North American Multi Model Ensemble (NMME) Performance of Monthly Precipitation Forecast over South Sulawesi, Indonesia

Abstract

The North American Multi-Model Ensemble (NMME) as one of the multi-model seasonal forecasting system, regularly generate monthly precipitation forecast for all globe with 0.5 – 11.5 months lead time. This useful information can be used as general input to regional and local precipitation forecast. This study quantifies monthly precipitation hindcast data performance in South Sulawesi provided by seven coupled models from the NMME during 29 years (1982 – 2010) period. Climate Hazards Group InfraRed Precipitation with Stations (CHIRPS) dataset and the Standardized Verification System (SVS) for long-range forecasts (LRF) analysis applied to asses monthly precipitation prediction. Almost all individual model shows relatively high skill when used to make June – November monthly precipitation prediction and low skill when used to forecast monthly precipitation in December – May periods. Multi-Model Ensemble (MME) performed using Simple Composite Method (SCM) increased performance of monthly precipitation prediction. Nevertheless, this improvement only applies at short lead time (< 3.3 months). This result also shows that NMME is more promising when it used to make precipitation forecast application during dry period (i.e. drought prediction) rather than wet period over South Sulawesi region.