Initial soil condition

[lucia]

Hello,
i am currently running a single-point run for the Qinghai-Tibet Plateau(alpine meadow) site with 2000_DATM%1PT_CLM50%BGC_SICE_SOCN_MOSART_SGLC_SWAV.
I only changed the soil texture (sand and clay content), but my simulation results were very different from the actual observations. I noticed that my soil moisture content was high on the first day of the year and showed a downward trend in the first 100 days, which was very different from the measured values.





In creating surface dataset, I have specified the percent clay(11%) and sand(32%) as given in the literature.
I would like to check the initial soil condition but I am not sure where and what to check.Maybe it has something to do with the energy imbalance, because I've only been spinping up for 600 years, and I'm not sure if the soil is in equilibrium.
Or, if there is any other suggestions regarding this issue, I would really appreciate them.

Thank you.

 

[oleson]

For release-cesm2.1.3, the soil moisture for a "cold start" is set in src/biogeophys/WaterStateType.F90. For vegetation it is initialized to 0.15 volumetric.
A 600 year spinup should be enough to ensure the soil moisture is in equilibrium, particularly the layer2 you are analyzing.
I would suggest checking the atmospheric forcing you are using to make sure it is being read in correctly and is appropriate for the site. In particular, I see a decline in CLM soil moisture in the first 200 days, while the observations show a large increase. This may indicate that the precipitation forcing is not enough to keep up with the sum of evaporation and runoff.

 

[lucia]

Oleson, Thanks for your reply.I have checked the atmospheric forcing data that I was using and I find that the precipitation forcing in the first 100 days was indeed not much, probably not enough to maintain the sum of evaporation and runoff, as you say.


However, I observed that the soil moisture content value of the first day simulated by clm was maintained at about 0.3, while the actual observed value was 0.2 or even lower. May I ask how to change the soil moisture content of the first day.I checked src/biogeophys/WaterStateType.F90 and I would like to ask if I want to change the soil water content on the first day ,should I change this%h2osoi_vol_col(c,j) = 0.3_r8, is it correct?

 

[oleson]

That is the initial soil moisture for the pervious road in the urban canyon model (if col%itype(c) == icol_road_perv).
For vegetation you would change the 0.15_r8 value.
From your comments, however, it looks like you've done a spinup. If you are initializing from that spinup, then the initial value must be about 0.3 taken from the initial file generated from the spinup presumably. The setting of 0.15_r8 is for a "cold start" when there isn't an initial file.

 

[lucia]

That is the initial soil moisture for the pervious road in the urban canyon model (if col%itype(c) == icol_road_perv).
For vegetation you would change the 0.15_r8 value.
From your comments, however, it looks like you've done a spinup. If you are initializing from that spinup, then the initial value must be about 0.3 taken from the initial file generated from the spinup presumably. The setting of 0.15_r8 is for a "cold start" when there isn't an initial file.

Thanks a lot! I kind of get it.
I have checked that when the model reached a steady state, the value of the first day of every year was around 0.3, which was consistent with what you described, obviously inconsistent with my measured situation. May I ask if it is possible for me to change this value?

what is more, My running process is as follows:
First I created case, using 1979-1989 data（My own data set） to complete the 600 year AD_spin up, I have used the 2000 compset for the two spinups but are using year 1979-1989 for atmospheric forcing
Finally I create case2 and complete the simulation using 1990-2018 data.(Use the initial file generated in the previous step)
I'm not sure if that's the right thing to do, but I don't have any data on atmospheric forcings going back any further than 1979.can you please give me some advise?

 

[oleson]

The value of 0.3 is what you arrived at, at the end of your spinup. I wouldn't then change it arbitrarily. You will be out of equilibrium again.
Your approach to spinup seems reasonable.

 

[lucia]

Oleson, Thanks for your reply. I still have some questions about the following points:
1. The soil moisture content value I got at the end of spinup on the first day was 0.3. What determined this 0.3? If I change the value of my atmospheric forcing data, does this 0.3 change?
2. If I change the site and change the vegetation type, will my soil moisture content be different from 0.3?
3. For clm, is it possible to make the soil moisture content simulated by me closer to the measured value?
4. I used atmospheric forcing data from 1979-1989 (rather than forcing a spinup on atmospheric data from 1901-1920) to carry out spinup,when the Industrial Revolution is supposed to have taken place. Is this spinup out of line with reality? Can you give me any advice?

 

[oleson]

1. The model interaction with the atmospheric forcing over the entire spinup determined that 0.3. Yes, if you use different atmospheric forcing, then the 0.3 is likely to change
2. Probably
3. Soil moisture is a result of very complicated interactions of the forcing data with the biogeophysics/biogeochemistry of the model. You'll have to experiment to see how different surface properties affect soil moisture.
4. I think this is fine.

 

[zzwflood]

为了进一步减少 CLM5.0在模拟径流时的偏差，怎样将土壤蒸发方案中形成 DSL的临界含水量设置为田间持水量？

 

[zzwflood]

谢谢