<![CDATA[Impact based forecast is one of the forecasting systems currently being developed by BMKG. The change of paradigm from ordinary weather forecasting systems to impact-based weather forecast is a step to improve the class of BMKG forecasts. The purpose of this study is to determine what weather parameters can be used as a predictor of flood disasters in Jakarta and determine the probability of numerical weather forecast model data for rain that has the potential to flood in Jakarta. This study uses 2015-2019 Jakarta flood events data, GFS models and disaster matrix data. This research begins with the collection of Jakarta flood events in 2015-2019, then performs a composite analysis of weather variables that can be used as a flood disaster predictor, quantifies flood events into impacts (minimal, minor, significant and severe), quantifies weather to likelihood (very low, low, medium and high) and presenting of impact based forecast maps. The results showed that the GFS numerical weather forecast model in Jakarta succeeded in describing the values of rainfall, relative humidity and temperature, CAPE and vertical velocity parameters, but the results were not in accordance with actual reality. The probability of each variable described by the GFS model included as likelihood is as follows, rainfall has a probability of 20-25% which is included as a very low likelihood category (<29%), relative humidity andtemperature has a probability of 50-55% which includes as a low likelihood category (30-59%), CAPE has a probability of 43-45%, which is included as a low likelihood category (30-59%) and vertical velocity has aprobability of 35-38% where it is included as a low likelihood category (30 -59%). In this study the GFS model successfully illustrates the probability and likelihood of relative humidity and temperature, CAPE and vertical velocity better than rainfall in flood events in the DKI Jakarta area.]]>
