C and N removed from the top and added to the soil,
which makes it possible to create a full organic matter balance that
does not include the surface or the plowing layer.  The "N Balance"
and "Carbon Balance" log parameterizations have been updated, so has
the "Organic Nitrogen Balance" summary.

There is a new "tillage" log variable in SoilWater, SoilNH4, SoilNO3
and the SoilChemical submodels, which mean that the "Total Water
Balance", "Soil NH4 Balance", "Soil NO3 Balance" and "Nitrogen
Balance" summaries show a balance, even if you do not log the entire
plowing layer.

* 3.39

** Fixed bug in convection transport model.

* 3.38

** No changes.

* 3.37

** 2% N lost as N2O during nitrification.

This default has been set down from 4%.  This time for real.

** Dead leaves are now logged in N Balance and summaries.

This fixes some minor N balance errors, especially if their were a
crop on the field at the end of the simulation.

* 3.36

** "NO3 Root Uptake" and "NH4 Root Uptake" new log parameterizations.

** DOM turnover is now affected by water.

* 3.35

** New information in the top summary.

** New log models "Chemical Content" and "Chemical Concentration".

** chemistry.dai reorganizaed, and new pesticides added.

** Bug fixes

The input listed in the OrnaicMatter section in the beginning of
daisy.log did not included carbon lost digesting the added matter.
Now it does.

** Visual C++ .Net 2003 port.

* 3.34

** New way to specify deposition in the default weather model

Instead of specifying deposition with the (NH4|NO3)(Wet|Dry)Dep
keywords in the .dwf file, you can use this set of keyword:

Deposition: kgN/year
  Total yearly deposition, given in kgN/year.
PAverage: mm
  Total yearly precipitation, given in mm.
DepDry: % 
  Dry fraction of total deposition (from air). The remainder will fall
  with the precipitation. 
DepDryNH4: %
  Fraction of NH4 in deposition from air.  The remainder will fall in
  the form of NO3. 
DepWetNH4: %
 Fraction of NH4 in deposition from rain and snow.  The remainder will
  fall in the form of NO3

You cannot specify both.  If you use the this specification, these
numbers will be used for calculating the equivalents of the classic
specification of deposition.  

Example:

Deposition: 16 kgN/year
PAverage: 493 mm
DepDry: 40 %
DepDryNH4: 60 %
DepWetNH4: 50 %

You do not have to specify DepDry, DepDryNH4, or DepWetNH4.  If you
don't the numbers above will be used (they are from the _Daisy
Standardisering_ project).

** Daisy will now write a summary of top soil organic matter initialization

The summary will be written in daisy.log.  If you set the top_summary
organic matter init parameter like this: 
  
  (defcolumn Andeby default
    "The B.And farm, Andeby, 2002."
    (OrganicMatter (init (top_summary "topsum.tab")))
    (Soil (horizons (-20 [cm] Ap) (-2.5 [m] C))
          (MaxRootingDepth 60.0 [cm]))
    (Groundwater deep))

Then the summary will also be written to the specified file for
further processing.
  
** Fixed some dimenesions.

** Updated answers to exercise 1 to 8.

* 3.33

** Bug fixes

** New "Soil N Residuals" and "Soil C Residuals" log parameterizations.

* 3.32

** New "sorg_height" harvesting parameter in the default crop model.

Set this to a negative number to ensure that harvesting implies as
tillage operation, and always kill the plant.

** Harvesting potatoes and beets now implies a tillage operation.

** New "Potato; SCRI" parameterization in potato.dai

This is based on hight quality data from SCRI, Scottland.

* 3.31

** The exercises have been updated.

The now reflect the new organic matter initialization model and
automatic aquitard horizon.

The answers have only been partly updated.

** Allow spaces at end of hyphen-only line for default weather model.

* 3.30

** Renamed "Majs" to "Silomajs" and adjusted parameters.

New dk-maize.dai file from Christian Thirup.

** New log model "Soil Water Potential (pF)".

Log water potential in all numeric layers as pF.  

* 3.29

** Logged Crop N uptake was larger than simulated Crop N uptake.

Basically, the log file would claim the crop kept taking up nitrogen,
even after the crop was ripe.  This only affected the log file, not
the simulation itself.

* 3.28

** Fixed minor bug in weather initialization.

* 3.27

** Adjusted parameterizations for winter crops (wheat, barley, and rape)

** Slightly finer automatic numeric discretions of soil.

* 3.26

** Made "weather default missing_years" more robust.

* 3.25

** Fixed bug in Freundlich adsorbtion model.

** Renamed Es_reduction_factor to vapor_flux_factor.

* 3.24

** EpInterchange moved from Surface to Vegetation.

** New Litter submodel of permanent vegetaion.

Here you can set 

*** vapor_flux_factor

This controls reduction of potential evaporation below litter.  The
default value is 1.0, meaning the litter does not block potential
evaporation at all.

*** interception_capacity

The water storage capacity of the litter layer.  By default, zero.

*** albedo

The albedo of the litter layer.  By default it will use the albedo of
the soil surface.

Example:

  (Vegetation Hardwood         
              (Litter (vapor_flux_factor 0.9 [])
                      (interception_capacity 2 [mm])
                      (Albedo 0.3)))

** "Bioclimate" and "Surface Water Balance" updated with litter.

This affects even runs without litter.  In particular, litter_ea will
include evaporation of rain "as it falls", which mean that pond_ea now
only included evaporation of ponded water.

* 3.22

** 2% N lost as N2O during nitrification.

This default has been set down from 4%.

** Fixed bug with partial specification of SOM_fractions.

If you specified "inactive humus" only, like in 

  (SOM_fractions -1 -1 0.35 [])

Daisy would ignore the specification is some circumstances.

* 3.21

** Added "Evapotranspiration" summary to "Surface Water Balance" log.

** Error messages can again be redirected under WinDOS.

* 3.20

** Bug fixes.

* 3.19

** Bug fixes.

* 3.18 

** Interception capacity for permanent vegetation now depends on LAI.

** The rs_min canopy parameter now affects the PM pet model.

The default value for rs_min is now 100.

** New FAO_PM pet model.

This model calculates Penman-Monteith on a reference surface, use the
old PM model if you have weather data measured at the field.

* 3.17

** New parameter yearlyLAI in permanent vegetation.

Use this parameter if you have LAI measurements for specific years.
The syntax is similar to "ForcedLAI" in the crops vegetation model.

(defcolumn Andeby default
  "The B.And farm, Andeby, 2003."
  (Vegetation Hardwood
              (YearlyLAI (1987 ((100 2.0) (200 4.0)))
                         (1988 ((100 0.0) (150 1.5) (200 5.0)))))
  (Soil (horizons (-20 [cm] Ap) (-2.5 [m] C))
        (MaxRootingDepth 60.0 [cm]))
  (Groundwater deep))

** [mm/h] and [mm/d] can now be used instead of [cm/h]

** The quarts_in_clay, quarts_in_silt and quarts_in_sand removed.

These where horizon parameters.  Let us know if you needed them, and
we will restore them.

** Bug fixes.

* 3.16

** Automatic choice of pet model

If you do not specify the pet model (potential evapotranspiration) in
Bioclimate, Daisy will choose one for you depending on the available
climate dats.

If reference evaporation is available in the climate data, Daisy will
use these (the "weather" pet model).

Else, if vapor pressure and wind are available, Daisy will use
Penman-Monteith (PM).

Else, if daily minimum and maximum temperature are specified, Samani
and Hargreaves (Hargreaves) will be used.
 
As a last resort, Makkink (makkink) will be used.

** New height parameter to crop_DM_over condition

Using this will only count DM above the specified height.  The default
value is 0.0, which corresponds to the old behaviour.  Example

Example:

    (wait (crop_dm_over "Grass" 4000 [kg DM/ha] (height 10.0 [cm])))
    (harvest "Grass" (stub 10.0 [cm])))

should result in a harvest of 4 Mg/ha.  Without the height parameter
you would get an earlier but smaller harvest, as the lowest 10 cm
grass would be counted in the total DM for the test, even though it
was not actually harvested.

** Updated tutorial.

It now describes the new organic matter initialization.

** New permanent vegetation parameterizations in vegetaion.dai.

Coniferous has been updated, Hardwood, Grass and Bush are new.  The
parameterizations are based on satelite meausrements for all of
Denmark.

Thanks to Eva Bøgh <Evb@geogr.ku.dk> for the parameterizations.

** Bug fixes.

Hourly values should work better now.

Organic matter in permanent vegetation should work less bad now.

* 3.15

** wait_days now wait one hour less.

It use to wait one hour more than the specified amount of time.  This
is no longer the case.

** New wait_hours command.

Waits the specified number of hours.

** Bioincorporation is now faster by default.

** New pesticide 'Ioxynil' in chemistry.dai.

** Irrigate commands now take 'hour' and 'days' arguments.

These determine how long time the irrigation will last.  By default,
irrgation will last one hour (which is also the minimum).

For example the command

  (irrigate_surface 1 [mm/h] (hours 24))

will last 24 hours give a total irrigation of 24 mm.

** New command 'irrigate_subsoil'.

This works like the other irrigate commands, except that the water is
incorporated directly in the soil.  It takes to additional parameters,
from and to, which determine where in the soil the water is
incorporated.  Example:

  (irrigate_subsoil 1 [mm/h] (hours 24) (from -5 [cm]) (to -20 [cm]))


** Daisy will now check dimensions for the IM submodel.

Example:

  (irrigate_subsoil 1 [mm/h] (hours 24) (from -5 [cm]) (to -20 [cm])
                    (solute (NH4 300 [mg N/l]) (NO3 50 [mg N/l])))

If you used the "unknown" dimension syntax, [?mg N/l], you must adjust
your setup by removing the question mark.

** The lr matrix water transport model should work better now.

It should give better results with the lysimeter, fixed and file
Groundwater models.

** Humus change and background mineralization in daisy.log file

The daisy.log file now contains two new lines:

  Expected humus change: -175.423 [kg C/ha/y], -4.50161e-07 [y^-1].
  Background mineralization: 15.9476 [kg N/ha/y].

They can work as a quick summary of the organic matter system
inititalization.  With an unchanged input level, we'd expect the
amount of humus in the soil to decrease over time, releasing 16 kg
N/ha/y.

* 3.13

** "N Blance" log parameterization changes.

New NH4-fertilizer-Surface, Volatilization-Surface and
NO3-fertilizer-Surface columns which ignore incorporated fertilizer,
and fertilizer in subsoil irrigation.

The "Nitrogen Blance" log summary has been updated to use these.  You
should now get a correct balance when using subsoil irrigation with
mineral fertilizer.

