All notable changes to this project will be documented in this page.
The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.
- Support in toml for dir_input and dir_output options. PR #140
- In setup_reservoirs: Global Water Watch compatibility for determining reservoir parameters.
- In setup_reservoirs: All dowloaded reservoir timeseries are saved to root in 1 csv file. Column headers indicate reservoir id.
- New setup_river_floodplain method PR #123. Note: requires pyflwdir v0.5.7
- New setup_floodplains method, that allows the user the choose either 1D or 2D floodplains
- Default tomls are now using the dir_output option to specify run_default folder.
- in setup_reservoirs: options 'usehe' and 'priorityjrc' are removed and replaced with 'timeseries_fn'. Options are ['jrc', 'gww']. By default None to use reservoir_fn data directly.
- in setup_areamap: name of the added map is based on column name of the vector data (col2raster) instead of name of the vector data file (area_fn). Allows to add several maps from one vector data file.
- Merged river part of setup_hydrodem into setup_rivers functionality, and moved land part of setup_hydrodem into setup_floodplains.
- Bugfix with wrong nodata value in the hydrography method which caused errors for model which where not based on (sub)basins PR #144
- Bugfix with wrong indexing in the river method that could cause memory issues PR #147
- fix error in setup_reservoirs when gdf contains no data in np.nanmax calculation for i.e. damheight #35
- write_forcing with time cftime.DatetimeNoLeap #138 by removing slicing forcing if missings (not needed)
- write_forcing automatic adjustment of starttime and endtime based on forcing content
- When clipping a model from a model with multiple forcing files, a single netcdf is made in write_forcing and the * is removed from the filename.
- add options to calculate daily Penman-Monteith potential evaporation using the pyet package. Depending on the available variables, two options are defined
penman-monteith_tdew(inputs: ['temp', 'temp_min', 'temp_max', 'wind_u', 'wind_v', 'temp_dew', 'kin', 'press_msl']) andpenman-monteith_rh_simple(inputs: ['temp', 'temp_min', 'temp_max', 'wind', 'rh', 'kin']).
New setup_staticmaps_from_raster method and river smoothing algorithm. Correct some bugs linked to soon deprecated staticmaps and staticgeoms objects in hydromt core to work with the new 0.6.0 release.
- update forcing example with multiple forcing files #122
- New window smoothing algorithm in setup_rivers to avoid cells with small river length. Set the min_rivlen_ratio argument to a value larger than zero to apply the smoothing. Note: requires pyflwdir v0.5.6 PR #92
- write_forcing with time of type cftime.DatetimeNoLeap #109
- write_forcing: re-write config in case of multiple forcing files
- read_forcing with multiple files (* key in toml)
- bug in setup_gauges in update mode with crs.is_epsg_code #108
- bug in self.rivers if no staticgeoms and rivmsk is found #113
- bug in wflow_build_sediment.ini template in examples
- wrong defaults in wflow_build.ini teamplate in examples #116
- temporary fix to update staticgeoms basins+rivers in clip_staticmaps (update when moving away from deprecated staticgeoms).
- fix wrong default value for lai_fn in setup_laimaps #119
We now use rioxarray to read raster data. We recommend reinstalling your hydromt and hydromt_wflow environment including the rioxarray package. This enables the writting of CF compliant netcdf files for wflow staticmaps.nc and inmaps.nc. Following an update in xarray, hydromt version should be >= 0.5.0.
- correct float32 dtype for all landuse based maps (by changing values in all lookup tables to floats)
- write CF-compliant staticmaps.nc and inmaps.nc
- CRS issue when deriving subcatch for user defined gauges in setup_gauges
- update times in config depending on forcing date range availability in write_forcing methods #97
- In the naming of the generated hydrodem map, it is now specified if a D4 or D8 conditionning has been applied for land cells.
- uint8 dtype wflow_rivers and wflow_streamorder maps
- except for coordinates (incl x_out and y_out) all variables are saved with at most 32 bit depth
- new dtype and nodata arguments in setup_constant_pars
- read boolean PCRaster maps with int type to be consistent with netcdf based maps
- use latest hydromt github version for the test environment files.
- in setup_glaciers predicate to intersects glacier data with model region is 'intersects' (the old 'contains' was not used anyway due to a bug in core).
- in setup_reservoirs and setup_lakes the predicate 'contains' to open data is now officially used after a bugfix in hydromt core (cf #150).
- nodata argument to setup_areamap with a default of -1 (was 0 and not user defined).
- setup_riverwidth method deprecated (will be removed in future versions) in favour of setup_rivers. We suggest to remove the setup_riverwidth component from your ini files.
