ESTIMATION AND UNCERTAINTY ANALYSIS OF PROBABLE MAXIMUM PRECIPITATION (PMP) CASE STUDY: WESTERN OF JAVA

• Main Author: Indriany
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/50390
• Institution: Institut Teknologi Bandung

Probable Maximum Precipitation (PMP) is a method to understanding the potential for rain that can trigger a very large flood and is taken into consideration in anticipating this event. PMP is usually entered in the calculation of the Probable Maximum Flood (PMF) used to measure hydraulic structures. One of the commonly used PMP estimation methods is Hershfield statistical method. Regardless of the method used, there is uncertainty in PMP estimation. PMP estimation with considering the uncertainty not much research yet. PMP estimation with uncertainty analysis can provide information to practitioners or stakeholders in making more informed decisions regarding dam failure and flood risk assessments. In this research, estimation and uncertainty analysis of PMP will be carried out in the Western of Java region, precisely in Jakarta, Bogor, Banten and Bandung. The data used to calculate of PMP are the observation rainfall at the BMKG and BBWS Citarum stations. Furthermore, uncertainty analysis is carried out by looking at difference in PMP return period from the selection of best and 4th best probability distribution. Based on the results of calculations in this study, PMP value is inversely proportional to the uncertainty value. A high PMP value tends to have low uncertainty, while a low PMP value tends to have high uncertainty. The highest PMP is in Tangerang Selatan station. Meanwhile, the lowest PMP was at the Geofisika Bandung Station. Then the difference in return period from best distribution and 4th best distribution is smallest at Kemayoran Jakarta station and Citeko Bogor station, while the biggest difference is at Geofisika Bandung station. Large return period differences indicate high uncertainty, so it is important to determine accurately the most suitable probability distribution.