* 3.12

** You can now specify daily max and min temperature.

There are two new fields in the ".dwf" files: T_min and T_max.  These
sepcify the minimum and maximum daily temperature.  

If you specify T_min and T_max, Daisy will create an hourly air
temperature variation by assuming T_min is reached at sunrise, and
that T_max is reached at 15:00, and interpret linearly in between.  No
attempt is made to make the generated hourly values average to the
value specified in AirTemp.  Models that rely on daily average
temperature rather than hourly temperature will still get the number
specified by AirTemp.  Also, in the rare cases where T_min for a given
day is larger than T_max for the preceding day, some of the hourly
temperatures will be outside the [T_min;T_max] interval.

If the above limitations are not acceptable, supply hourly values for
AirTemp, and let Daisy find T_min and T_max from these.

** New "Hargreaves" potential evapotranspiration (pet) model.

This model estimates potential evapotranspiration based on the daily
temperature variation, i.e. the T_min and T_max described above.

** Fixed bug that prevented bioincorporation in some cases.

** "Grass to grain" will now survive cut at height 0 and after ripeness.

It used to be that only "Grass" would survive that.  The NEWS entry
for 2.43 got it wrong.

If you want to undo these changes, use

(defcrop "MyCrop" "Grass to grain"
  "This grass don't survive cut as well."
  (Prod (IntDSRelRtRes 0.80))
  (Harvest (DSmax 0.99)))

** New Danish crop kalibrations.

Majs, Aert and Graes are now available from dk-crops.dai, kalibrated
to average Danish yield.

* 3.11

** Updated the biomod clayom model.

It should now correspond to the article _CN-SIM - a model for the
turnover of soil organic matter._ by Bjørn M. Petersen et al,
submitted to Soil Biology & Biochemistry, 2003.

* 3.10

** New parameterizations and parameter files.

dk-crop.dai and dk-fertilizer.dai are new to the distribution, they
contain parameterizations developed from the Daisy standardization
project.  They are based on average numbers for Danish farms, rather
than specific field experiments.  All names are in Danish.

** New mineral fertilizers.

A number of new mineral fertilizer parameterizations have been added
to fertilizer.dai.

** New command line option.

From the command line

  daisy setup.dai -l am table

will generate a tab seperated table containing all the fertilizers
that are either build-in or loaded from the file setup.dai (just an
example, substitute with the files you want to summarize).

** Bug fixes.

Fixes crashes on some situations with low mineral nitrogen content.

* 3.09

** Log summaries now handle logging a soil interval.

This involves better tracking of bioincorporation (it might be
incorporated outside the soil interval), root uptake (from outside the
soil interval) and root death (outside the soil interval).

* 3.08

** The "N Balance" log model now produce a total nitrogen summary.

This includes all nitrogen in the system.

** The "N Balance" log model now log Crop-N and Surface-Org-N.

This is needed for the above balance.

* 3.07

** Handle high temperature soil.

Daisy used to give an error when the soil temperature got over 
40 dg C because the heat factor for various processes were not defined
above that level.  Now the heat factor max out at 37 dg C, and
gradually decrese down to 0 (no activity) at 60 dg C, as suggested by
J.A. van Veen and M.J.Frissel. 

* 3.06 

** Changed default value for K_aquitard from 1e-4 to 1e-3.

In the groundwater pipe model. 

* 3.05

** You can now specify a single SOM pool.

If you want to specify the amount of inactive humus in a specific
horizon, and let Daisy figure out the partitioning among the active
pools itself, you can specify -1 for the active pools in the
SOM_fractions horizon parameter.  For example

  (defhorizon foo default
    (SOM_fractions -1 -1 0.05)
    ;; More parameters.
  )

will specify that 5% of the humus in the 'foo' horizon is inactive
(i.e. is in SOM3), and let Daisy figure out how much of the rest is in
SOM1 and SOM2 by itself.

In general, the special value -1 in SOM_fractions means Daisy should
figure out the content itself.

* 3.04

** dk-horizon.dai updated with new and better numbers.

** Bug fix: K_to_pipes is now optional.

* 3.03

** 4% N lost as N2O during nitrification.

By default, 4% of the ammonium consumed during the nitrification
process does now escape as nitrous oxide.  The remaining well be
converted to nitrate.

The typical effects will be sligtly smaller simulated N harvest or N
leaching.

The "N Balance" log model has been opdated to log both NH4 consumed,
and N2O and NO3 produced.  The "Denitrication" log model also log the
N2O produced, as this is sometimes mistaken for denitrification.

You can control the N2O production with the new "N2O_fraction"
parameter in the nitrification models.

** New K_to_pipes parameter in the pipe groundwater model.

This parameter indicates how fast water flow to the drainpipes in
saturated soil.  By default the K_sat and anisotropy parameters of the
horizon is used.  

You can use it together with K_aquitard to calibrate deep percolation
and drain flow.

* 3.02

** You can now supress the line with initial values in the log files.

Set the flag "print_initial" to "false" when you specify what log
files to generate.

Example:

  (output ("N Balance" (print_initial false)))

** Faster denitrification.

Denitrication has been recalibrated, and is significantly faster now. 

** Better error message when weather file is missing.

* 3.01

** daisy.exe should be fast again.

* 3.00

** The "PrecipCorrect" in the .dai file renamed to "PrecipScale".

I.e. the new syntax is

  (weather default "weather.dwf"
           (PrecipScale 2.0 2.0 2.0 2.0 2.0 2.0 2.0 2.0 2.0 2.0 2.0 2.0))

The PrecipCorrect in the ".dwf" file is unchanged.

** Fixed (yet another) bug in OrganicMatter init.

There is a new 'efficiency' parameter which specifies the efficiency
the C input is integrated into the SMB pools.  This has the default
value of 0.5.  Before it was integrated with an implicit 100%
efficiency rate.

* 2.99

** dk-horizon.dai now has C/N values.

** You can now specify PrecipCorrect in the .dai file.

To double precipitation in "weather.dwf":

(weather default "weather.dwf"
         (PrecipCorrect 2.0 2.0 2.0 2.0 2.0 2.0 2.0 2.0 2.0 2.0 2.0 2.0))

If you specify PrecipCorrect in with the .dai file and the .dwf file,
both will be applied.

It is suggested that the numbers in the .dwf should represent
corrections based on systematic errors in the measurement process,
while corrections done as experiments and attempts to use the data at
different locales should be located in the .dai file.

* 2.98

** Bug fixes.

* 2.97

** New K_fast denitrification parameter.

This parameter limits denitrification based on activity in fast OM
pools.  By default, it is identical to K.  K now only limit the slow
pools. 

This may mean slightly faster denitrification in some situations.

* 2.96

** Smother groundwater pipe table curves.  

Groundwater may now be in the middle of a node, or above ground.

** Made UZrichard more robust.

* 2.95

** Simplified implementation of the pipe groundwater model.

The new implementation should be more robust.

* 2.94 

** Fix bug that rendered anisotropy irrelevant.

** Updated dk-horizon.dai with new data.

* 2.93

** Groundwater pipe now survives flooding.

The pipe groundwater model does not works for a totally saturated soil
profile.  To work around this, it will now push water to the surface
before the first node get saturated.  This approximates some of the
effects of a groundwater table that lies above the soil surface.  But
if you log the groundwater table, it will be shown somewhere below the
surface.

You will get messages about flooding beginning and ending when this
happens.  If this happens a lot in reality, you should suggest the
farmer to improve his drainage system, it is obviously inadequate.  If
it only happens in simulations, you should increase either anisotropy
or K_aquitard, depending on whether you want the excess water in the
pipes or not.

Anyway, the effect is that the simulations survives in many cases
where they didn't use to.

* 2.92

** New 'anisotropy' horizon parameter.

Set this to a number larger than 1.0 to make water flow faster towards
drain pipes.  

* 2.91

** Daisy now only print the date on the screen when something happens.

To restore the old behaviour, add

  (print_time daily)

to your setup file.

You can use any condition, so e.g. 

  (print_time hourly)

and

  (print_time (weekly 2))

will print time each our or every second week, respectively.

* 2.90

** Updated dk-horizon.dai with new conductivity parameters.

* 2.89

** Denitrification parameters changed meaning.

The 'alpha' and 'water_factor' parameters now only applies to the slow
pools.  On the other hand, the 'alpha_fast' and 'water_factor_fast'
parameters are now by default identical to the non-fast versions, so
everything should work as usual.

** New 'redox_height' Denitrification parameter.

All NO3 below this depth in the soil will be immediately denitrified
chemically.  You can log this with the new converted_redox log
variable, which is also in the Denitrification log parameterization.

By default, no chemical denitrification occurs.

* 2.88

** New 'M_vGp' hudraulic model by Børgesen et.al..

Compared to a straight Mualem + van Genuchten, it has two extra
parameters, 'h_m' which indicated the transition point for macropores,
and 'f' which is a shape parameter for the conductivity function.

** Clarified documentation for the DSnew harvesting parameter.

** Slight speedup in logging.

* 2.87

** Denitrification can now be affected by fast OM pools.

For a definition of "fast OM pools", see entry 2.71.

This is controled by two new parameters to the Denitrification
submodel, alpha_fast and water_factor_fast. You can set them like
this:

    (defcolumn Andeby default
      "The B.And farm, Andeby, 2002."
      (Denitrification (alpha_fast 0.3)
                   (water_factor_fast (0.5 0.0) (0.8 0.1)))
      (Soil (horizons (-20 [cm] Ap) (-2.5 [m] C))
      (MaxRootingDepth 60.0 [cm]))
      (Groundwater deep))
    
    (output "Denitrification")

These adds to the denitrification just like the old alpha and
water_factor parameters, but the total denitrification is still
limited by the K parameter.

By default, CO2 from fast pools have no special effect on
denitrification.  

You can get detailed information about potential and actual
denitrification with the new "Denitrification" log parameterization.

** CO2_fast is no longer printed in "Carbon Balance".

* 2.86

** More information from the "Carbon Balance" log parameterization

When choosing the "Carbon Balance" log parameterization, you will get
a summary like the following printer to the screen as well as in
daisy.log.   

Applied fertilizer =  300 [kg C/ha/y]
         Residuals = 1194 [kg C/ha/y]
             Roots =  325 [kg C/ha/y]
-------------------------------------
     Total C input = 1819 [kg C/ha/y]

In general, it is now possible to make Daisy print such summaries for
log files generated by the 'table' log model.  Look at the
documentation for the new 'summary' parameter, and at the "Carbon Balance"
parameterization in 'log.dai' for an example.

