Scholarlink

<p>The study evaluates the effectiveness of assimilating satellite radiance and Pre-Processed Binary Universal Form for the Representation (PREPBUFR) datasets, including regional background error statistics, to improve cyclone forecasts over the Bay of Bengal using the Weather Research and Forecasting (WRF) model and its 3-Dimensional Variational (3DVAR) data assimilation system. A doubly nested domain with a finer resolution of about 4&#xa0;km under convective permitting scale was employed to simulate the four intense cyclones. Initial and boundary conditions were provided using National Centers for Environmental Prediction Final Analysis (NCEP-FNL) data. Twelve numerical experiments were conducted for four intense tropical cyclones: Phailin (2013), Hudhud (2014), Vardah (2016), and Mocha (2023), using control (CNTL, without data assimilation) and 3DVAR (data assimilation) experiments, with PREPBUFR alone and a combination of PREPBUFR and satellite radiance data (DA). The assimilation results suggest that initial conditions improved through data assimilation. Hence, it can improve the cyclone predictions regarding track, intensity, landfall location, and timing. The study found that the radiance data significantly influenced moisture and temperature profiles, especially for Phailin and Vardah. Model simulations demonstrated that data assimilation experiments improved cyclone track predictions by reducing errors by 13.95%, with mean track errors of 55&#xa0;km, 98&#xa0;km, 95&#xa0;km, and 175&#xa0;km from day 1 to day 4, respectively and absolute errors in maximum surface wind speed (MSW) reduced by up to 22.3% compared to the PREPBUFR experiment on Day-3 of the forecast and 30%, compared to CNTL experiment. The DA experiment simulations outperformed the PREPBUFR-only and CNTL experiments in the prediction of central sea level pressure (CSLP), achieving 54.5% reduction in CSLP error on Day-3. The DA experiment predicted the peak intensity more accurately for the majority of tropical cyclones (TCs) compared to the PREPBUFR and CNTL experiments. In addition, the DA experiment simulations provided more accurate predictions of precipitation, reflectivity, and warm core structure than the CNTL experiment simulations, aligning well with IMD best-fit track data. 24&#xa0;h of simulated accumulated rainfall from Cyclone Phailin were also compared with Global Precipitation Measurement (GPM) rainfall and 174 Automatic Weather Stations (AWS) to evaluate the model performance of the DA experiment and it is well compared with observational data. Incorporation of NCEP PREPBUFR of meteorological data, satellite radiation observations and regional background error statistics (RBES) through the 3DVAR system significantly improved peak intensity and cyclone prediction over the Bay of Bengal (BoB) region.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Evaluating the impact of satellite radiance observations on the simulation of intense tropical cyclones over the Bay of Bengal on a convective permitting scale using a data assimilation technique

  • 期刊论文

  • 发表日期：

    2025年7月19日

查看全文
  • Thatiparthi K.,
  • Kuvar Satya Singh

摘要

The study evaluates the effectiveness of assimilating satellite radiance and Pre-Processed Binary Universal Form for the Representation (PREPBUFR) datasets, including regional background error statistics, to improve cyclone forecasts over the Bay of Bengal using the Weather Research and Forecasting (WRF) model and its 3-Dimensional Variational (3DVAR) data assimilation system. A doubly nested domain with a finer resolution of about 4 km under convective permitting scale was employed to simulate the four intense cyclones. Initial and boundary conditions were provided using National Centers for Environmental Prediction Final Analysis (NCEP-FNL) data. Twelve numerical experiments were conducted for four intense tropical cyclones: Phailin (2013), Hudhud (2014), Vardah (2016), and Mocha (2023), using control (CNTL, without data assimilation) and 3DVAR (data assimilation) experiments, with PREPBUFR alone and a combination of PREPBUFR and satellite radiance data (DA). The assimilation results suggest that initial conditions improved through data assimilation. Hence, it can improve the cyclone predictions regarding track, intensity, landfall location, and timing. The study found that the radiance data significantly influenced moisture and temperature profiles, especially for Phailin and Vardah. Model simulations demonstrated that data assimilation experiments improved cyclone track predictions by reducing errors by 13.95%, with mean track errors of 55 km, 98 km, 95 km, and 175 km from day 1 to day 4, respectively and absolute errors in maximum surface wind speed (MSW) reduced by up to 22.3% compared to the PREPBUFR experiment on Day-3 of the forecast and 30%, compared to CNTL experiment. The DA experiment simulations outperformed the PREPBUFR-only and CNTL experiments in the prediction of central sea level pressure (CSLP), achieving 54.5% reduction in CSLP error on Day-3. The DA experiment predicted the peak intensity more accurately for the majority of tropical cyclones (TCs) compared to the PREPBUFR and CNTL experiments. In addition, the DA experiment simulations provided more accurate predictions of precipitation, reflectivity, and warm core structure than the CNTL experiment simulations, aligning well with IMD best-fit track data. 24 h of simulated accumulated rainfall from Cyclone Phailin were also compared with Global Precipitation Measurement (GPM) rainfall and 174 Automatic Weather Stations (AWS) to evaluate the model performance of the DA experiment and it is well compared with observational data. Incorporation of NCEP PREPBUFR of meteorological data, satellite radiation observations and regional background error statistics (RBES) through the 3DVAR system significantly improved peak intensity and cyclone prediction over the Bay of Bengal (BoB) region.