Rainfall Estimation Using Machine Learning Approaches with Raingauge, Radar, and Satellite Data

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Citations (Scopus)

Abstract

Rainfall is an essential variable in studying weather and has a vital role in the hydrological cycle. Therefore, accurate rainfall information becomes an urgent need. Machine learning approaches have also shown remarkable developments in various areas of life including rainfall estimation. This research combined rain gauge, radar and satellite data with several algorithms such as Decision Tree, Random Forest, Adaptive Boosting, and Gradient Boosting. Interestingly, it was found that the Gradient Boosting learning ensemble algorithm is best applied in rainfall estimation, with a correlation value (R2) is 0.740 and RMSE value is 1,765. On the other hand, weather radar reflectivity data significantly influences the integration of the data built into the model, suggesting that details of optimizing data are significantly required.

Original languageEnglish
Title of host publication2022 International Conference on Electrical Engineering and Informatics, ICELTICs 2022
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages25-30
Number of pages6
ISBN (Electronic)9781665487665
DOIs
Publication statusPublished - 2022
Event2022 International Conference on Electrical Engineering and Informatics, ICELTICs 2022 - Banda Aceh - Jakarta, Indonesia
Duration: 27 Sept 202228 Sept 2022

Publication series

NameProceedings of the International Conference on Electrical Engineering and Informatics
Volume2022-September
ISSN (Print)2155-6830

Conference

Conference2022 International Conference on Electrical Engineering and Informatics, ICELTICs 2022
Country/TerritoryIndonesia
CityBanda Aceh - Jakarta
Period27/09/2228/09/22

Keywords

  • estimation
  • gradient boosting
  • machine learning
  • rainfall

Fingerprint

Dive into the research topics of 'Rainfall Estimation Using Machine Learning Approaches with Raingauge, Radar, and Satellite Data'. Together they form a unique fingerprint.

Cite this