Question About RRS_443 Variability in MODIS Aqua L2 Data

[cassandra21]

Hello everyone

I am currently using MODIS Aqua L2 data to study the spatial and temporal variability of RRS in a specific region of interest (ROI). While analyzing data from the 17th of May 2024, I noticed a significant difference in RRS_443 values between two scenes taken less than three hours apart: one at 10:50 UTC and another at 12:30 UTC (see files attached).

Using the same threshold and visualization settings for both images, RRS_443 values drop from approximately 0.004 sr-1 at 10:50 to around 0.001 sr-1 at 12:30. I was wondering if such variability is expected within such a short timeframe, and what factors might explain it. As I’m just beginning to work with this dataset, I would greatly appreciate any insights or guidance on interpreting these results correctly.

Thank you in advance for your help!

[PODAAC - celiaoued]

Hello,
It may help to mention the download links you used, or where you got the data from. I also recommend adding the tag for OB.DAAC to your post.

[cassandra21]

Hello,

Thank you for your response!

I used the Aqua MODIS Level-2 Regional Ocean Color (OC) data, version 2022.0, which I retrieved from NASA's Earthdata Search. Below are the direct links to the specific NetCDF files I used:

1. https://obdaac-tea.earthdatacloud.nasa.gov/ob-cumulus-prod-public/AQUA_MODIS.20240517T105001.L2.OC.nc

2. https://obdaac-tea.earthdatacloud.nasa.gov/ob-cumulus-prod-public/AQUA_MODIS.20240517T123001.L2.OC.nc

Thank you

[PODAAC - celiaoued]

Hi, I can't help you with these files, but OB.DAAC probably can!

[jared]

Hi,

The high temporal variability could be an artifact, possibly results from the atmospheric correction. This region from those two MODIS granules has a relative humidity higher than 95%, which is the upper threshold of the aerosol LUTs used for atmospheric correction. Under this situation, the aerosol information maynot be properly interpolated from the LUTs. We are going to find a way to resolve this problem.

Best,
Minwei

[cassandra21]

Hi,

Thank you for the clarification!

In the meantime, I’ve been extracting the information on the flags active within my region of interest to better assess data quality. I was wondering if there is any documentation or guidance on the thresholds or specific flag conditions that define whether a pixel should be considered invalid. For example, should every pixel with any flag active be treated as NaN, or can some flagged pixels still contain valid RRS values?

Any insights on how to interpret these flags in the context of selecting the most reliable data would be very helpful

Best,
Cassandra

[jared]

Hi Cassandra,

You can find the flag information here: https://oceancolor.gsfc.nasa.gov/resources/atbd/ocl2flags/

Best,
Minwei