RECONSTRUCTION OF SEA SURFACE TEMPERATURE IN WPPNRI 713 USING DINEOF
Indonesia is a tropical country with intense atmospheric convective activity, this makes the satellite difficult to obtain information and results blank data due to a strong cloud cover. This study aims to reconstruct Sea Surface Temperature (SST) data using the Data Interpolating Empirical Orthogon...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/71145 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | Indonesia is a tropical country with intense atmospheric convective activity, this makes the satellite difficult to obtain information and results blank data due to a strong cloud cover. This study aims to reconstruct Sea Surface Temperature (SST) data using the Data Interpolating Empirical Orthogonal Functions (DINEOF) method in the area of WPPNRI 713. This study used a daily SST data Level 3 MODIS Aqua night time (MID-IR) and daytime (Thermal) from 2003 – 2021, and also SST Level 4 GHRSST data in 2012 and 2020 for verification. DINEOF reconstruction on a 4-year data range produces an expected error of 0,4363°C, this performance can be improved by using an additional 5-year dataset. This addition can reduce the expected error to 0,2646°C, but this improvement is accompanied by + 2 times increase in running time. In the reconstruction with a time span of 1 year, the results of DINEOF show a lot of noise in the spatial plot which indicates that DINEOF will be less stable in the short time span. From the comparison of before and after of DINEOF reconstruction on SST Level 4, an RMSE of <0,01°C and a correlation of >0,9 was obtained and indicate good reconstruction results.
|
|---|