UTILIZATION OF BIAS CORRECTION GLOBAL PRECIPITATION MEASUREMENT (GPM) IMERG SATELLITE DATA RAINFALL FOR SURFACE RUNOFF MODEL (CASE STUDY: BRANTAS WATERSHED)

• Main Author: Karinta Hapsari, Rana
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/57593
• Institution: Institut Teknologi Bandung

The Global Precipitation Measurement (GPM) mission provides next-generation global observations of rain and snow. The Integrated Multi-Satelite Retrievals for GPM (IMERG) algorithm combines information from the GPM satellite constellation to estimate precipitation over most of the Earth’s surface. However, it is necessary to evaluate the relevance and accuracy of GPM IMERG products to determine the compatibility with the value of the rain gauge. In this study presents a correction factor for GPM IMERG over Brantas Watershed. The correction of the GPM IMERG product from January 2015 to December 2019 rainfall data which was conducted temporally and spatially by combining Quantile Mapping (QM) method and the Multiple Linear Regression (MLR) method. The combination of QM and MLR methods aims to obtain a correction factor for 132 grids in the Brantas Watershed. The QM was used for bias correction by using five rain probability classes. It shows a significant decrease in the bias value of the monthly rainfall GPM IMERG product. The rBias decreased from 19.3% to 3.2%, while the initial NSE value increased from 0.80 to 0.84. The MLR method aims to obtain the equations used for the correction factor. Equations in MLR include longitude (X), latitude (Y), and elevation (Z). The collaboration of these three parameters produces a different correction factor for each grid. Factor correction for three different catchment area shows decrease rBias 29,3% in Kali Bodo catchment area. For surface runoff model using the corrected GPM rainfall data shows NSE increase to 0,64 in Kali Bodo catchment area.