Numerical climate modeling and verification of selected areas for heat waves of Pakistan using ensemble prediction system

6th Vacuum and Surface Sciences Conference of Asia and Australia

Amna, S.^1 ; Samreen, N.^1 ; Khalid, B.^1 ; Shamim, A.^1

Department of Environmental Sciences, International Islamic University, Islamabad, Pakistan^1

关键词: Climatic conditions;    Ensemble prediction systems;    General circulation model;    Modeling and verifications;    Probabilistic forecasts;    Probabilistic measures;    Relative operating characteristics;    Statistical temperature;   

Others  :  https://iopscience.iop.org/article/10.1088/1742-6596/439/1/012041/pdf
DOI  :  10.1088/1742-6596/439/1/012041

来源: IOP

【 摘 要 】

Depending upon the topography, there is an extreme variation in the temperature of Pakistan. Heat waves are the Weather-related events, having significant impact on the humans, including all socioeconomic activities and health issues as well which changes according to the climatic conditions of the area. The forecasting climate is of prime importance for being aware of future climatic changes, in order to mitigate them. The study used the Ensemble Prediction System (EPS) for the purpose of modeling seasonal weather hind-cast of three selected areas i.e., Islamabad, Jhelum and Muzaffarabad. This research was purposely carried out in order to suggest the most suitable climate model for Pakistan. Real time and simulated data of five General Circulation Models i.e., ECMWF, ERA-40, MPI, Meteo France and UKMO for selected areas was acquired from Pakistan Meteorological Department. Data incorporated constituted the statistical temperature records of 32 years for the months of June, July and August. This study was based on EPS to calculate probabilistic forecasts produced by single ensembles. Verification was done out to assess the quality of the forecast t by using standard probabilistic measures of Brier Score, Brier Skill Score, Cross Validation and Relative Operating Characteristic curve. The results showed ECMWF the most suitable model for Islamabad and Jhelum; and Meteo France for Muzaffarabad. Other models have significant results by omitting particular initial conditions.