Pemanfaatan Model Weather Research and Forecasting Environmental Modeling System (WRF-EMS) untuk Memprediksi Hujan

Abstract

Numerical models can be used to predict the weather. Numerical model used in this study is a model of Weather Research and Forecasting Environmental Modeling System (WRF-EMS). The purpose of this study is to examine the output prediction of rain on 1 or two days ahead using the WRF-EMS model at 6 weather observations stations and the TRMM Data on 8 initials and boundary conditions, and to examine the output prediction of rain spatially using the WRFEMS model on the initial and boundary conditions that are different. In this study, the configuration of the standard scheme using the WRF-EMS model is maintained; however, the cumulus scheme is changed to the Betts-Miller-Janjic cumulus scheme. The accuracy of rainfall prediction results were assessed by comparing the model results with the rainfall synoptic data and TRMM. The accuracy of the model can also be seen in the range of the threat values score (TS) of rain predicted on the 1 and 2 days later. The result shows that the model predictions of rain using WRF-EMS is good enough to simulate a no-rain events on the island of Sumatra and Kalimantan. This is shown by the results of the rainfall prediction intensity and the data shown on TRMM. The WRF-EMS model can predicts rain in the morning, while the result of the prediction is over estimated in the afternoon and evening. The model can produce a fairly well prediction in the territorial waters of south Sumatra and Java Island. The WRFEMS model has a consistent high TS value at Citeko station and horizontal resolution of 27 and 3 km for the prediction of rain on the next couple of days. The TS value at Citeko weather observation station of the model with a horizontal resolution of 27 km for the prediction of 1 and 2 days in advance of each ranged from 0 – 0.33, and 0 – 0.30, while the TS value of the model with a horizontal resolution of 3 km for rainfall predictions 1 and 2 days in advance each ranged from 0.17 to 0.75 and 0.38 - 0.63. The increased of horizontal resolution from 27 km to 3 km can also increase the value of TS. This indicates that the model with a horizontal resolution of 3 km is more accurate for predicting rain compared to the model horizontal resolution of 27 km. The rainfall prediction was spatially over estimated when using the WRF-EMS Model, and this model is more sensitive in predicting rain at Citeko Station.