Page 1 of 1
Question related SST
Posted: Fri Sep 17, 2021 2:09 pm America/New_York
I have downloaded the level 3 daytime and nighttime SST data from MODIS Terra. When I process the data, I saw that some of the years and months show nighttime SST is higher than daytime SST. Is it possible? Or it is the sensor problem.
If possible, please, suggest some references on what situations the ocean shows the abnormality?
In general, daytime SST is higher than nighttime SST as far as I know.
Re: Question related SST
Posted: Fri Sep 17, 2021 5:35 pm America/New_York
There are many reasons for the general SST_day > SST_night expectation to fail, including diel patterns in wind velocity and direction, clouds, and biases in ancillary products.
Years ago, while working on methods to determine dynamic boundaries for Longhurst's regions, I compared day and night SST using Pathfinder daily data for the Northwest Atlantic and found that (SST_day - SST_night) showed a decreasing trend over time with some negative values in the most recent years. I discussed this with the physical oceanographers in our group, but the bottom line was that the day-night differences were small compared to the change in temperature across major current systems (Gulf Stream in the study area). Since we were also looking at chlor_a, there were practical advantages like savings in internet bandwidth and disk space for data where you could get both from the same granual (e.g., daytime).
Re: Question related SST
Posted: Tue Sep 21, 2021 8:56 am America/New_York
It is possible for the night-time SST retrievals to be warmer than those of the previous day, but it could also be the result of errors in the retrievals resulting from algorithm imperfections, especially in anomalous atmospheric conditions.
The effects of diurnal heating usually result in the “daytime” SST retrievals being warmer than the “night-time” values, especially when “daytime” means mid-afternoon. But depending on when and where the measurements were taken and under which circumstances, this may not be the case. Terra is in a daylight descending orbit with an equator crossing time of 10:30 am and pm. This is a mean sun time, and around the year will change by up to 20 minutes, relative to the position of the sun in the sky. Also, for measurements taken away from the equator, the local sun time can vary by several hours, dependent on the latitude. For Terra, the measurements at northern latitudes are taken at a later sun time during the day, and at an earlier sun time for southern latitudes. In the southern hemisphere, the night-time measurements are taken earlier in the evening, and in the northern hemisphere, they are taken later into the night. Thus, the sampling of any diurnal heating and cooling signal will vary depending on the latitude, with the “day” and “night” labels of the measurements not being necessarily appropriate.
It would help us if you could give me some details of what you are seeing, including when and where you have noticed the anomalous signals, and how big they are. Also, knowing the files names and the source of the data would also help us determine whether the signals are environmental or artifacts of the algorithms. Please contact me using my firstname.lastname@example.org