- setup_rivers calculate river width and depth based on the attributes of the new river_geom_fn river geometry file. We suggest adding "river_geom_fn = rivers_lin2019_v1" to the setup_rivers component of your ini files.
- In setup_soilmaps the interpolation of missing values (interpolate_na function) is executed on the model parameters at the model resolution, rather than on the original raw soilgrids data at higher resolution. This change will generate small differences in the parameter values, but (largely) improve memory usage.
- Possibility to use any dataset and not just the default ones for setup_laimaps, setup_lakes, setup_glaciers. See the documentation for data requirements.
- Possibility to write_forcing in several files based on time frequency (fn_freq argument).
- setup_hydrodem method for hydrological conditioned elevation used with "local-inertial" routing
- workflow.river.river_bathymetry method to derive river width and depth estimates. Note that the new river width estimates are different and result in different model results.
- moved basemaps workflows (hydrography and topography) from HydroMT core. Note that HydroMT_Wflow v0.1.3 there should be used together with HydroMT v0.4.4 (not newer!)
- new ID columns for the outlets staticgeoms
- new
index_colattribute to setup_gauges to choose a specific column of gauges_fn as ID for Wflow_gauges
- Calculation of lake_b parameter in setup_lakes.
- Add a minimum averaged discharge to lakes to avoid division by zero when computing lake_b.
- When writting several forcing files instead of one, their time_units should be the same to get one Wflow run (time_units option in write_forcing)
- Filter gauges that could not be snapped to river (if snap_to_river is True) in setup_gauges
- Avoid duplicates in the toml csv column for gauges
- Fill missing values in landslope with zeros within the basin mask
- prevent writing a _FillValue on the time coordinate of forcing data
This release adds pyflwdir v0.5 compatibility and a data_catalog of the used data to the write_method.
- write data_catalog with the used data when writing model
- tests on staticmaps dtype
- TOML files only contains reservoir/lake/glacier lines when they are setup and present in the model region.
- pyflwdir v0.5 compatibility: changes from stream order bugfix and improved river slope
- Fixed docs with rtd v1.0
- Wrong dtype for Wflow_gauges
- Removed unnecessary glacier/lake/reservoir lines from the TOML, fixes a bug if missing glacier
This release implements the new results attributes for Wflow.
- Add results attributes for Wflow and read_results method (including test+example).
- Add f_ parameter in soilgrids
- Support soilgrids version 2020
- Setup_areamap component to prepare maps of areas of interest to save Wflow outputs at.
- Support Wflow_sediment with vito landuse.
- New utils.py script for low_level Wflow methods.
- wfow_sbm.toml remove netcdf output.
- Wflow_soil map is now based on soil texture calculated directly from soilgrids data
- test cases change toml and Wflow_soil.map
- Wflow_sbm.toml now includes links to staticmaps of glacier parameters and outstate of glacierstore is added.
- Fix f parameter in soilgrids
- Full reading and writing of Wflow filepaths depending on the toml file (including subfolders).
- The Wflow_gauges now contains river outlets only (instead of all outlets).
- Added Wflow_plot_results example.
- Fixed staticmaps_to_mapstack example.
This release adds more functionnality for saving forcing data for Wflow and fixes several bugs for some parameter values and soilgrids workflow.
- Write the forcing with user defined chunking on time (default is 1) and none on the lat/lon dimensions (makes Wflow.jl run much faster).
- Rounding of the forcing data with user defined number of decimals (by default 2).
- Progress bar when writing the forcing file.
- Remove unused imports.
- Fixed a mistake in the computation of the lake_b parameter for Wflow.
- Missing no data values for soilgrids workflows.
- Streamorder reclass function for Manning roughness.
- New behavior of apply_ufunc from an update of xarray for passing attributes (need to specify keep_attrs=True).
- Added changelog.
- Tests without hydroengine for the reservoirs (too long).
Initial open source release of HydroMT Wflow plugin, also published on pypi. Noticeable changes are listed below.
- Minimum HydroMT plugin template in the plugin-boilerplate branch.
- Default filename for the forcing file created by HydroMT (when the one in config already exists).
- Implement new get_basin_geometry from HydroMT core.
- Consistent setup functions arguments for data sources ('_fn').
- Rename hydrom_merit source to merit_hydro (updated version of data-artifacts).
- Bugs using the clip functions
- Initial version of the documentation on github-pages.
- Latest and stable version of the documentation.
- Setup Binder environment.
- Add examples notebooks for the documentation.
- Initial tests for Wflow and Wflow_sediment.