A SYNTHETIC PRECIPITATION MODEL TO DETERMINE THE USAGE OF THE DISTRIBUTION OF RAIN AS A DATA INPUT IN CALCULATING THE DESIGN FLOOD

Abstract

In calculating the design-flood discharge, engineers often use the frequency-distribution analysis of rainfall data as the basis for obtaining the magnitude of the design flood. The distribution of rain stations is an important factor in determining the distribution of regional rainfall; however, not all catchments have sufficient rain stations to represent the distribution of precipitation in a watershed. This results in a flood hydrograph—obtained using a rainfall–runoff model that is based on the calculation of the estimated rainfall—with a low level of correlation with the observed hydrograph. This research aims to set up a synthetic rain station that corrects the difference between the simulated hydrograph and observed hydrograph to achieve a better correlation level. Usually, researchers use the available rain stations for analyzing the design rainfall and design flood; however, in this research, a synthetic rain station has been built. Rainfall–runoff modeling methods for ungauged catchments (UCs), i.e., watersheds without observation stations, are widely available. Each method has different parameters to determine the unit hydrograph in a watershed. During the development of rainfall–runoff models at UCs, the precipitation in the upstream of a watershed is often assumed to be uniform, but it is not the case. The methodology used in this study collects hourly rainfall data from the ARR and hourly flood-discharge data from the AWLR in one catchment, then calculates the Thiessen coefficient before and after the set up of the synthetic rain station, and finally, calculates the rainfall at the synthetic rain station by trial and error. From the results of the synthetic rainfall modeling, the equation for the synthetic rainfall is obtained:ArRR=(-0.127A-0.260FB-0.307Re+1.227Q+0.51)/(C.K_Thiessen ) with a correlation coefficient of R = 0.818 in a watershed area below 300 km2.