Comparison of interpolation methods for sea surface temperature data

Denny Wijaya Kusuma (Institute for Marine Research and Fisheries Ministry Of Marine Affair and Fisheries)
Ari Murdimanto
Bambang Sukresno
Dinarika Jatisworo

JFMR (Journal of Fisheries and Marine Research)
Vol 2, No 2 (2018): JFMR VOL 2 NO 2

Interpolation methods have been used in many applications to produce continuous surface data based on point data. The common interpolation methods for Sea Surface Temperature (SST) data are Inverse Distance Weighted (IDW), Kriging, Natural Neighbor Interpolation (NNI), and Spline. In this study, those four interpolation methods will be reviewed and compared to find the satisfactory method. The Argo float data was chosen as SST point data and Aqua MODIS image as validation data. Each method will be reviewed and compared to Aqua MODIS data to find the best performance. The assessment for testing the best interpolation model is smooth performance, Maximum and Minimum comparison, mean comparison, Root Mean Square Error (RMSE) and Standard Deviation Difference. The result shows that IDW interpolation is the best way to make spatial interpolation for SST.