Potential bugs in calculation Secchi depth in GOTM output file

[Jorrit Mesman]

Dear all,

I have encountered a problem with the Secchi depth ("secchi" output variable) when using the "time_method: mean" option to generate output. This is part of the "v5.2.2-au" version of GOTM - unsure about the integration of Secchi depth into a "main"/"lake" version of the model. In my case, I used a model integration time step of 1 hour. During night, when incoming shortwave radiation is 0, the Secchi depth is computed as 0, whereas during the day, it is computed normally (see attached plot of hourly values). The "mean" time_method averages the Secchi depth values, therefore ending up at about half of the real Secchi depth. It would be better if the mean Secchi depth was computed from the mean shortwave radiation ("rad") profile over that period, but I don't know if this is feasible? I can circumvent the issue by generating point output for Secchi depth at noon, but it seems like an issue that many people could run into. 

There is a secondary potential issue with the computation of Secchi depth, although not as serious. Secchi depth (computed in the diagnostics.F90 script), is based on the "rad" variable. However, "rad" is affected both by g2 (visible) and g1 (non-visible shortwave radiation), and Secchi depth is theoretically only affected by visible light. I don't think this will make a large difference, but perhaps this could be included in the computation. 

Best regards,

Jorrit Mesman

[Jorn Bruggeman]

Hi Jorrit,

 

My starting point would be the expression for Secchi depth that is being used, to see if that could be modified to depend only on attenuation, not on irradiance (e.g., I know of formulations that depend only on the vertical attenuation coefficient of diffuse downwelling irradiance). That would allow Secchi depth to be properly defined at night too. Should that fail, it would be an option to give the Secchi depth a dedicated “missing value” when it is not defined, instead of zero, and to make the temporal averaging scheme exclude such missing values. But that’s more involved.

 

In any case, the trouble is that I’m not familiar with that "v5.2.2-au" branch of GOTM. It does not seem to be in official repository at https://github.com/gotm-model/code. Without seeing the code, it is difficult to comment further.

 

Cheers,

 

Jorn

--
You received this message because you are subscribed to the Google Groups "GOTM-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gotm-users+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gotm-users/50e86313-42bd-404d-973e-07887008ea3bn%40googlegroups.com.

[Jorrit Mesman]

Thanks for the suggestions Jorn! Would you mind sharing the formulations using the attenuation coefficient? This GOTM version is on the Aarhus University / WiT repository (I believe Karsten is involved), but I could see if this could get implemented; it seems the easiest and most sustainable solution...

Best,

Jorrit

Op vrijdag 14 april 2023 om 12:29:58 UTC+2 schreef Jorn Bruggeman:

[Jorn Bruggeman]

Hi Jorrit,

 

Here’s some info:

 

https://doi.org/10.1016/j.rse.2007.06.009

https://ereefs.aims.gov.au/ereefs-aims/gbr4/bgc/hindcast/secchi_kd-490_epipar-sg

 

Ultimately, they use an inverse relationship between Secchi depth and attenuation, which seems to come from this 1929(!) paper:  https://doi.org/10.1017/S0025315400029829.

 

However, these studies describe the link between near-surface attenuation and Secchi depth. Since attenuation is potentially depth-dependent in GOTM (if using feedbacks from biogeochemistry), you may want to somehow consider its vertical variation, e.g., by finding the depth where the remaining radiation is some fraction (e.g., exp[-1.7]=0.18 according to Poole & Atkins) of the surface downwelling value. That could still be done without needing the actual radiation, and therefore should work at night.

To view this discussion on the web visit https://groups.google.com/d/msgid/gotm-users/8869893b-58b7-49e5-a717-b38d018e69b9n%40googlegroups.com.

[Karsten Bolding]

Hi Jorrit

FYI I don't work at Aarhus University anymore and the others working with the GOTM-WET repository at AU has also quit effectively leaving the repository un-maintained. The GOTM part of the repository has anyway not been updated for a very long time as far as I know.

Best regards,

Karsten

To view this discussion on the web visit https://groups.google.com/d/msgid/gotm-users/8869893b-58b7-49e5-a717-b38d018e69b9n%40googlegroups.com.

--

Karsten Bolding

kar...@bolding-bruggeman.com

+4564422058