Chlorophyll-A algorithm question

samsiuca · by **samsiuca** » Thu Mar 16, 2017 4:40 am America/New_York

Hello,

The algorithm product summary https://oceancolor.gsfc.nasa.gov/atbd/chlor_a/ describes the result as "near surface concentration". How deep into the water does "near-surface" imply? An infinitesimally thin layer on the surface of the water? Or perhaps some centimetres/metres down (the Hu et. al paper referenced in the product page uses <30m deep waters for testing)?

Thanks,
- Sam

gnwiii · by **gnwiii** » Thu Mar 16, 2017 7:23 am America/New_York

Roughly speaking, you get a weighted contribution from the underwater concentration. The wavelengths used in chlor_a algorithms can penetrate many meters, but due to attenuation of the light on both downward and return paths, the contribution to Rrs is heavily weighted towards the surface. You can get a rough idea of the weighting by plotting exp(-2*Kd*z) (using z for depth). One practical consequence is that remote sensing chlor_a underestimates the contribution of deep chl maxima (doi:10.3390/rs9030210 gives to references to several papers that discuss the problem).

samsiuca · by **samsiuca** » Thu Mar 16, 2017 11:48 am America/New_York

Thanks for the quick response gnwiii!

When you say ""Kd", I assume you are referring to https://oceancolor.gsfc.nasa.gov/atbd/kd_490/? Is there a particular paper(s) or study from where the -2 coefficient comes from (as https://oceancolor.gsfc.nasa.gov/docs/technical/NASA-TM-2016-217551.pdf just writes exp(?Kd*z))?

- Sam

gnwiii · by **gnwiii** » Fri Mar 17, 2017 3:45 am America/New_York

The factor of two comes from the double length of the path down and back up. The simple exponential formula gives a rough and ready approximation for the contribution from various depths, so Kd_490 can be used "for purposes of illustration".

samsiuca · by **samsiuca** » Fri Mar 24, 2017 8:59 pm America/New_York

Hello,

Within the file https://oceancolor.gsfc.nasa.gov/docs/ocssw/chl_8h_source.html , I see that a chlmax variable has been set to 1000.0 , which I assume is the same limit Sean Bailey mentioned in https://oceancolor.gsfc.nasa.gov/forum/oceancolor/topic_show.pl?pid=25280;hl=chlor%20100. Is there a reason why this cap is in place (it seems a bit arbitrary)?

Best Regards,
- Sam S.

by **OB.DAAC - SeanBailey** » Sat Mar 25, 2017 11:41 am America/New_York

Yes, it is arbitrary, but it is also an order of magnitude larger than any value we would have any expectation to be real (and thus quite extreme).
The data set used to generate the algorithm has very few concentrations higher than 50mg/m3, and none at the CHLWARN flag max of 100 mg/m3.
So, if the algorithm returns a value of 1000, odds are quite high that it is pure fantasy.

Sean