VERIFICATION OF CITARUM RIVER DISCHARGE USING CLIMATE FORECAST SYSTEM (CFS) VERSION 2 (Case Study Upper Citarum Watershed, Bandung, West Java)

• Main Author: PUTU PRADIPTIANTI LESTARI (NIM : 12813012), LUH
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/22928
• Institution: Institut Teknologi Bandung

Citarum River serves as a water supplier for the main reservoir in West Java, namely Saguling Reservoir. Debit prediction is important in terms of reservoir operations. This study describes the use of Climate Forecast System (CFS) Version 2 data that has been done statistical downscaling to predict the discharge in Upper Citarum watershed using hydrological model. The statistical downscaling method used in this research is constructed analogue multiwindows by using chi 850 hPa and psi 850 hPa data with predictors used by AUSMI and WNPMI to predict daily rainfall. The daily rainfall prediction is then used as an input for the SWAT hydrologic model to generate predictive discharge in the Upper Citarum basin. Verification is done by determining the low discharge event <54.8 m3/s which influences the turbine movement especially in the period of December-January-February 2011 - 2016 by using Brier Score and Brier Skill Score. <br /> <br /> <br /> By the obtained verification it is known that CFSv2 is able to represent the event of discharge less than 54.8 m3/s that occurs for 10 - 14 days in one month with good accuracy value of 0.068 and has good skill in predicting the occurrence. In addition, CFSv2 data has a fairly good accuracy in predicting high discharge categories in the period December-January-February. However, CFSv2 data do not have the skill in predicting low, normal, and high discharge categories.