| dc.description.abstract | The skill of Regional Climate Model 4 (RegCM4) in simulating rainfall is evaluated through a comparison of observations, using Asian Precipitation-Highly Resolved Observational Data Integration towards Evaluation (APHRODITE), and Climate Research Unit (CRU), and 19-year simulation period of 1982-2000 in Kalimantan. The analysis includes how RegCM4 capture the rainfall type, the rainfall characteristics when ENSO, the error and the diversity of data, and dominant cycles in Kalimantan. The analysis is continued with a calculation of Maximum Dry Spell Length (MDSL). It is found that RegCM4 produces overestimated value, but it is capable in capturing rainfall type, rainfall characteristics when ENSO, and dominant cycles in Kalimantan. The dominant cycles in influencing Kalimantan rainfall are ENSO and Annual Oscillation. The main model weakness is an overestimation of precipitation in mountainous regions. It is proven by the value of Root Mean Square Error (RMSE) and standard deviation is higher in mountainous regions (>500 m). Furthermore, based on simple correlation, the model is better in simulating monthly resolution than daily resolution. It is evidenced by the correlation up to 0.8 in monthly resolution. In calculating MDSL, RegCM4 reproduces the same pattern (the higher MDSL when El Nino, and lower value when La Nina) with APHRODITE. | en |
