Dynamical downscaling of a multimodel ensemble prediction system: Application to tropical cyclones

Abstract

This study attempts dynamical downscaling to improve north Indian ocean (NIO) tropical cyclone prediction from a global multimodel ensemble prediction system using weather research and forecasting (WRF) model. A total of 16 ensembles are used in the WRF simulations, these ensembles are biascorrected prior to downscaling for model climatological errors. The ensemble mean constructed from the output of all downscaled ensembles is analyzed for added value to global predictions. This mean is also compared against observation as well as high-resolution (12 km) deterministic forecast from global forecast system (GFS). Two devastating NIO tropical cyclone cases of year 2017 which were not reliably predicted by global systems have been selected for this study. The results show that downscaled predictions well simulate the intensity and spatial distribution of the rainfall and relative vorticity associated with these cyclonic storms. The wind and temperature vertical profiles during cyclone mature stage are also captured more accurately than raw prediction and high-resolution global deterministic forecast. The study affirms the adequacy of dynamical downscaling in predicting the cyclonic storms over global real-time weather forecasting system.

Multi-Model Multi-Physics Ensemble: A Futuristic Way to Extended Range Prediction System

“…These atmospheric ICs are obtained from National Center for Medium Range Weather Forecasting and oceanic ICs from Indian National Center for Ocean Information Services for CFSv2. Additionally, the real-time sea-surface temperature (SST) from CFSv2 after bias-correction is used as forcing to GFS (detailed technique can be seen in Abhilash et al, 2014a, and Kaur et al, 2020. This ERP system was developed and thoroughly tested for skill at Indian Institute of Tropical Meteorology (IITM) under NMM.…”

Tropical cyclone simulations over Bangladesh at convection permitting 4.4 km & 1.5 km resolution

“…Downscaling of ERA5 is reported in a few other studies: Bonanno et al 3 downscale ERA5 using the Weather Research and Forecasting (WRF) model to produce a new 7 km reanalysis over Italy; preliminary work by Taddei et al 4 use ERA5 to force the BOlogna Limited Area Model-MOdello LOCale (BOLAM-MOLOCH) regional model for the purposes of coastal risk assessment in the North Western Mediterranean sea, and Wang et al 5 use ERA5 to run a 10 km WRF domain over high mountain Asia. Specifically examining tropical cyclones, many studies use variations of the Weather Research and Forecasting (WRF) Model 6 , such as Kaur et al 7 who use WRF to downscale the National Center for Environment Prediction (NCEP) Climate Forecast System (CFSv2) and its atmospheric component Global Forecast System (GFS) to 9 km over the north Indian Ocean for two historical cases (Mora and Ockhi), with analysis focusing on the spatial accuracy of rainfall and 850 hPa vorticity, and the vertical profiles of wind and temperature. They conclude that the downscaled model significantly improves the spatial distribution of rainfall, maximum vorticity evolution, wind, and temperature profiles for mature phase cyclones.…”

Deep learning–based downscaling of summer monsoon rainfall data over Indian region

“…Several methods are available for downscaling meteorological information. They can be broadly classified as dynamical and statistical downscaling focusing on regional climate simulations[ (Kaur et al 2020)(Nobre et al 2001Díez et al 2005;Shukla and Lettenmaier 2013;Xue et al 2014)]. The statistical downscaling methods [(von Storch et al 1993;Zorita and von Storch 1999;Wilby and Dawson 2013;Sahai et al 2017)] have been applied in several studies mainly focussing on the downscaling of precipitation [(von Storch et al 1993;Vrac and Naveau 2007;Benestad 2010;Sahai et al 2017)].…”

“…Dynamical downscaling involves using the initial and boundary conditions from a global model and then running a high-resolution regional model to generate a local forecast. For example, city-level forecasts, which are often computationally expensive and time-consuming, require high-performance comput resources [ (Benestad and Haugen 2007;Benestad 2010;Kaur et al 2020)]. On the other hand, statistical downscaling uses the outputs from the global dynamical models or observations as inputs to statistical models (ranging from simple univariate to complex multivariate schemes) to generate high resolution (e.g., city-scale) information.…”