Page 1 of 1

Satellite PAR formula and Gregg-Carder light model

Posted: Wed Nov 24, 2021 2:57 pm America/New_York
by clays_bio
Hi everyone!

I'm modelling hourly surface irradiance from 400-700nm using the Gregg-Carder model and trying to figure out how to properly convert it to the same units as the satellite PAR product so it can be scaled, but I'm having trouble following the steps in the PAR model. I would like to use the satellite planar value in mW/cm2/um instead of the scalar value in Einsteins/m2/day.

For Gregg-Carder irradiance, I end up with a matrix of irradiance values in Watts/m2/nm that I can convert to mW/cm2/um, where each row is an hour interval and each column is a waveband.

For satellite PAR, my understanding is that the satellite PAR value in mW/cm2/um is an average value over the whole PAR spectrum and the whole day, and that multiplying it by a factor of 1.193 accounts for the transformation of units, planar to scalar geometry, and integration over waveband and time, with a little bit of error, giving us the final PAR product in Einsteins/m2/day.

If this is right, then I should divide satellite PAR by 1.193 to get it back to units of mW/cm2/um, and take the average value of my Gregg-Carder matrix in mW/cm2/um units for comparison.

Can anyone confirm is this is the correct way to do it?



Re: Satellite PAR formula and Gregg-Carder light model

Posted: Mon Nov 29, 2021 11:00 am America/New_York
by OB.DAAC - SeanBailey
Yes. PAR is calculated internally in mW/cm2/um, and at the very end of the code:
/* Convert to E/D/m^2 */
if (par[ip] != BAD_FLT)
par[ip] *= 1.193;

So...dividing by 1.193 will get you back into radiance.


Re: Satellite PAR formula and Gregg-Carder light model

Posted: Thu Dec 02, 2021 11:01 am America/New_York
by clays_bio
Thanks Sean!

I think my main confusion is the value that the satellite planar radiance represents, since I want to make sure I end up with the equivalent value modelled from Gregg-Carder. I was mixing up steps of integrating vs averaging and converting between energy and number of photons per waveband, and then confusing myself even more trying to relate it to the satellite model.

I just realized my method of calculating the average of the Gregg-Carder matrix is a little different too -- I'm taking the average of the entire matrix, but I should be averaging over waveband at each hour interval, and then integrating over day and dividing by 24 hours, which gives a slightly different result. Can you confirm that the latter is the value I should be using from my GC matrix to get a value equivalent to the satellite radiance in mW/cm2/um?


Re: Satellite PAR formula and Gregg-Carder light model

Posted: Fri Dec 03, 2021 7:35 am America/New_York
by OB.DAAC - SeanBailey
Well, I can't *deny* that approach is's effectively what the satellite algorithm is doing :D