Weekly sea surface temperature from SNPP-VIIRS data using open source software

Abstract

The objective of this study is to make weekly mean sea surface temperature (SST) from SNPP-VIIRS satellite data. The data used were SST data every afternoon pass, from 13 to 19 May 2019. The data were obtained from Remote Sensing Technology and Data Center, LAPAN. From 12 samples data used showed that pixel size of each data are different, with range between 0.006737548 - 0.006858158 degree. The pixel size should be exactly the same to do data overlay. Then, the pixel size was resized become 0.00675 degree. This process was carried out using Reprojection Menu in QGIS software. The next process was calculating weekly mean SST every pixel. The mean SST was calculated only on pixels that have SST values. The calculation was not carried out on cloud, cloud shadow, land, and out of coverage data. The tolerance of coordinate position between pixels are less than 0.5 of pixel size. Weekly mean SST calculation process was executed using mean function in R software. The result of mean SST showed some possible error SST values. The errors were found in these conditions: 1) in coastal areas where possible mixed pixels of land and water 2) near cloud, where possible mixed pixel of cloud, cloud shadow, and water, 3) pixels with (extremely) high SST values near clouds. To overcome these conditions, some median type filters were applied. Therefore, weekly mean SST from SNPP-VIIRS data that smooth and minimum cloud cover can be obtained.