Welcome to the Earthdata Forum! Here, the scientific user community and subject matter experts from NASA Distributed Active Archive Centers (DAACs), and other contributors, discuss research needs, data, and data applications.
by gnwiii » Mon Nov 30, 2020 10:46 am America/New_York
By "time steps" do you mean different level-2 files? Each level-2 file has unique geolocation of the pixels, so comparing different passes requires mapping. Normally, when mapping, you set the SST quality levels you want to use, so for comparisons you could generate maps with different settings for quality levels. Many of the factors that affect SST quality change with time, so for each level-2 pass and choice of quality levels you will get very different images and potentially rather small areas of overlap for a given quality level.
My aim is to estimate the temperature difference for two consecutive years. But, to find the difference we need to take the same pixels for these two years. And in these two years I am not getting the same quality flags (one year has 0 quality flag and the other year has a quality flag 2 for the same pixel). So the difference between this is considerable or not. Can we move further by considering these values?
by gnwiii » Tue Dec 01, 2020 12:25 pm America/New_York
A lot of energy has gone into ways to compare temperatures at different times for climate change investigations. The ocean is very dynamic, so conditions at a given location can often change quickly, as with a storm driven mixing event. Usually this variabilty is handled by some sort of averaging in space and time, so you may want to work with level-3 data.
I think in the earlier message I could not reach you with the exact question. So, herewith I explain clearly with describing my doubt on the SST data set. For my study, I am using MODIS AQUA Level 3 monthly data of SST. Suppose for a specific pixel, the temperature is 25Â° C (Time step 1) with flag 0 (which represents the best quality) and in Time step 2 the pixel value is 26Â° C with quality flag 2 (which is questionable). Can we estimate the temperature deviation between these two time steps for the same pixel with different quality flags?
The vast majority of the conditions that can set the quality level for a pixel to '2' have to do solely with the detection of cloudy pixels. See https://oceancolor.gsfc.nasa.gov/atbd/sst/flag/ for a description of the sst flags and how they are combined to define the quality levels. At level three, only the highest quality data are kept when binning the L2 data. So, if a quality level 1 or 0 pixel falls into a bin that already includes a quality level 2 data, the higher quality data replace the lower quality data. If a monthly product retains quality level 2 data, it is likely that the region is rather cloudy (or determined cloudy by the cloud detection decision tree), so you should use with caution. The quality levels do not provide information on the uncertainty of the retrieved SST. See https://oceancolor.gsfc.nasa.gov/atbd/sst/ (and references therein) for more details on how the product is derived and assessed.
by arunima_1994 » Mon Dec 28, 2020 12:55 pm America/New_York
Thank you so much sir for the explanation. Still I have one doubt, between the quality flag 0 and quality flag data, the difference is only with the change in satellite zenith angle. I want to know that, how the change in the satellite zenith angle affects the SST data. Can we consider the SST data having quality flags 0 and 1 together for one observation or we have to consider it separately?
Larger zenith angles correspond to a longer path length, so more air between the sensor and the ground. The longer the path length, the more chance for water vapor to affect the retrieval. As there is currently no correction for water vapor contamination, the data are simply lowered in quality. The quality levels are: 0 = BEST 1 = GOOD 2 = QUESTIONABLE 3 = BAD
If GOOD is good enough for your purposes, you can include them with BEST.