To get rid of the summary, specify an empty summary parameter for the
log parameterization, like this:

(output ("Carbon Balance" (summary)))

** The "N Balance" log model print out summaries for NH4 and NO3.

Similar to above.

** The "Root Zone Water Balance" log model now prints a total water balance.

Similar to above.

** The "Root Zone Water Balance" log model now includes surface storage.

Needed to create a total water balance.

* 2.85

** Minor cleanups to logging.

* 2.84

** New log table time_columns parameter.

If you specify this, you will automatically get year, month, mday and
hour columns.  By default, time_columns will be true if you do not
specify any time columns in the entries list.  All the log
parameterizations in log.dai has been updated to take advantage of
this, which gained approximately 10% speed.

* 2.83

** Minor cleanups and speedups to logging.

** The Surface lake parameter might work with lr.

* 2.82

** Minor cleanups and speedups to logging.

** The Surface lake parameter might work again.

* 2.81

** No user visible changes.

* 2.80

** Fixed missing "when" in log.dai.

* 2.79

** Log files now contain a line with initial content.

This line display the initial content of all non-flux variables.  It
only affect log parameterizations that has any non-flux variables.

** Changed default for OrganicMatter input root parameter to 800 [kg C/ha/y]

This reflect simulation results made by Henrik Svendsen.

** New duplicate warning.

If you specify the same attribute twice in the same list, you will get
a warning.  Same if you define a parameterization twice, or try to
overwrite a build-in model.

* 2.78

** The old log file syntax is restored.

You can now again begin log file parameterizations with

  (entries (year (path time))
           (month (path time))
           (mday (path time))
           (hour (path time))

but you will get a warning to shift to the new syntax mentioned in the
entry for 2.73.

** More control over the 'hypres' hydraulic model.

By default, the top horizon will be initialized by 'hypres' as a
topsoil, and the rest as subsoil.  You can now control this by
specifying the topsoil parameter.

(defhorizon Ap default
  (hydraulic hypres (topsoil true)))

(defhorizon C default
  (hydraulic hypres (topsoil false)))

** Changed default value for K_aquitard to 1e-4 [cm/h].

This groundwater pipe parameter used to default to 1e-5 [cm/h].

* 2.77

** A bit more speedups and cleanups to logging.

Having lots of log files is no longer expensive,

** The 'flux' parameter for all select models.

Instead of writing (when hourly) to ensure flux variables are getting
accumulated between log file timesteps, write (flux true).  The old
syntax will contibue to work for some time, but generates a warning.

** New dk-horizon.dai parameterization file.

This file contains median values from the DJF "Profil Database".

** You can now use [cm/d] instead of [cm/h].

* 2.76

** Logging is much faster.

Generation of log files is now seven times faster than before.  Lots
of stuff was changed in order to achieve this speedup, so it is likely
than some features have broken.  Please let me know in that case.

** Removed the table1 log model.

The one place it was used in log.dai (the PMSW log parameterization)
has been converted, fetch a new file.

* 2.75

** Minor fix in organic matter initialization.

* 2.74

** New automatic aquitard horizon.

Drained soil will now automatically get an aquitard horizon, unless
you specify the discretization (zplus).  The aquitard horizon is a
third of the size of the aquitard itself.

This affect logging, since the log models by default log down to the
end of the last horizon.  I therefore suggest you always specify the
'to' parameter when using the 'pipe' groundwater model.

The aquitard has the following parameters

(defhorizon aquitard default
  "Aquitard horizon used by drained soil."
  (hydraulic hypres (K_sat *K_aquitard*))
  (clay 50)
  (silt 20)
  (sand 29.99)
  (humus 0.01)
  (dry_bulk_density 2.0 [g/cm^3]))

where *K_aquitard* is the value of that parameter in the pipe
groundwater model.  

If you define your own horizon named "aquitard", that will be used
instead of the above.  The only requirement is that the hydraulic
model used must accept a 'K_sat' parameter.  Most do.

* 2.73

** Log file syntax changed.

To write year, month, day of month and hour to the log file, instead
of 

  (entries (year (path time))
           (month (path time))
           (mday (path time))
           (hour (path time))

the syntax is now

  (entries (number (path time year))
           (number (path time month))
           (number (path time mday))
           (number (path time hour))

The log parameterizations in log.dai have been updated.

** The "Organic Matter" log model now log SOM3 as well.

* 2.72

** New 'C_per_N' horizon parameter.

The intent of this parameter is to represent the measured C/N ratio
of the soil in the horizon.  Use it like this:

(defhorizon Ap default
  ;; Insert other parameters here.
  (C_per_N 16.0))

Note that the effect of setting this parameter is entirely unproven,
so use it with great caution.

The idea is that the varlues in the old 'SOM_C_per_N' parameter now
represents the "goal" C/N for the SOM pools, but not necessarily the
initial value.  The idea is that the farmland always end up with a C/N
ratio of 11, but that newly cultivated land may have a different
ratio.  We model this by saying that C/N entering the pool will always
have the ratio given in 'SOM_C_per_N', but C/N leaves the pool with
the current ratio, initially very close to the value specified by
'C_per_N'.

Note that the initial C/N ratio will not be identical to the number
spcified by the 'C_per_N' parameter, as 'C_per_N' represent the
average of all initial pools, including the AOM and the SMB pools.
The SOM pools will get what is left after the AOM and SMB pools have
been initialized.  This number will usually be very close to 'C_per_N'
though, as the SOM pools tend to much larger than the other pools.

Setting 'C_per_N' instead of 'SOM_C_per_N' for a soil with a high
measured C/N ratio will give significantly less mineralization, as
add organic matter will no longer for forced to have the same high C/N
ratio as the original content.

Currently all SOM pools will have the same initial C/N ratio when you
specify the 'C_per_N' parameter, but we might chang this so fast pools
ar closer to the goals specified in 'SOM_C_per_N' than slow pools.

* 2.71

** New CO2_fast log variable in OrganicMatter.

It is logged in the "Carbon Balance" log model.

It tells the CO2 produced by fast pools, where fast mean it has a
turnover rate larger than CO2_threshold, a new OrganicMatter
parameter.  The default value of 1e-4 means the SMB2 pool and all AOM
pools are considered fast with the default parameterization.

* 2.70

** New turnover_factor horizon parameter.

This parameter affects the turnover of all organic matter pools, in
effect making th whole system go slower (or faster if you set it
larger than 1).  Use it like this

(defhorizon C default
  ;; Insert other parameters here.
  (turnover_factor 0.1 []))

to make the organic matter system move 10 times as slow in the "C"
horizon. 

* 2.69

** Fixed bug that prevented B_C hydraulic model from being used.

* 2.68

** SOM1 now limited to be between 0.3 and 0.7.

If you let Daisy calculate the initial organic matter partitioning,
that is, do not specify the SOM_fractions horizon parameter, Daisy
will no longer come up with SOM1 fractions below 0.3 or above 0.7,
which is what our simulations have shown SOM1 stay within after change
in farming practice for typical Danish conditions.

These limits are controlled by three new OrganicMatter init
parameters, the default values are:

  (OrganicMatter (init (SOM_limit_lower 0.3 0.7 0)
                       (SOM_limit_upper 0.7 0.3 0)
                       (SOM_limit_where 0)))

Check the reference manual for a description of each.

** You can now specify K_sat or K_at_h for th hypres hydraulic model.

If you do so, that value will be used instead of the value predicted
by the HYPRES model.  All other hydraulic parameters will still be
calculated by HYPRES.

(defhorizon Ap default
  ;; Insert other parameters here.
  (hydraulic hypres (K_sat 3 [cm/h])))

** Allow automatic partitioning for subsoil with low humus content.

Previously, it would result in a negative SOM3 value.  Now we
repartition it as a top_soil, which mean SOM3 will be zero and SOM1
will not be in equilibrium.

* 2.67

** New K_at_h hydraulic parameter in most models.

This is an alternative way of specifying K_sat.

  (K_at_h 2 [pF] 5e-8 [m/s])

will mean the hydraulic conductivity have been measured to 5e-8 m/s at
pF 2.  From this, Daisy will find K_sat automatically.

** Changed EpFac for the simple and default crop models to 1.15.

Expect significant changes in simulation results.

You can change it back to 1.0 with

    (sow ("Spring Barley" (Canopy (EpFac 1.0))))

in order to reproduce older results.

** 'soil_h' state variable removed from phenology models.
** 'soil_temperature' and 'partial_soil_temperature' state variables moved.

From phenology models to root_system submodel.  This will not affect
most people.

* 2.66

** The `weekly' condition should work now.

It never worked before.

* 2.65

** You can now initialize organic matter based on background miniralization.

If you specify 'background_mineralization' in OrganicMatter init and
do not specify 'SOM_fractions', the system will attempt to choose SOM1
and SOM2 values so the total mineralization from the SOM (and SMB)
pools, but not the AOM pools, are that amount.

Example

  (OrganicMatter (init (input 1400 [kg C/ha/y])
                       (background_mineralization 30 [kg N/ha/y])))

The net mineralization is the background mineralization plus the
amount of added organic nitrogen.

* 2.64

** There is now an inert SOM3 pool by default.

You can use this to "park" part of the humus so it will not contribute
to the mineralization by specifying a third number to the
SOM_fractions horizon parameter.  E.g if you instead of 

  (SOM_fractions 0.6 0.4)

write 

  (SOM_fractions 0.3 0.2 0.5)

half the humus in the system will be inactive, cutting 50% off the net
mineralization.

If you do not specify the third number, it will be assumed to be zero.
In effect, the current behavior (with no SOM3 pool) will be
maintained.

** Default organic matter initialization of subsoil changed.

If you do not specify SOM_fractions for a subsoil horizon, Daisy will
automatically park sufficient humus in the inactive SOM3 pool to get a
zero net mineralization.

** The AOM pools are now considered part of the humus for initialization.

The effect of this is that the SMB and SOM pools will be
correspondingly smaller, and you will get a slightly smaller
mineralization.

* 2.63

** New PenClayFac parameter for root penetration.

Example:

    (sow ("Spring Barley"
          (Root (PenClayFac (0.0 0.8) (0.1 1.0) (0.3 1.1)))))

This mean root penetration will be 20% slower with zero clay content
and 10% faster with 30% clay in the soil.  The default root
penetration speed is valid for soil with 10% clay.

** Use average daily temperature for root penetration.

We used to to use the temperature at midnight.  This fix might affect
simulation results slightly.

* 2.62

** Bug fixes and cleanups.

** svat PMSW now obeys the screen height climate parameter.

This means it is useful for tall vegetation, like forests.

