The simulation results (GPP and LE) of CLM5.0 differ significantly from the observed values

[liangxiaoru]

Dear all,

I used CLM5.0 in CESM2.1.2 for single point simulation, and the simulated site was FR Pue (Puechabon, France) (43.7413, 3.5957) The time is from 2000 to 2014, and the vegetation type is evergreen broad-leaved forest.

Atmospheric forcing data from half-hour observations at Fluxnet2015, surface data in 2000 were made by ./mknoocnmap.pl, mkmapdata.sh,./gen_domain and ./mksurfdata.pl utilities.

The compset I have chosen is I1PtClm50Bgc
（2000_DATM%1PT_CLM50%BGC_SICE_SOCN_SROF_SGLC_SWAV）.

I used the entire data loop to run Spinup for 300 years, bringing the model to a stable state.

However, the variables of interest GPP and LH obtained differ significantly from actual observations, with LH being two to three times larger than the observed values. Here are the single point simulation results.

The attachment contains the Atmospheric driving and surface data, and log files.
Thank you for your feedback.

 

[slevis]

A way to troubleshoot this may be to consider what could lead to overestimated LH and then pursue your hypotheses. For example, maybe the soil is too wet. If the precipitation and the soil texture data are fine, then you may need to come up with other hypotheses. If everything seems fine in the inputs, then you may need to consider how the model parameterizations could cause this behavior.

When you have an update, please let us know; however, we are not able to help during this troubleshooting, unless you report a problem with the model.

 

[slevis]

Of course the community is always welcome to offer feedback and/or help.

 

[liangxiaoru]

A way to troubleshoot this may be to consider what could lead to overestimated LH and then pursue your hypotheses. For example, maybe the soil is too wet. If the precipitation and the soil texture data are fine, then you may need to come up with other hypotheses. If everything seems fine in the inputs, then you may need to consider how the model parameterizations could cause this behavior.

When you have an update, please let us know; however, we are not able to help during this troubleshooting, unless you report a problem with the model.

Thank you for your response!
We have checked the input atmospheric forcing data,
Discovered an error in precipitation data (we entered the cumulative amount every half hour, which should actually be in millimeters per second),
Perhaps excessive precipitation caused LH to be overestimated.
After correction, we obtained simulation results that approximate the observed values.