Kd490 data VIIRS vs MODIS

[tomtomascik]

Hi,
I would greatly appreciate it if someone could explain why there is such a large discrepancy in the Kd490 values taken by VIIRS and MODIS instruments. Here is an example:

VIIRS: https://coastwatch.pfeg.noaa.gov/erddap/griddap/nesdisVHNSQkd490Daily.graph?kd_490%5B(2020-02-11T12:00:00Z)%5D%5B(0.0)%5D%5B(20.643749999999997):(20.568749999999994)%5D%5B(92.26875000000001):(92.38125000000002)%5D&.draw=surface&.vars=longitude%7Clatitude%7Ckd_490&.colorBar=KT_turbid%7C%7C%7C0%7C2%7C&.bgColor=0xffccccff

MODIS: https://coastwatch.pfeg.noaa.gov/erddap/griddap/erdMH1kd4901day.graph?k490%5B(2020-02-11T12:00:00Z)%5D%5B(20.64583):(20.5625)%5D%5B(92.27084):(92.39584)%5D&.draw=surface&.vars=longitude%7Clatitude%7Ck490&.colorBar=KT_turbid%7C%7C%7C0%7C6%7C&.bgColor=0xffccccff

Thank you for your assistance.
Cheers,
Tom

[OB.DAAC - SeanBailey]

Tom,

The site you reference is a NOAA site, not NASA. The site does indicate that the MODIS data they present comes from NASA, however the VIIRS data is from NOAA. That said, I looked at the corresponding NASA product, and overall, the Aqua MODIS and SNPP VIIRS Kd490 products agree quite well. Certainly, there are some areas where they diverge, but they are consistent overall. Below are data from the day you referenced, but on a larger geographic scale to give you context:

viirs_kd_11Feb2022.png (10.5 KiB) Not viewed yet

aqua_kd_11Feb2022.png (5.83 KiB) Not viewed yet

You had zoomed into the spot where the two differ the most. The reason for that difference boils down to the algorithm itself combined with calibration differences between these two aging instruments.

The current Kd490 algorithm employs a band ratio (well, a maximum band ratio...see the ATBD for details) to relate the remote sensing reflectance signal to the diffuse attenuation at 490nm. That relationship takes on a sigmoidal shape, which means that small differences in the ratio will result in relatively larger changes at the extreme ends of the Kd490 values. So, for high Kd490 (the region you identified has a high retrieved Kd490 value) subtle differences in the ratio between the two instruments can result in big differences in the retrieval.

I hope this helps,
Sean

[tomtomascik]

Hi Sean,
Thank you very much for taking time to explain the data, much appreciated. I will be using the VIIRS data set. My main objective is to use the Kd490 data for that small 12 quadrant region to compare the 'relative' turbidity among 4 seasons. For those that may be interested, the area in question (St. Martin's Island) represents the northern-most extent of scleractinian coral communities in the northeastern Bay of Bengal that falls into the Myanmar north and Bangladesh Ecoregion ER024. High turbidity is one of the key environmental factors that has prevented coral reef development in the area, even though corals settle and survive on the rocky substrate that surrounds the island.
Sea, I would greatly appreciate it if you would be kind enough to let me know how you produced the VIIRS image of the larger area in your reply. Thanks.
Cheers,
Tom

[OB.DAAC - SeanBailey]

Tom,
I used the tools we make available with the SeaDAS package to generate those images of the L3 mapped Kd490 data, specifically the l3mapgen program.
Regards,
Sean

[tomtomascik]

Hi Sean,
Thank you for your reply. I downloaded the SeaDAS package a while back, but still trying to figure out how to actually use it. I wish there were better tutorials, or a manuals, for those like me with zero knowledge of these systems. Thank you for your help.
Cheers,
Tom

[OB.DAAC - amscott]

Hi @tomtomascik ,

You can find tutorials here; https://seadas.gsfc.nasa.gov/tutorials/video_tutorials/

[OB General Science - guoqingw]

To add on to Sean and Alicia's replies. Here is the help page for data processors including "l3mapgen": https://seadas.gsfc.nasa.gov/help-8.3.0/processors/ProcessOverview.html

[tomtomascik]

Thank you!
Tom