Improvement in WRF model prediction for heavy rain events over North Sumatra region using satellite data assimilation. Issue 1 (November 2021)

• Record Type: Journal Article
• Authors: Saragih, Immanuel Jhonson Arizona; Mukhsinin, Huda Abshor; Tarigan, Kerista; Sinambela, Marzuki; Situmorang, Marhaposan; Sembiring, Kurnia; Humaidi, Syahrul
• Is Part Of: IOP conference series. Volume 893:Issue 1(2021)
• Publication Date: 2021-11
• Subjects: WRF model -- satellite assimilation -- heavy rain; Earth sciences -- Periodicals; Environmental sciences -- Congresses; Environmental sciences -- Periodicals; 550.5
• DOI: 10.1088/1755-1315/893/1/012040
• Languages: English
• ISSNs: 1755-1307

Abstract

Abstract: Located adjacent to the Indian Ocean and the Malacca Strait as a source of water vapour, and traversed by the Barisan Mountains which raise the air orographically causing high diurnal convective activity over the North Sumatra region. The convective system that was formed can cause heavy rainfall over a large area. Weather Research and Forecasting (WRF) was a numerical weather model used to make objective weather forecasts. To improve the weather forecasts accuracy, especially for predict heavy rain events, needed to improve the output of the WRF model by the assimilation technique to correct the initial data. This research was conducted to compare the output of the WRF model with- and without assimilation on 17 June 2020 and 14 September 2020. Assimilation was carried out using the 3D-Var technique and warm starts mode on three assimilation schemes, i.e. DA-AMSU which used AMSU-A satellite data, DA-MHS which used MHS satellite data, and DA-BOTH which used both AMSU-A and MHS satellite data. Model output verification was carried out using the observational data (AWS, AAWS, and ARG) and GPM-IMERG data. The results showed that the satellite data assimilation corrects the WRF model initial data, so as increasing the accuracy of rainfall predictions. The DA-BOTH scheme provided the best improvement with a final weighted performance score of 0.64.