Forecasting of tropical cyclone using global and regional ensemble prediction systems of NCMRWF : A review

Abstract

NCMRWF global ensemble prediction system (NEPS-G) is being used for operational forecasting of tropical cyclones (TC) since the year 2016. The NEPS-G was upgraded and its horizontal resolution was increased from 33 km to 12 km in the year 2018. The upgraded NEPS-G has shown improved performance in forecasting tropical cyclones over the North Indian Ocean (NIO). The average 120 hours track forecast errors of NEPS-G for six TCs of 2019 and four TCs of 2020 are about 250 km and 220 km, respectively. The regional ensemble prediction system of NCMWRF (NEPS-R) has 4 km horizontal resolution. Average track forecast error of NEPS-R at 72 hours forecast lead time is less than that of NEPS-G. The strike probability forecast of NEPS-G shows good reliability and the area under the relative operating characteristic curve is greater than 0.5 for four TCs of 2020. The forecasting of peak TC intensity is better in NEPS-R. It could also predict the rapid intensifications of the TCs ‘Phani’ and ‘Amphan’. NEPS-G was unable to predict rapid intensifications. The prediction of time of peak intensity is still an issue with the models. Both NEPS-G and NEPS-R overestimate peak intensity of TC ‘Nisarga’. The overestimation is more in NEPS-R forecast.