The Utilization Of Weather Radar Data For Weather Early Warning In Bengkulu Province Using K-Means Method

Abstract

The importance of weather information in Indonesia is because weather is a variable that determines climatic conditions, one of which is rainfall. Due to the importance for the public to know about weather developments in Indonesia, BMKG must be precise and accurate in providing weather information by utilizing Weather Radar. This radar applied the K-Means Method.The Weather Early Warning application uses the Python programming language and the Weather Radar Library (WradLib) by utilizing one of the data mining methods, namely K-Means. The grouping results are based on file.vol data obtained from Weather Radar (data attached). From this file.vol will be processed in the Python program and will produce 2 groups, namely Cluster C1 Wide Area and Cluster C2 Medium - dense area.The results of this study indicate that the use of weather radar for weatherearlywarning in Bengkulu Province using the K-Means method produces several iterations. There is a slight difference between Iteration 1 and Iteration II, because the data file.vol from the weather radar contains thousands of data so that in 1 sub-district there can be 2 areas, namely Wide Areas and Medium - Dense Regions, but that does not affect the final result of this study, namely Warning Early Weather.