Error messages

In progress.... By far not an exhaustive list of the error messages that ED can throw.

Error Message: Bad data for namelist object nl%XXXXX

This is a problem with the ED2IN settings file. More often than not the error is actually with the variable coming before or after the one actually throwing the error.
Examples:

• Including a variable name that doesn't exist for the version of the code you're running
• Not including a variable that does need to exist
• Punctuation errors

Error Message: RKMS Integration Ends

There are many potential causes of the integrator crashing but broadly they fall into two categories, either the model fails to produce biologically and physically meaningful output (i.e. values fall outside of pre-specified ranges) or the stepsize of the integrator gets to small. There are a lot of messages generated when the integrator stops and the clues to the cause are usually found there. Generally one should start from the top of these messages because the first message generally states which category the crash falls into and the cause of the error (i.e. which variables were unrealistic).

Diagnosis:''' Leaf Temperature increases to >> 300K in one time step

• First check: make sure that the meteorological driver is correctly set. This includes making sure that data provided to ED2 are in UTC (instead of local time), and that NL%IMETAVG in ED2IN is set consistently with each meteorological driver (i.e., do not assume NL%IMETAVG can be the default, there is no default, and it may need to be defined differently for different sites). This is likely the case if the model is crashing when simulating times near sunrise or sunset.
• preferred fix: reduce integrator time steps (NL%DTLSM, NL%RADFREQ e.g. from 900 to 300)
• "desperation" fix: increase lai_min, (at 900s time step, increasing it to 0.05 reduced frequency of this error). Note that this will cause any cohort below this threshold to be ignored for almost all calculations (including photosynthesis).
• This error usually occurs when there is low leaf area in a cohort and the environmental conditions are changing. The integrator has difficulty converging due to there being many flux terms in the leaf energy budget and little thermal mass to buffer. Decreasing DTLSM helps because it allows the shortwave and longwave radiation terms to be updated more frequently
• If this error occurs at the end of the month or right after an event, it is likely that structural changes to the vegetation have caused the instability (and the above fixes are less likely to help)
• If this occurs during leaf flush or leaf fall it is likely that the transient "small" amount of leaf mass is the culprit.