* 2.61

** No user visible changes.

* 2.60

** No user visible changes.

* 2.59

** Bug fixes and cleanups.

** Daisy will now refuse to run unless it can open an daisy.log file.

* 2.58

** No user visible changes.

* 2.57

** Bug fixes and cleanups.

* 2.56

** Changes to reference manual.

Simple default values (like numbers) are now printed in "sample"
sections.

** More parameter checks.

** daisy.exe is fast again

For mysterious reasons, the new compiler started generating very slow
code, even slower than pre-2.47.  Switching to another computer seem
to have solved that, so Daisy is up to the speed of 2.47.

* 2.55

** Removed the obsolete NH4_evaporation fertilizer parameter.

Use 'volatilization' instead.  

** Cleanups.

* 2.54

** No user visible changes.

* 2.53

** You can now specify the minimum concentration for nitrogen uptake.

Example:

(defcrop "MyGrass" "Grass"
  "Don't assume zero sink for nitrogen uptake."
  (CrpN (NO3_root_min 1e-8 [g N/cm^3])
        (NH4_root_min 1e-9 [g N/cm^3])))

* 2.52

** No changes.

* 2.51

** Fixed bug in organic matter initialization.

* 2.50

** Macropores now enabled by default for clay soil.

If the combined humus and clay fraction of the top horizon is above
5%, macropores will be enabled by default.  The macropores will go
down to 150 cm, or the first non-clay horizon, or the position of any
drain pipes, which ever is higher.

You can disable macropores like done in this example:

(defcolumn Andeby default
  "No macropores"
  (SoilWater (macro none))   ; Disable macropores.
  (Soil (horizons (-20 [cm] Ap) (-2.5 [m] C))
        (MaxRootingDepth 60.0 [cm]))
  (Groundwater deep))

The macropores will be initiated when the presure go above -3 cm.

** Automatic generated discretion now limited by dispersivity.

No interval size will be more than twice the dispersivity.

** Checkpoints work again.

** More robust error handling.

* 2.49

** Fixed bug introduced by 2.48 that prevented use of organic fertilizer.

* 2.48

** SMB initialization improved.

The SMB pools are still initialized assuming equilibrium with the rest
of the system, but the method is now exact rather than an
approximation.  This will cause small changes in mineralization, less
if you are using a warming up period.

** Alternative SOM initialization method.

If you do not specify the 'SOM_fractions' horizon parameter, the SOM
pools will instead be initialized from the total yearly C input, with
an assumption that SOM2 will be in equilibrium with the rest of the
system.   You specify the C input with a new 'init' parameter to
organic matter, like this:

  (OrganicMatter (init (input 1400 [kg C/ha/y])))

If you specify this parameter, it will also be used for initializing
the SMB pools.  If you do not specify the 'init' 'input' parameter, a
default value will be calculated from the current input rate of the
initial AOM content specified by the 'am' parameter to
'OrganicMatter'.  The default value for the 'am' parameter is

  (OrganicMatter (am (root (weight 1.2 [Mg DM/ha]))))

corresponding to 1.2 tons of dead roots.  This value not likely to
produce any useful numbers whatsoever.

You can estimate the yearly input by creating a likely crop rotation
for the rpevious decades, and run it for 10 years with

  (output ("Carbon Balance" (when yearly)))

add the "Fertilizer_C", "Residuals_C_top" and "Residuals_C_soil"
columns, and take the average yearly value.

Having done that, you will likely want to initialize the 'root'
parameter to 'init' as well.  The 'root' should be the average
"Residuals_C_soil".  A full specification for the 'init' parameter
would be:

  (OrganicMatter (init (input 1400 [kg C/ha/y]) ;Default: use am input.
                       (end -20 [cm])           ;Default: end of first horizon.
                       (fractions 0 1 0 [])
                       (root 480 [kg C/ha/y])
                       (dist 7 [cm])
                       (T 7.8 [dg C])           ;Default: from weather file.
                       (h -100.0 [cm])))

The equilibrium depends on soil clay content, temperature and water
content.  The clay content will always be taken from the horizons.
The temperature will be taken from the yearly average specified in the
weather file, unless overwritten by the 'T' parameter above.  By
default, the water content is assumed to be at field capacity.

The input is assumed to come from two sources.

1) Old roots whose density will decrease exponentially with the depth
   as specified by 'dist', for every 'dist' cm you go deeper, the root
   density will be decrease to half, until it reach the
   'MaxRootingDepth' soil parameter.

2) (Non-root) crop residuals and organic fertilizer, which will be
   distributed uniformly into the first horizon, or if specified the
   value of the 'end' parameter.

Note that 'input' is the total input, including roots.  The default
values for the 'root' parameter corresponds to 1.2 Mg DM/ha/y.

The 'fractions' parameter determine where the AOM pools will end up,
the three numbers represent SMB1, SMB2 and SOM2 in that sequence.  By
default, everything will be eaten by SMB2 as shown in the example.

Most users will only want to set the 'input' and perhaps 'root'
parameters.

You can see the calculated initial values for the SOM and SMB pools in
the 'daisy.log' file, and follow their changes with the
"Organic Matter" log model.

If you specify a very high input for a soil with a very low humus
content, Daisy will be unable to use this method for initialization,
and stop with an error.  The amount of SOM2 where the system is in
balance will be higher than the total amount of C in the soil.  In the
real world the system will not be in short-term balance either, you
will see that as a quick build-up of humus in the soil.

In reality, most systems will not be in short-term balance, to reach
even a short-term balance you need to run with similar input levels
for decades which is unusual.  Therefore, you should set SOM_fractions
explicitly if you have any data that will allow you to estimate the
mineralization.

If you have measured nitrogen content at harvest, you can fit the
SOM_fractions parameter until the simulated nitrogen content matches
the measured nitrogen content.  Increase the SOM2 fraction if the
simulated nitrogen content is too low to speed up mineralization, and
increase the SOM1 fraction if the simulated nitrogen content is too
high to slow down mineralization.

* 2.47

** More parentheses required for fetilize action.

Before, you could write 

  (fertilize mineral (weight 80 [kg N/ha]))

Now you must write

  (fertilize (mineral (weight 80 [kg N/ha])))

This fixes an old bug, and makes it consistent with the syntax for
incorporation where the parentheses have always been necessary:

  (fertilize (mineral (weight 80 [kg N/ha]))
             (to -20 [cm]))

I apologize for the inconvenience.

** Made checkpoints work with hypres and Cosby_et_al hydraulic models.

** NH4 lost to volatilization  is now always counted in "fertilized_NH4".

Previously, it was only counted during surface application, not during
incorporation.

** Better handle end of weather data.

** Daisy can now be compiled with Intel C++ 7.0.

In my tests, it produces slower and larger code than GCC 3.2.

* 2.46

** Fix bug with van Genuchten hydraulic models (including hypres).

If you had multiple horizons using the hydraulic model, all used the
same value for M, affecting root uptake.

** Allow Daisy to run even with very steep hydraulic conductivity curves.

** Switched to from Borland C++ to GCC for building daisy.exe

This seems to cut nearly half of the simulation time, and the size of
the executable to less than half the old size.

Technically, the new version is created by cross-compiling from Linux
to MinGW32 with GCC 3.2, while the old version was compiled natively
with Borland C++ 5.0.1.

* 2.45

** New missing_years attribute for the default weather model.

This allows you to reuse the weather data for additional years.  For
example, if you have weather data for 1991 to 2000 in a file named
"weather.dlf", but want to run a simulation from 1987 to 2001, you can
specify the weather model like this.

(weather default "weather.dwf"
         (missing_years ((1987 1990) (1993 1996))
                        ((2001 2001) (1991 1991))))

This should be read as an order to Daisy to, in the period from 1987
to 1990, use the weather data from 1993 to 1996, and for 2001 use the
weather data from 1991.

This is particularly useful for running Daisy with a "warming up"
period where we don't have actual measurements.

In the example above we map two periods (1987-1990 and 2001), but you
can have as many periods as you like, so you can e.g. run a 100 year
simulation with just 10 years worth of climate data.

Note that Daisy will always use the actual data when available, so if
you in the example above have some data from 2001 but not all, Daisy
will not switch to 1991 data before it have used up the real data.

* 2.44

** New "ForcedLAI" vegetation attribute.

You can now force the CAI to have a specific value for part of the
year.  This is useful if you have measured the total CAI and want
Daisy to use those values for photosyntheses and transpiration (and
interception capacity).  Here is an example:

(defcolumn Andeby default
  "The B.And farm, Andeby, 2003."
  (Vegetation crops
              (ForcedLAI (1987 ((100 2.0) (200 4.0)))
                         (1988 ((100 0.0) (150 1.5) (200 5.0)))))
  (Soil (horizons (-20 [cm] Ap) (-2.5 [m] C))
        (MaxRootingDepth 60.0 [cm]))
  (Groundwater deep))

The code above should be read: In 1987 we start using forced CAI day
100 where it is 2.0 and end forced CAI in day 200 where it is 5.0.  We
use liniear interpolation between these two points, so in day 150 it
will be 4.0.  Before day 100 and after day 200 we use the simulated
CAI values.  In 1988 we start day 100 at CAI 0.0, increase linearly to
1.5 day 150, and from there again linearly to 5.0 day 200, where we
switch back to using simulated CAI.  All other years we use simulated
LAI.

Only paramterizations of the "default" crop model will be affected by
the "ForcedLAI" attribute.

If there are mulitple crops, the forced CAI will be distributed among
them corresponding to the relative size of their simulated CAI.  When
forced CAI is active, this is the only use of the simulated CAI.  

But the simulated CAI will be calculated anyway and can be logged.
Most log files will only log the CAI actually used which, when Forced
LAI is in effect, will not be the sum of Leaf, Stem and SOrg AI.  The
new "Forced LAI" log model will log both the used, forced and
simulated CAI.  By default, it will log the sum of all crops, set
$crop to log a particular crop alone as below:

(output "Forced LAI"
        ("Forced LAI"
         (set "$crop" "Spring Barley")
         (where "sbarley.dlf"))
        ("Forced LAI"
         (set "$crop" "Grass to grain")
         (where "grass.dlf")))

** Bug fix for permanent vegetation.

The litter was not put back to the system.

** You can now overwrite individual parameters in lists of submodels

For example, if you want to create a version pig_slurry (form
fertilizer.dai) with faster turnover, you can do it

  (defam my_slurry pig_slurry
    "Pig slurry with faster turnover rates."
    (om ((turnover_rate 5.0e-4))
        ((turnover_rate 3.0e-3)) 
        ()))

This specifies a slurry with three OM pools.  The pools have the same
parameters as pig_slurry, except that the first two pools have a
faster turnover rate.  Note that pig_slurry itself have three OM
pools, a slow pool, a fast pool, and a pool that go directly to SOM2.

