Generating himawari-8 time series data for meteorological application

Abstract

Optical remote sensing images have been widely used for temporal monitoring. The data is acquired by sensors on satellites with better spatial resolution compared to in-situ measurements by meteorological stations. The problem with utilizing optical images is the cloud, which blocks the ground and near-ground information collected by satellites. To overcome this problem, especially when dealing with thermal bands, we propose a procedure including aggregation and spatial interpolation methods to obtain time series data over a region. There is still no reference to selecting the data period to calculate the aggregate value and apply spatial interpolation. An assessment is proposed by applying Yamane’s formula in the time domain and thresholding the number of pixels in the spatial domain. Himawari-8 data was utilized and collected on an hourly basis over Java Island. This algorithm is applied to a sequence of periodic datasets to obtain a time series of aggregate data for meteorological applications. The result of this study is a recommendation to use three-month periods of data over the eastern part of Java.