Application of TRMM for runoff predictions in Ungauged basin, Case study: Nanjung Catchment

Abstract

Rainfall-runoff models rely on reliable observation data in order to calibrate and verify the model parameters. Calibration and verification process are necessary because of the uncertainty of the watershed and mostly the model parameters could not derive directly from the physical characteristics. However, the vast majority of the river basin in the world, especially in the developing country or even least developed country, lack of observation data. TRMM (Tropical Rainfall Measuring Mission) was a collaboration mission between NASA and Japan Aerospace to collect and study the rainfall. This paper studied the reliability of TRMM data for simulate runoff during the flood event. The study took place in Nanjung catchment, West Java. The analysis is using HEC-GeoHMS to combine the terrain data and NRCS-SCS model is assigned to calculate the water loss. The results show TRMM data give unsatisfied predictions; the NS value is -5.212. Regression formula is arranged to calibrate the TRMM data and using GS data as a comparison. There are two scenarios that have been analyzed, first scenario by using existing monthly basis regression equation (Y=2.747X0.351) and the second scenario by governing regression formula based on the magnitude of rainfall. The results show both scenarios improves the objective function values, but none of the scenarios give satisfied result when it applied to the hydrology model. Although the TRMM failed for flood simulation, the TRMM data can be still be used for design flood, such as frequency analysis or GEV.