I am currently conducting a simulation study on GPP (Gross Primary Productivity) for July 2010 using CLM5.0. I have encountered a problem. I used the output hourly file variables "LAISUN", "LAISHA", "PSNSUN", and "PSNSHA" to calculate GPP as follows: GPP=(LAISUN * PSNSUN + PSNSHA * LAISHA） * 1.2e-5 * (12/44) * 3600. However, I cannot obtain the monthly GPP results that match the model output. What could be the reason for this?
CLM5.0 GPP
- Thread starter yuecheng
- Start date
I'm not sure, are you missing a conversion factor? You have 12/44 which presumably is gC/gCO2, but I think you need to multiply by 44 gCO2/1 mole CO2, and I'm not sure what 1.2e-5 represents.
I see this in the code:
gpp(p) = psnsun_to_cpool(p) + psnshade_to_cpool(p)
So I would check to see if GPP = PSNSUN_TO_CPOOL + PSNSHA_TO_CPOOL in the history output.
I see this in the code:
gpp(p) = psnsun_to_cpool(p) + psnshade_to_cpool(p)
So I would check to see if GPP = PSNSUN_TO_CPOOL + PSNSHA_TO_CPOOL in the history output.
Hi, Keith Oleson
1.2e-5 represents the conversion of umol CO2 m2 s to g C m2 s. After removing 12/44, I calculated the GPP for July 2010 using the hourly LAISUN, PSNSUN, PSNSHA, and LAISHA variables from CESM output for July 2010, and validated it with the CESM output of July GPP. However, I found that it still could not match. In the figure, GPP (hour) represents the GPP for July calculated using the hourly LAISUN, PSNSUN, PSNSHA, and LAISHA variables, while GPP (CESM) uses the GPP value for July 2010 from CESM output. Note: I have already converted the seconds in the units to months.
1.2e-5 represents the conversion of umol CO2 m2 s to g C m2 s. After removing 12/44, I calculated the GPP for July 2010 using the hourly LAISUN, PSNSUN, PSNSHA, and LAISHA variables from CESM output for July 2010, and validated it with the CESM output of July GPP. However, I found that it still could not match. In the figure, GPP (hour) represents the GPP for July calculated using the hourly LAISUN, PSNSUN, PSNSHA, and LAISHA variables, while GPP (CESM) uses the GPP value for July 2010 from CESM output. Note: I have already converted the seconds in the units to months.
Attachments
I think you have the untis conversions correct now. Although you say that you've converted from seconds to months, but the units on your plot say gC/m2/s.
The code also uses 1.2011e-5, not 1.2e-5, but that shouldn't be causing that big of a difference. So I'm not sure why there is a difference. Have you tried the same calculation using output at the model time step (half-hourly)? In other words, compare the GPP output at half-hourly with that calculated from LAISUN, LAISHA, PSNSUN, PSNSHA at half-hourly. Do they disagree at every time step or not? Maybe that would point to where the differences start occurring. Maybe there is some daytime/nightime inconsistency there for some reason.
The code also uses 1.2011e-5, not 1.2e-5, but that shouldn't be causing that big of a difference. So I'm not sure why there is a difference. Have you tried the same calculation using output at the model time step (half-hourly)? In other words, compare the GPP output at half-hourly with that calculated from LAISUN, LAISHA, PSNSUN, PSNSHA at half-hourly. Do they disagree at every time step or not? Maybe that would point to where the differences start occurring. Maybe there is some daytime/nightime inconsistency there for some reason.