If my_slurry had more OM pools than pig_slurry, you would have had to
specify all parameters for the additional OM pools.

The last empty set of parantheses in the definition of my_slurry is
needed, otherwise you would not have got a copy of the third pool from
pig_slurry. 

** Fixed assimilate partioning for several crop paramterizations.

Specifically "Winter Wheat" and varieties, "Pea" and "Ikuwala Maize".

This might cause some changes in yields, possibly apart from
"Ikuwala Maize" these changes should be insignificant.

** New checks and better descriptions for many parameters.

As these checks found errors in several crops, you need to update your
lib files.

* 2.43

** "Grass to grain" grass model now regrow after a full or late cut

I.e., this grass needs a tillage operation in order to kill it.  

WARNING: This also affects regrowth after a "normal" cut, ans the
paramterization is no longer backed by experimental observations.
I.e. it produce bogus results!  Use only when bogus results are better
than no results.  Otherwise, use the "Grass" parameterization.

** New TSum phenology model.

Simply specify a temeprature sum for each fase:: emergence, vegetative
and reproductive.

 (Devel TSum 
        (EmrTSum 100.0 [dg C d])
        (VegTSum 760.0 [dg C d])
        (RepTSum 740.0 [dg C d]))

* 2.42

** Fixed bug in hypres implementation.

K_sat was 24 times to large.

** irrigation_subsoil and irrigation_total

New log variables in the default bioclimate model.  The total
irrigation is now included in the "Root Zone Water Balance" log model.

* 2.41

** New "original" pseudo-model.

This keyword allows you to reuse the original parameters of an component
attribute, when creating a specialization. 

For example, if we have this definition of a foo parameterization:

  (deffoo default "old foo"
    (bar default (a 1) (b 2)))

and want a new parameterization that is identical, except "b" should
be 42, we had to write

  (deffoo "new foo" "old foo"
    (bar default (a 1) (b 42)))

i.e. specify the type and all attributes for "bar" once again.

Now we can change "b" alone

  (deffoo "new foo" "old foo"
    (bar original (b 42)))

Se next change for a more specific example.

** Change to default crop model.

The value of the "Devel" attribute in the "default" crop model should
now be a parameterization of the new "phenology" component.

In practice, this means "(Devel " should be replaced with
"(Devel default" in the crop parameterization files.  This has already
been done with the files in the daisy-lib.zip file.

For any derived parameterization where you only changed a few
parameters of the "Devel" submodel compared to the original, you can
use the new "original" keyword, like this

  (Devel original (EmrTSum  300))

to overwrite the EmrTSum parameter only.

** Total precipitation now in "Weather" log file.

This is just snow + rain.

** New "l" paramter in the "B_vG" and "M_vG" hydraulic models.

You can now specify the "l" parameter to van Genuchten.

** Bug fixes.

* 2.40

** You can now just specify "sand" instead of "fine_sand" and "coarse_sand"

In the horizon.

** New default hydraulic model: hypres.

This model replace "Cosby_et_at" as the default hydraulic model.  

"hypres" is a pedotransfer function for estimating parameters for the
van Genuchten/Mualem equation.  We believe it works better than
Cosby_at_al for European soil.

"hypres" requires you to specify the dry_bulk_density horizon
parameter, if you for some reason does not have that value, we suggest
you go back to the "Cosby_et_al" pedotransfer function, by adding 

    (hydraulic Cosby_at_al)

to your horizons.  Otherwise, add

    (dry_bulk_density 1.5 [g/cm^3])

to the horizon, replacing 1.5 with the measured value.

As usual, all the pedotransfer functions are unreliable, you should
use measured values for the hydraulic properties when at all possible.

** Don't require leaf to survive harvest, as stem may provide photosyntheses.

** Better messages at harvest.

* 2.39

** Changed default dispersivity to 5 from 6.

** New calibration of organic matter model.

Daisy now uses the numbers from Sander Bruun et al, "Calibration and
Validation of the Soil Organic Matter Dynamics of the Daisy Model with
Data from the Askrov long-term experiments", Soil Biology &
Biochemistry, 2003.

###########################################
# THIS MEANS OLD SETUP FILES ARE INVALID! #
###########################################

You need to readjust the 'SOM_fractions' horizon parameter in your
setup files.  For a system in perfect equilibrium (input of organic
matter equal tuinover), the value will be

  (SOM_fractions 0.49 0.51)

For farms with a long term humus build-up, more of the humus will be
in the second (fast) pool.  For farms where the amount of humus is
decreasing, slow first (slow) fraction will be larger.  New farmland
and farmland using mineral fertilizer tend to have decreasing humus,
and thus a larger fraction of humus in the slow pool.  Note that we
are talking changes over decades and centuries, if you switched to
organic fertilizer five years ago, that is unlikely to have a
measurable effect on the SOM partitioning.

For horizons below the root zone, if there is any significant amount
of humus may be leftover from old tree roots, and slowly decreasing.
In that case, the vast majority (maybe 95\%) is in the first, slow
pool.  

All added matter must be adjusted to the new parameters, the
fertlizer.dai file and the crop parameterizations in the daisy-lib.zip
file have been updated accordingly.

* 2.38

** You can now use "all" as crop argument to the crop_DM_over condition.

This will include the combined  weight of all crops on the field in
the test. 

* 2.37

** (OrganicMatter (clay_factor ...)) renamed.

It is now a component.  To get the old behaviour, use

(OrganicMatter (ClayOM old (factor ...)))

** New biomod clayom model.

This is still very experimental.

* 2.36

** Daisy can now handle dissolved organic matter.

You can have any number of DOM pools, they are specified as a "dom"
attribute of organic matter.  For each DOM pool you must add an extra
fraction to the end of the SOM and SMB fractions.  You _can_ add extra
fractions to AOM pools, in order to direct fertlizer directly to the
DOM pools.

The DOM pools are eaten by the SMB, and tranported with the water
unless adsorbed to the soil.

By default, Daisy has no DOM pools.

Here is an example of OrganicMatter with one pool:

          (OrganicMatter (smb ((C_per_N 6.7)
                               (efficiency 0.6 0.6)
                               (fractions 0 0.6 0 0.4 0.0)
                               (maintenance 7.5e-05)
                               (turnover_rate 7.708e-06))
                              ((C_per_N 6.7)
                               (efficiency 0.6 0.6)
                               (fractions 0 0.4 0 0.6 0.0)
                               (maintenance 0.000416667)
                               (turnover_rate 0.000416667)))
                         (som ((efficiency 0.4 0.4)
                               (fractions 1 0 0 0 0)
                               (turnover_rate 1.125e-07))
                              ((efficiency 0.5 0.5)
                               (fractions 0.9 0 0.1 0 0)
                               (turnover_rate 5.83333e-06)))
                         (dom ((C (M 0.0 [g/cm^3]))
                               (N (M 0.0 [g/cm^3]))
                               (adsorption none)
                               (diffusion_coefficient 4.6e-6 [cm^2/s])
                               (efficiency 0.0 0.4)
                               (fractions 1 0)
                               (turnover_rate 0 [h^-1]))))

And here is an example of an AOM pool that goes directly to DOM:

(defam C_N_high organic
  (om ((turnover_rate 1.0)
       (efficiency 1.0)
       (fractions 0 0 0 1)))
  (dry_matter_fraction 1.0)
  (total_N_fraction 0.1)
  (total_C_fraction 0.4))

*** New "DOM Pools", "DOM-C" and "DOM-N" log models.

The first give a summary of two DOM pools in the specified interval.
The two later give the carbon and nitrogen content in the DOM pools of
all layers.  You can get the content of a specific DOM pool with the
"$pool" option, otherwise you will get the content of all of them.

** You can now specify global radiation in [MJ/m^2/d] in the weather file.

** You can now specify speed in [m/s] instead of [cm/h].

** Accept fertilizer with only a single OM pool.

** Better error message when using undefined parameterizations.

* 2.35

** fertilized NH4 in log files now include NH4 lost to volatilization.

* 2.33

** More internal changes in organic matter... beware.

* 2.32

** Internal changes and bug fixes to organic matter.

Please report any problems.

** Better timestamp in daisy.log.

** Code cleanup.

* 2.31

** More information i daisy.log.

** Complex default values are now also written in the reference manual.

** Bug fix in the transport convection model.

** C API change:  daisy_printer_alist now takes four arguments.

* 2.30

** Assertion failures are now logged in daisy.log.

* 2.29

** Made 'height_start' and 'height_end' optional in default macropore model.

It will now use the last and first point of the 'distribution'
parameter as defaults.  Also, some consistency checks have been added.

* 2.28

** You can now specify heat_factor and water_factor for individual OM pools.

This allows you to specify that heat and water affect some types of
organic matter more than it affects other types.

Here is an example where heat factor has been specified for AOM1 and
water factor has been specified for AOM2

(defam "Andeby Pig Slurry" pig_slurry
  (om ((initial_fraction 0.72)
       (heat_factor (0.0 0.0) (20.0 1.0))
       (C_per_N 100)
       (turnover_rate 2.0e-4)
       (efficiency 0.60 0.60)
       (fractions 0.5 0.5 0.0))
      ((initial_fraction 0.18)
       (water_factor (-10000.0 0.0) (-100.0 1.0) (0.00 0.0))
       (turnover_rate 2.0e-3)
       (efficiency 0.60 0.60)
       (fractions 0.0 1.0 0.0))
      ((C_per_N 11)
       (turnover_rate 1.0)
       (efficiency 1.0)
       (fractions 0.0 0.0 1.0))))

You can also specify heat and water factor for the SOM and SMB pools.  

(defcolumn Andeby default
  (OrganicMatter 
   (smb ((C_per_N 6)           ; SMB1
         (heat_factor (0.0 0.0) (20.0 1.0))
         (water_factor (-10000.0 0.0) (-100.0 1.0) (0.00 0.0))
         (efficiency 0.6 0.6)
         (turnover_rate 7.708e-6)
         (maintenance 7.500e-5)
         (fractions 0 0.6 0 0.4 ))
        ((C_per_N 10)          ; SMB2
         (heat_factor (0.0 0.0) (20.0 1.0))
         (water_factor (-10000.0 0.0) (-100.0 1.0) (0.00 0.0))
         (efficiency 0.6 0.6)
         (maintenance 4.16666666667e-4)
         (turnover_rate  4.16666666667e-4)
         (fractions 0 0.6 0 0.4))))

And similarily for SOM.

