DATA OPTIMIZATION OF RAIN RADAR AND ITS COVERAGE EXPANSION USING RADAR NETWORK

Abstract

This research developed network of two Santanu rain radars of the Center for Atmospheric Science and Technology LAPAN installed in Bandung (SANTANU 1) and Sumedang (SANTANU 2). The objective is to get larger coverage area which also comply data each other. The first step is data optimation to each radar detection results, particularly due to rain attenuation, and then validation with the transportable radar image. The shape analysis on a rain object detected in the slice of both radar images indicates 0.8672 of R² which show the width of both objectcs has a good similarity. Data optimation due to rain attenuation using attenuation correction for regional weather radar (KA) and path-integrated attenuation (PIA) indicate that the reflectivity improvement by KA is 10 times better than PIA. The object detected by Santanu rain radar is similar to the transportable radar for reflectivity larger than 30 dBZ.  The next step is integration of both rain radar images to create mosaic image and then validation using the transportable radar data. The production of the mosaic image is tried out using three methods: image stitching, averaging, and Cressman scheme. The best result is presented by Cressman scheme which show more integrated slice area and maintain the features of the object. However, the processing time of this scheme is 17 minutes which is longer than  image stitching method. The processing time depend on the number of grid data and the hardware memory. Validation on the mosaic result by using transportable radar data by implementing structural similarity index (SSIM) method yields 0,32 of average value. The improvement on SSIM value is obtained if speckle removal is applied by recording 59% of enhancement