STUDY OF SPATIAL TEMPORAL RAINFALL DISTRIBUTION EFFECT ON FLOODS IN UPPER CITARUM RIVER BASIN

• Main Author: Akbari Habibi, Suyudi
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/42327
• Institution: Institut Teknologi Bandung

Indonesia is a country that has high natural resource potential and high disaster risk. Statistically during the last 15 years the flood disaster has the highest percentage of 33.8% of all disaster events in Indonesia. One area that is often affected by floods is the Bandung area because it is naturally a flood valley formed by the Citarum River. This flood basin is surrounded by mountains that form several sub-watersheds with different characteristics such as the width of the watershed, slope of land, land use. The difference in characteristics results in the distribution of rainfall varies with time and space and is thought to result in the contribution of each watershed to flood events. Therefore, a study of temporal spatial rainfall distribution patterns is needed to determine the dominant rainfall distribution in influencing flooding and inundation. In addition, some factors that influence the flow may be caused by external factors such as the presence of geothermal infrastructure that causes runoff to be reduced. The focus of the research carried out was to determine the characteristics of spatial temporal rainfall distribution patterns of synthetic unit hydrographs and how their correlation with the flood inundation along the Upper Citarum River. The methodology used in the research phase is secondary data collection, literature review, analysis of the contribution of flood hydrographs to the hydrological model, and analysis of flood inundation with hydraulic models as well as conclusions and suggestions for future research development. The data used in this study are rainfall data, duration hours, land use data, topographic data from DEMNAS and AWLR data in the Upper Citarum River. The method that used for the analysis of the contribution of flood hydrographs is compared the lumped model method and the semi distributed model. Comparison of the two methods is to compare the calibration process using Nash-Sutcliffe model efficiency (NSE). Calibration by comparing the discharge data from AWLR observations at Nanjung Station and the modeling discharge data. After that, a return period flood hydrograph analysis is performed and to calibrate it will be compared with the results of statistical analysis of the distribution of the maximum annual flood discharge that has occurred in the Upper Citarum River that accumulates in Nanjung. The best calibration results will be used in determining the contribution of flood hydrographs. After that, a flood inundation analysis is performed using a hydraulic model whose output is a prediction of a flood inundation and a prediction of inundation for a flood return period. Simulations to be carried out in the modeling are 6 flood events, namely 3 years of flood events in 2017, 1 year of flood events in 2018 and 1 year of flood events in 2019. These flood events simulations are very important in comparing different spatial and temporal flood characteristics. The hydrograph contribution of each event will be the basis for determining the average contribution of the event that occurred. For hydrograph analysis that is considered suitable with the Citarum sub-watershed is HSS-Snyder. Hydrograph analysis using the SCS-Curve Number from NCRS obtained a lower calibration value. The most suitable hydrograph for Majalaya Sub-watershed is Synthetic Unit Hydrograph (HSS) SCS where the average value of NSE HSS SCS (NSE = 0.90) is better than the Nakayasu HSS average value (NSE = 0.88). During flood conditions the peak discharge values vary depending on the coefficient values of Ct, n and ? (alpha) of the HSS SCS-Snyder and HSS Nakayasu parameters. The best calibration flood hydrograph is SCS-Snyder flood hydrograph (NSE value = 0.72) while SCS-CN (NSE = 0.65). Based on the lumped model and the semi distributed model, the results show that the semi distributed model provides better results than the lumped model. The average NSE value for the semi-distributed model is 0.88 (very good) and for the lumped model is 0.64 (good). After that, an analysis of the contribution of the flood hydrograph which obtained the results of five rivers that had the greatest influence on the floods in the Upper Citarum Watershed from the floods that occurred was the Upper Citarum River (41.97%),Cikeruh River (13.97%),Cisangkuy River (11.60%), Citarik River (8.59%), and Ciwidey River (7.99%). Five rivers that had the greatest influence on flooding in the Upper Citarum Watershed for 2 year return period rainfall were the Upper Citarum River (24.83%), Cisangkuy River (17.59%), Citarik River (11.77%), Ciwidey River (9.61%), and Cikeruh River (9.33%).