** Dimension of canopy_washoff_coefficient chemical parameter has changed.

It is now a fraction, 0.0 means it stick to the canopy and is never
washed off, 1.0 means it is fully dissolved in the intercepted water.

You must update chemistry.dai and all pesticide parameterizations.

** New log variable `water_stress_days' in "Crop Production".

This gives an estimate of the number of days worth of production lost
due to water stress.  It is created by multiplying the water stress
for each hour with the fraction of the daily radiation received that
hour, and accumulating the result.

* 2.27

** Residuals are now logged in "N Balance" and "Carbon Balance".

** Allow macropores to go below the lowest node.

* 2.26

** Specifying Soil zplus is now optional.

If you do not specify it, Daisy will attempt to create a reasonable
default discretitation.

** New log parameterizations "Organic Matter" and "AOM Pools".

These log the content of the individual organic matter pools.
"Organic Matter" logs SOM, SMB and all the AOM's.  "AOM Pools" logs
all the AOM's, or a specific AOM if "$am" is set.

Example: 

  (output "Organic Matter"
          ("AOM Pools" (set "$am" "Grass/root")))

** New spec attribute in select entries of the log table model.

A "spec" is a suplement to "path" for specifying which daisy attribute
should be logged.  If you specify "spec", Daisy will be able to either
find the dimension itself, or if you specify the dimension, attempt to
convert from the dimension used internally.

Example:

  (interval (path column "$col" SoilWater Theta)
            (spec column default SoilWater Theta)
            (tag "Water")
            (dimension "mm"))

The advantage of doing it this way is that Daisy will be able to
detect any errors.

The feature is still experimental, and will only work in few cases.
You will get an error for the cases where Daisy cannot convertion the
dimensions.

* 2.25

** Changed the equilibrium tranformation model.

The 'A' and 'B' parameters that specify the names of the two chemicals
are no longer ordered.

** You can now specify 'turnover_halftime' in organic matter pools.

This is an alternative to 'turnover_rate'.

** Changes to the default chemical model.

*** decompose log increment for pesticided moved from solute to chemical.

So now write 

 (defchemical Foo default
    (lag_increment (0.0 1e-5) (1.0 1e-4))
    ...)

instead of 

  (defchemical Foo default
    (solute (lag_increment (0.0 1e-5) (1.0 1e-4))
            ...) ...)

*** canopy_dissipation_rate_coefficient made obsolete.

Instead, you the new name 'canopy_dissipation_rate' or the new
parameter 'canopy_dissipation_halftime'.

*** New decompose_halftime parameter

It can be used instead of 'decompose_rate'.

** New [d] to [h] convertion.

Particular useful in konjunktion with the new halftime parameters.

* 2.22

** Soil water conductivity is now temperature dependent.

The specified water conductivity (K_sat) is assumed to apply at 
20 dg C, and scaled according to the water viscosity at the current
soil temperature.

* 2.21

** Added cross references to reference manual.

In particular, add information about where each component, model,
parameterization and submodel is used.

* 2.20

** Change the way to specify chemicals for logging.

You can now use a (set "$chem" Atrazine) parameter for "Surface
Chemicals", "Soil Chemicals" or "Chemicals" to log Atrazine alone, or
leave it out to log the sum of all chemicals in the soil.  As part of
this cleanup, the "Chemical" and "Soil Chemical" parameterizations has
been moved to 'log-old.dai'.

Also, Daisy will now automatically delay the Chemicals log until the
chemical enters the soil.  This means the number of columns will be
correct from start.  

** Change the way to specify organic matter for logging.

Update the "Total Soil Content", "N Balance", AOM and Bioincorporation
log parameterizations.  For AOM, use like (set "$am" "grass/root") to
log a specific AOM.  Compared to before, the "?" is missing.

* 2.18

** Daisy now creates a daisy.log file.

It contains most of the information printed during the run.

* 2.17 

** Fixed many (small) memory leaks.

Also added workaround for a memory bug in Groundwater pipe, and fixed
some uninitialized variables in svat PMSW.

** More cleanup in log.dai.

As usual, look in log-old.dai for missing log parameterizations, and
remember to use an explicit "when" if you don't want hourly values.

** New warnings and errors on invalid input.

* 2.16

** New "Prod RtDR_T_factor" parameter to the default crop model.

Temperature dependent factor for root death rate.

Use it like this

(defcrop MyGrass Grass 
  (Prod (RtDR_T_factor (0.0 0.0) (10.0 1.0))))

The will stop root death when soil soil freeze, and let it gradually
grow to full power when the soil temeprature reached 10 dg C.

** New PrecipCorrect keyword for the default weather model.

Add a 

  PrecipCorrect: 1.12 1.10 1.00 1.20 1.40 1.60 1.12 1.10 1.00 1.20 1.40 1.60

to the keyword section of your .dwf file to get mothly correction for
the precipitation.  The Precip-Correct should be followed by 12
numbers, each a factor to multiply to the precipitation for that
month.  The first number corresponds to January, the last to December.

** Made "Soil Temperature" and "Surface Water Balance"log models hourly.

** Moved "Daily Water Flux" and "Hourly Water Flux" to 'log-old.dai'.

* 2.15

** Amount of solute in drainage was entirely wrong.  Fixed.

** Moved all V1 crop parameterizations to 'crop-old.dai'.

** New soil_water_content_above condition.

See the section about irrigation in the tutorial for an example.

** Updated the pig_slurry and cattle_slurry parameterizations. 

** New pig_manure, cattle_manure, and horse_manure parameterizations.

** Updated Rye, Sugar Beet, and Spring Rape parameterizations.

** Renamed Maize to Pioneer Maize, and added new Maize parameterization.

The new maize is a sort common in Denmark.

** New Pendimethalin pesticide.

** New Potato crop parameterization.

** The reference manual is improved.

It now list standard parameterizations, and index all component,
model, submodel, attribute and parameterization names.  The
introduction is also updated.

* 2.14

** Fixed bug introduced in previous version that disabled macropores.

** Made check for mass balance more robust.

* 2.12

** Fixed error with extra CH2O in the default crop model.

It should not have measurable effect on production, but could produce
warnings during simulation.

** Made the default mactrans model more numerically robust.

This also removed some rare warnings during simulation, but has no
effect on the results.

* 2.11

** Fixed mass balance error in mineralization.

When removing tiny organic matter pools, their remaining C was
incorporated into the SOM, but not their N.  This could add up to 1 kg
N inbalance over 10 years under some circumstances.

** Imobilization is now restricted by available nitrogen.

It used to run on full speed, even when there was no nitrogen to
immobilize.

** Fixed bug in the Cosby hydraulic model.

The estimated conductivity was far too low to be useful.

** The cd transport model can switch to smaller timesteps.

This will happen in the cases where it gave negative answers, and
eliminates the "BUG: M[0] = -1.4115e-7 after transport" messages.  The
effect of the bug would be a slightly too low estimate on leaching for
nitrogen starved systems.

* 2.09

** There is now an index of logable variables in the reference manual.

** The 'default' bioclimate model is now default.

I.e., you don't have to specify it explicitly.

** If you don't specify DSnew in the default crop model, it will use 
the stub length to calculate the DS after a cut. 

* 2.08

** Fix bug preventing macropores and drainage to be used together.

** Fix checkpoints.

** You can now specify [d^-1] instead of [h^-1].

** You can now specify fractions with [%].

* 2.07 

** New "Cosby_et_al" hydraulic model.

This is a modified Bruce Campell model with parameters derived by
pedotransfer functions as specified by Cosby et al.

This model is now default, so you don't have to specify hydraulic
parameters for a horizon (you should, though, the pedotranspher
functions aren't that reliable.)

* 2.06 

** You can now specify [m] or [pF] instead of [cm].

** Syntax for specifying dimensions in arrays changes.

Dimensions now applies only to numbers before them.  For example:

  (foo 10.0 20.0 [cm] 1.0 1.5 [m] 100.0)

Here the two first entries (10.0 and 20.0) are in cm,  the two next
(1.0 and 1.5) are m, and the last is whatever the default dimension
for foo is.

* 2.05 

** New 'linear' Pedotransfer model.

This allows you to to specify parameters as a linear function of
horizon attributes.  Here is an example:

(defpedotransfer "van der Zee & van Riemsdijk" linear
  (description "0.5 (Al_0 + Fe_O)")
  (factors ("Al_O" 0.5)
           ("Fe_O" 0.5))
  (offset 0.0))
           
(deftransform "Phosphor Fast<->Slow" equilibrium
  (A "Phosphor Fast") (B "Phosphor Slow")
  (equilibrium Langmuir
               (K 1e-5 [g/cm^3])
               (pedo_my_max "van der Zee & van Riemsdijk")
               )
  (k_AB 1e-4))

Currently only the 'K' and 'my_max' attributes to the Langmuir
equilibrium model, and the 'k_AB' and 'k_BA' attributes to the
equilibrium transform model can be specified with pedotransfer
functions. 

The "Al_O" and "Fe_O" attributes must be specified in the horizon,
like this:

(defhorizon top default
  (attributes ("Fe_O" 1e-5)
              ("Al_O" 5e-6)
  ...)

* 2.04

** Transformations.

You can now create arbitrary transformations between soil chemicals. 

Example:

(defcolumn test default
  (Transformations (equilibrium (A "Phosphor Solute") (B "Phosphor Solid")
                                (equilibrium Langmuir
                                             (K 1e-5 [g/cm^3])
                                             (my_max 1e-2 [g/cm^3]))
                                (k_AB 1e-2 [h^-1])))
  ...)

Right now, "equilibrium" is the only implemented transform model.  It
transforms towards an equilibrium, using k_AB (and k_BA) as speed.
k_BA are by default the same as k_AB.

Langmuir is the only implemented equilibrium model.  

All the numeric parameters (K, my_max, k_AB, k_BA) are arrays, so they
can differ for each soil interval.  There are "pedo_" versions
(pedo_K, etc.) of them all, making it possible to specify them using
pedotransfer functions.  No pedotransfer functions are currently
implemented. 

** New "full" adsorbtion model

Use it with "none" transport to get non-solute soil chemicals.

** New "solid" standard chemical parameterization.

It uses the full adsorbtion model together with the "none" transport. 

** New "Soil Chemical" log parameterization.

These allows you to log a single soil chemical.

** New "Transform" log parameterizations.

This allows you to log transformations.

