Sensitivity of WRF-EMS Model to Predict Rainfall Event on Wet and Dry Seasons Over West Sumatra

Abstract

Sensitivity of WRF-EMS model using combination of convective and planetary boundary layer was analyzed using eight experiments that were conducted during wet and dry seasons over West Sumatra. Parameter used were forecast error, root mean square error (RMSE), threat score (TS), probability of detection (POD), false alarm ratio (FAR) for 24 and 48hours rainfall accumulation. WRF-EMS Model with combination of BMJ-ACM2 is the best interm schemes of the consistency of pattern and intensity of rainfall compared with TRMM for 24 and 48hours accumulation. Model verifications on Tabing station revealed that during wet season the values of RMSE were 0.64 and 0.38mm and TS values were 0.86 for 24hours and 0.73 for 48hours. Meanwhile, during dry season RMSE were 0.27 and 0.45 with TS values were 0.50 for both 24 and 48hours. POD and FAR values were 1.0 and 0.14 (24hours), 0.89 and 0.20 (48hours) during wet season. In addition, POD and FAR values were 0.67 and 0.34 (24hours), 0.75 and 0.40 (48hours) during dry seasons. Thus, we found that WRF-EMS model is more accurate to predict rainfall during the wet season than dry season over Tabing, West Sumatra.