** Bug fixes.

* 2.03

** You can now specify dimensions for numbers in input files.

The dimensions are in brackets, like this:

(defhorizon top default
  (hydraulic M_vG 
             (K_sat 19.3 [cm/h])
             (Theta_res 0.05 [cm^3 H2O/cm^3])
             (Theta_sat 0.424 [cm^3 H2O/cm^3])
             (alpha 0.069 [cm^-1])
             (n 1.527 []))
  (SOM_fractions 0.66 [] 0.34 [])
  (SOM_C_per_N 11 [g C/g N] 11))

You must write the dimension exactly as is specified in the manual,
with the convention that ^ is used for superscript.  Also, fractions
and dimensionless values are both denoted as [].  Some values in Daisy
have unknown or unbspecified dimensions in the manual.  For these,
Daisy will accept any dimension in the input files, as long as it
start with a question mark.

There are no conversions yet, only the dimension used in the manual
may be specified.  In a furture release, Daisy will be able to make
some simple conversions, like [m] to [cm], allowing you to use what is
most convenient in the input file.

For arrays, if you specify the dimension for an entry (like
SOM_C_per_N in the example), that dimension is assumed to apply to the
rest as well, until something 

* 2.02

** You can now check on temperature sums.

To sow spring barley after the temperature sum reaches 110 dg C d,
write:

    (wait (TSum_above 110))
    (sow SpringBarley)

TSum is the sum of the daily average air temperature, ignoring frosty
days.  By default, it is reset March 1 every year.  

* 2.01

** New root density models

*** DS_Depth

Example:

    (sow ("Spring Barley"
          (Root (rootdens "DS_Depth" 
                          (entries (0.0 (0.0 1.0) (10.0 0.0))
                                   (1.0 (0.0 1.0) (50.0 0.0))
                                   (2.0 (0.0 1.0) (40.0 1.0) (50.0 0.0)))))))

This specifies the root distribution at DS 0.0 (emergence), 1.0
(flowering) and 2.0 (ripe) respectively.  At all times, the highest
density is at the top.  At emergence, the density diminish linearly
until 10 cms depth, below that there is no root.  At flowering, it
still diminish linearly, now down to 50 cm.  At ripe, it has maximum
desnity down to 40 cm, at which point it diminish linearly downto 50
cm. 

Daisy will interpolate between these three development stages.

*** DS_Rel

Example:

    (sow ("Spring Barley"
          (Root (rootdens "DS_Rel" 
                          (entries (0.0 (0.0 1.0) (1.0 0.0))
                                   (1.0 (0.0 1.0) (1.0 0.0))
                                   (2.0 (0.0 1.0) (0.8 1.0) (1.0 0.0)))))))

This works like DS_Depth, except that instead of using absolute
numbers for soil depth, we now use numbers relative to the current
potential root depth, as calculated by the root submodel.  With
DS_Depth, the calculated root depth is ignored.

*** Depth_Depth

    (sow ("Spring Barley"
          (Root (rootdens "Depth_Depth" 
                          (entries (00.0 (0.0 1.0) (10.0 0.0))
                                   (50.0 (0.0 1.0) (50.0 0.0))
                                   (75.0 (0.0 1.0) (40.0 1.0) (75.0 0.0)))))))

This works like DS_Depth, except that instead of using the development
stage as index, we now use the calculated root depth.  Obviously, this
doesn't work when the root distribution change while the root depth
remain constant, as implied by the DS_Depth example.

In general, it is recommended that DS_Depth or Depth_Depth is used
when actual measurements has been made, and DS_Rel is used for general
xscrop parameterizations since soil conditions may prevent the root from
reaching the same depth as the measured values.

* 2.00

** GNU/Linux i386 port.

* 1.99

** Harvest cut_delay in the default crop model changed.

It is now a function of the amount of dry matter removed (measured in
kg/ha), rather than of the fraction of dry matter removed.

This affects none of the standard parameterizations.

** Harvest cut_delay now limits development as well as production.

** Crop height is now limited by WStem in the default crop model.

The Canopy parameter 'HvsWStem' put an upper limit on the crop height
based on WStem, relative to the crop height at flowering.  By default,
it need 200 g DM / m^2 to reach full height.

This affects stem and leaf harvest for monocrops, and light
competition for multicrops.

** Change LAI distribution in default crop parameterizations.

LAIDist0 and LAIDist1, which controls how the leafs are distributed
vertically, have been modified for all parameterizations of default
crop model to something that makes more sense.  

This will only affect leaf harvest for most simulations, but have
great effect for grass with multiple cuts, and for multicrops.

** You can now set MinDens in the Gerwitz+Page74 root density model.

** New "Root Density" log model.

* 1.98

** Better errors for (some) illegal parameter values.

You now get line and column number for where they appeared.  

Lots of code was changed to do this, so if some of the errors seem
wrong, let me know.

* 1.97

No user visible changes.

* 1.96

** New 'cut_stress' factor in the standard crop model.

After a cut production may be delayed for some days, depending on how
large a fraction that was removed.

(defcrop "Grass with Cut Delay" Grass 
  (Harvest (cut_delay (0.4 7) (0.8 28))))

** Bug fixes.

* 1.95

** Bug fixes.

* 1.94

** Log file change.  

Moved remaining CrpAux variables to Prod in the standard crop model.
The log.dai file has been updated.

* 1.93

** Log file change.  

StemRes was moved from CrpAux to Prod in the standard crop model.

* 1.92

** Log file change.  

Phenology has been merged into Devel in the default crop model.  This
mean log parameterizations logging DS must be updated.  

A number of log variables have been moved from CrpAux to CrpN in the
default crop model.  The most likely to be logged are Fixated and
AccFixated. 

The parameterizations in 'log.dai' has been updated.

** "N Balance" now logs fixation.

* 1.91

** Added description string for fertilizer parameterizations.

* 1.90

** Now give better error old style root AOM initialization.

** Deposition is now logable, and logged in "N Balance".

* 1.89

** Root density is now a component.

This means that the 'SpRtLength' and 'DensRtTip' root parameters have
changed, and that parameterizations that sets these must be updated.

Instead of 

(defcrop MyCrop default
  ;; ...
  (Root (SpRtLength 100.0)
        (DensRtTip 0.1)

use 

(defcrop MyCrop default
  ;; ...
  (Root (rootdens "Gerwitz+Page74"
                  (SpRtLength 100.0)
                  (DensRtTip 0.1))

If you don't have any better value for SpRtLength or DensRtTip than
the above, you don't need to specify them.  They are the default
values.
 
* 1.88

** "N Balance" and "Carbon Balance" now contains harvest information.

* 1.87

** Mark "CrpN SeedN" as obsolete in default crop model.

Use "Prod NCrop" instead.  You will need to update the crop library to
get rid of the warnings.

** Make it possible to log nitrogen added with seeds.

It has been added to the "Fertilizer" log parameterization.

** Lots of updates to the log parameterizations.

* 1.86

** Count initial AOM when estimating initial SMB.

** By default, assume 1.2 t DM/ha old root residuals at initialization.

To disable, specify an empty value for "am" in OrganicMatter.

The effect if the above two changes is a much reduced need for a
"warming period" at the start of the simulation. 

** The syntax for "root" AM has changed, everything has default values.

Typically, you would just override the "weight" variable with your own
estiimate.

* 1.85

** New top level attributes "directory" and "path".

You can now specify where Daisy shall put log files, and where it
shall look for input files, directly from the ".dai" file.  The
following should be put near the top of the dai file:

   (directory "C:/my/daisy/setup/output")
   (path "." "C:/my/daisy/setup" "C:/daisy/lib")

This will make Daisy run in the directory "/my/daisy/setup/output" (or
"C:\my\daisy\setup\output"), and also generate its log files there.

It will search for input files first in the directory where it run
("." means "current directory"), if it doesn't find them there it will
look in "C:\my\daisy\setup", and if it isn't there either, it will
finally look in "C:\daisy\lib" before giving up.

PS: Remember to use "/" instead of "\" for file names in ".dai" files.

** Bug fixes.

* 1.83

** More Borland work.

** Don't log C harvest per default.

* 1.82

The harvest log model now gives information about the carbon content.

* 1.80

** Now compiles with C++Builder 5.0 via the commandoline and Cygwin make

** Now includes a C++Builder 5.0 project file.

** Now compiles out of the box with Cygwin and "make daisy.exe"

** Bug fixes.

* 1.79

** Log the evapotranspiration used, not just the one in the weather file.

** Bug fix: ignored last root node in some circumstances.

* 1.78

** Keep track of C in crop.

Note: This means the CStem, CDead, CSOrg, CRoot and CLeaf harvest
parameters of the standard crop module can no longer be set.  Instead
set the E_Leaf, E_Stem, E_SOrg, E_Root production parameters.  You can
also log the C content with the CStem, CSOrg, CRoot and CLeaf in
production. 

* 1.77

** Ported to GCC 3.0

** You can now plant emerged crops.  

Define the crop like below, and use the normal "sow" command.

(defcrop "Maize for Planting" "Maize"
 (Phenology (DS 0.269081))
 (Prod (CH2OPool 12.6672)
       (NCrop 1.41396)
       (WLeaf 15.7116)
       (WRoot 26.4654)
       (WStem 10.9263))
 (Root (Depth 51.9679)))

* 1.76

Fix logging state for 'with-column'.

* 1.75

Fix logging of speed in Bioincorporation.

* 1.74

** More flexible dimensions in logs.

The character & in a dimension string in a log description will be
replaced by "h" if you use "hourly" (with no arguments) as a log
condition, "d" if you use "daily" (with no arguments) as a log
condition, "w" if you use "weekly" (with no arguments) as a log
condition, "m" if you use "monthly" (with no arguments) as a log
condition, "y" if you use "yearly" (with no arguments) as a log
condition, and "dt" otherwise.  All the standard log descriptions
have been updated to use "&".

This means that if you for example put

  (output ("N Balance" (when weekly)))

in your setup, "N Balance" will be logged weekly, and the flux
dimensions will be correct.

** Fixed off-by-one error for flux bottom in logs.

* 1.73

Fixed logging incorporated inorganic matter.

* 1.72

Fixed some dimension strings.

* 1.71

Fixed bug in PMSW checkpointing.

* 1.69 

Fixed bug in Penman-Monteith.

* 1.68

NH4 Volatilization is now in the Fertilizer log.

* 1.66

Disabled 'old' and 'old2' hydraulic models, and 'file' and 'hourly'
weather models.

* 1.60

** New 'EpFacDS' Canopy parameter.

This PLF allow the crops potential evapotranspiration adjustment to
depend on the development stage of the crop.

* 1.59

** Checkpoints should work now.

* 1.58

** Tillage can now affect background mineralization.

See the 'som_tillage_factor', 'smb_tillage_factor' and 'tillage_age'
parameters to 'OrganicMatter'.

* 1.55

** New 'density' flag to the 'interval' select method for logging.

Set 'density' to true to make Daisy log the average (in height)
content of the interval, rather than the total content.

** Fixed bad initialization crash.

Found by ML.

** The PM pet module now gives more realistic results.

Bugs found by SH.

* 1.54

** Adjusted DSLAI05 in 'Grass' parameterization.

Should give larger initial harvest.

** The ScreenHeight weather parameter is now respected.

This affects the 'PM' potential transpiration model.

* 1.51

** Checkpoints now include input statements and definitions automatically.

The 'libraries' attribute has been removed.
Checkpoints still doesn't work, though.

* 1.50

** Minor reference manual enhancements.

It will now state if submodel as a full or partial value.

* 1.48

** Fix 'monthly' and 'yearly' conditions.

Used to match every hour in the month or year, except the last.
Reversed that.

** You can now log applied fertilizer.

Use the new 'Fertilizer' log model, defined in 'log.dai'.

** Fix loading global files (bug introduced in 1.47).

* 1.47

** New default SMB and SOM parameters. 

These should have little effect on mineralization, but give a more
realistic estimate of the size of the SMB pools.

It *will* have a large effect if you are running without a starting
period, and have large initial AOM pools.

** New fertilizer types in 'fertilizer.dai'.

** More work on the winter wheat and winter barley parameterizations.

* 1.46

** Fixed link order dependency crash (MAYBE).

* 1.45

** Fixed crash related to the am/cleanup AOM introduced in 1.40.

** Defaults for the 'minimum', 'from' and 'to' precision fertilize params.

* 1.44

** New 'first_year_utilization' fertilizer parameter.

Use this together with the new 'equivalent_weight' fertilize
parameter, to let Daisy calculate the amount on organic fertilizer to
apply, to get an inorganic nitrogen equivalent.  Example:

  (fertilize (pig_slurry (first_year_utilization 0.20))
             (equivalent_weight 50.0)) ; kg N/ha

Here, we specify that 20% of the nitrogen (organic or inorganic)
contained in pig slurry can be utilized the first year, and that we
want to fertilize with an amount of pig slurry corresponding to 50 kg
N/ha.  Daisy will then apply, say, 3 ton DM/ha pig slurry, depending
on the N content of the slurry.

** New 'precision' fertilize parameter.

This allows you to let the amount of fertilizer you apply depend on
the nitrogen content of the soil in a specified zone.  Use it like
this: 

  (fertilize (n50)
             (precision (target 150.0) ; kg N/ha
                        (minimum 50.0) ; kg N/ha
                        (from 0.0) (to -100.0)))

Daisy will measure the amount of fertlizer in the specified zone (from
the soil surface to 100 cm depth), and supply enough extra nitrogen to
reach the specified target (150 kg N/ha).  If less than minimum (50 kg
N/ha) nitrogen is needed in order to reach the target, no fertilizer
will be applied.

You can use the 'precision' parameter with organic fertilizer as well,
in that case you must also specify the 'first_year_utilization'
fertilizer parameter. 

* 1.43

** New 'set_subsoil_irrigation' action.

For irrigating directly in the soil.  Unlike the other irrigation
actions, this one specifies a permanent flux.  This means the
irrigation will continue until you explicitly turn it off by
specifying a flux of zero.

Use it like this

    (set_subsoil_irrigation 1.5 (from -5.0) (to -10.0)
                            (solute (NH4 0.1) (NO3 0.7)))

This will specify an irrigation flux of 1.5 mm/h to be incorporated in
the soil between 5 and 10 cm depth.  The water will contain 0.1 mg
NH4-N/liter and 0.7 mg NO3-N/liter

If you call 'set_subsoil_irrigation' multiple times, the last call
will have effect.

Use

    (stop_subsoil_irrigation)

to turn off subsoil irrigation.

** Changed irrigation solute units from g/mm to mg N/liter.

* 1.42

** 'irrigate_top' action changed to 'irrigate_overhead'.

The old command is still available for compatibility reasons, but
marked obsolete and will be removed in at future version.

Also, the 'irrigation_top' log variables in the 'default' bioclimate
model hgave been renamed to 'irrigation_overhead'.  The 'log.dai' file
has been updated accordingly.

* 1.41

** New 'NEWS' file.

Read this to keep track of Daisy.

** Changed default log file ending from '.tab' to '.dlf'.

Those log files that contains a header now end with '.dlf' instead of
'.tab'.  This affects the build-in 'harvest' log model, as well as all
the 'table' parameterizations found in 'log.dai'.  The 'PMSW'
parameterization uses the old 'table1' log model, doesn't have a
header, and still output to 'pmsw.tab'.

** SoilNO3 and SoilNH4 now have default values.

You no longer have to specify the initial nitrate and ammonium content
when using the 'default' column model.

By default, the nitrate concentration in the water will be half of the
WHO recommended maximum for drinking water, i.e. 50 mg NO3 per liter.
The total ammonium content will a tenth of that.

Suggested by: SH.

* 1.40 

** New 'repeat' action.

The 'repeat' action take a single unnamed argument, which itself
should be an action.  That action will be repeated from the start, as
soon as it is finished.  If you for example have defined a 'sbarley'
manegement strategy, and you want to keep growing spring barley for
ever and ever, you can specify the manangement like this

        (manager (repeat sbarley))

More likely, you will use a 'while' to limit the repetition.

        (manager activity
           (while (wait (at 1998 7 7 1))
              (repeat sbarley))
           stop)

This will keep growing spring barley until 7/7 1998.

** New 't' action.

The 't' action will do nothing, just like 'nil', but will never be
"done".  This means that

        (while t (activity sbarley stop))

will do the same as

        (activity sbarley stop)

** New 'am/cleanup' AOM pool.

Small AOM pools will be merged into the 'am/cleanup' pool instead of
being mineralized instantly.  This way we avoid mysterios "spikes" in
the mineralization log data when small AOM pools are discarded.  

** Fixed log after 'default' crop get ripe.

In the 'default' crop model, production and respiration halts after
the crop gets ripe.  However, the last active production and
repsiration values would continue to be logged.  Now, these are
zeroed.

Bug found by: SH.

** Surface irrigation didn't work.  Now it does.

Surface irrigation was ignored.  Now it is applied correctly.

Bug found by: Mette Lægdsmand <Mette.Laegdsmand@agrsci.dk>

** Fixed numeric error in the 'default' bioclimate model.

The symptom was a rare assertion failure.

Bug found by: SH.

* 1.38

** New 'inorganic' column model.

Works like the 'default' model woth regard to water and soil
chemicals, but does not keep track of organic mater or nitrogen in the
soil or surface.  The crops will always have potential N uptake.

** 'Surface Water Balance' now logs production stress.

* 1.37

** New 'vegetation.dai' file, for permanent vegetation parameterizations.

** Fix 'PMSW' log, broken by changes in 1.36.

** Use 'S_external' to log incorporated stuff.

The 'N Balance' and 'Soil Chemicals' logs are updated to do this.

* 1.36

** Log at end of the day.

This means that the data in the daily log files now corresponds to the
day listed.  The used to correspond to the previous day.  Fixed
'daily', 'weekly', 'monthly' and 'yearly' conditions to always match
the last hour in the period.

* 1.35

** A zero stress factor now means no stress.

Previously it went maximum stress.  Specifically, the meaning of
'water_stress' and 'nitrogen_stress' has been inverted.  Zero is no
stress, one is full stress.

** SVAT model integration changes.

The 'PMSW' svat model give feedback
to the rest of the system, by calculating a 'production_stress' which
will be used instead of 'water_stress'.

* 1.33

** SVAT model integration changes.

The 'PT' parameter of the 'default' bioclimate model is gone, instead
there now is a 'svat' parameter.

** The 'crop' action fixed so it no longer will harvest before sowing.

The symptom was that a crop was skipped when started at the wrong
time. 

Bug found by: HS.

* 1.32

** Optional parameters of the 'crop' action wasn't optional.

Now they are.

Bug found by: HS.

* 1.30

** Fix compilation with Borland C++ 5.01.

* 1.29

** Default value for 'dispersivity' (6.0 cm).

** Default value for 'rs_min' is now 30.0 s/m.

* 1.28

** New 'permanent' vegetation model.

It is intended for column on a multi-column simulation where we have
very little knowledge, basically we just need to know the LAI at
different times of the year.

* 1.27

** Column specific weather.

It is now possible to have separate weather for each column.  You can
still have a global weather, it will be used as default for columns
where no weather have been specified.

* 1.26 

** Log files now print a header.

Specifically, the 'harvest' and 'table' log models will now -- by
default -- create a header containing information about the data.

** New 'simple' crop model.

Intended for crops where we have no knowledge of reactions to water or
nitrogen stress.

* 1.24

** New Bentazon, IPU, and MCPP parameterization in 'chemistry.dai'.

** The 'crop' action now supports spraying.

** Added support for ridging.

* 1.23

** Log and harvest dead leaves.

* 1.22

** Added domain and range for PLF's in reference manual.

* 1.21

** Renamed CSMP to PLF.

* 1.19

** New 'fixed' and 'deep' groundwater models.

* 1.18

** New 'set_heat_source' action.

Use this to make a football field green a larger part of the year.

* 1.17

** New 'set_porosity' action.

Use this to change the porosity of the soil, to simulate the effect of
applying heavy machinery on the field.  Works only with the
'M_vG_compact' hydraulic model. 

** New 'M_vG_compact' hydraulic model.

This one has variable porosity.

** New 'set_detention_capacity' action.

This one is intended for simulating changed to the soil surface
ruggedness. 

* 1.14 

** New 'AOM' log model.

To keep track of a specific pool of added organic matter.

* 1.10 

** Use 'volatilization' instead of 'NH4_evaporation' in AM.

* 1.08

** Made heat and moisture influence user setable.

Affects mineralization, nitrification, and denitrification.

* 1.07 

** Made soil chemicals mix with surface chemicals.

* 1.06

** Added bioincorporation of organic matter from surface to soil.

* 1.05 

** New 'crop' action.

This action specifies a strategy for managing a specific crop or
multicrop.  To create a crop rotation, you can combine these 'crop'
action sequentially using the 'activity' action.

* 1.04

** New 'lysimeter' groundwater model.

* 1.00

** New 'max', 'min' and 'average' selections for the 'table' log model.

** New 'assert', 'message', 'warning', and 'error' actions.

* 0.99

** Added surface runoff.

Controllable with the new 'DetentionCapacity' and 'ReservoirConstant'
surface parameters. 

** Added mixing of inorganic nitrogen between surface and soil.

* 0.96

** New 'fodderbeet' crop parameterization.

** New 'default' weather model.