Scholarlink

<p>This study determines the best combination of short-wave (SW), long-wave (LW) radiation schemes available in the Weather Research, Forecasting Model (WRF) to characterize cloudbursts in the North West Himalayan (NWH) region by inspecting the rainfall location, time, intensity against the Integrated Multi-Satellite Retrievals for GPM (Global Precipitation Measurement) (IMERG). We designed nine experiments with a unique pair of SW (Dudhia, Rapid Radiative Transfer Model (RRTM), Rapid Radiative Transfer Model for GCM (RRTMG)), LW (RRTMG, New Goddard, RRTM) radiation schemes, analysed the multiple cloudburst events of 19th–20th August 2022 in parts of Uttarakhand, Himachal Pradesh. We inferred that [RRTMG, RRTM] ([SW, LW]) has an intrinsic potential to resolve cloudbursts. The combination showed high accuracy in location, evident from the False Alarm Ratio (FAR), Probability of Detection (POD) ranging from 0.07 to 0.17, 0.89 to 0.97, respectively, for the low-elevation regions of Sarkhet village, Dhanoulti, Maldevta in Uttarakhand. The time series analysis showed that its peak rainfall led IMERG by only 30&#xa0;min. The statistical tests revealed that it has a small Root Mean Square Error (RMSE) (in the range 4.50–6.00&#xa0;mm), Mean Absolute Error (MAE) (in the range 2.40–3.30&#xa0;mm), bias (in the range −&#xa0;1.30 to 0.28&#xa0;mm), correlates with IMERG rainfall at 5% significant level, particularly in Uttarakhand. The categorical metrics for intensity also revealed a relatively higher proportion of ‘True-Hits’. We substantiated the performance of [RRTMG, RRTM] further by observing a sharp drop in the Outgoing Longwave Radiation (OLR) ~ 9–10&#xa0;h before the cloudburst in Sarkhet village, Dhanoulti in Uttarakhand, which was also prominent in ERA5 reanalysis data. We also inferred that the model has a regional bias, wherein its performance is relatively poor in Himachal Pradesh. We attributed this to the poor representation of local factors in the model.</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

Inspecting the sensitivity of radiation schemes for numerical modelling of cloudburst events in the North West Himalayan region

  • 期刊论文

  • 发表日期：

    2025年6月28日

查看全文
  • Sreyasi Biswas,
  • Charu Singh

摘要

This study determines the best combination of short-wave (SW), long-wave (LW) radiation schemes available in the Weather Research, Forecasting Model (WRF) to characterize cloudbursts in the North West Himalayan (NWH) region by inspecting the rainfall location, time, intensity against the Integrated Multi-Satellite Retrievals for GPM (Global Precipitation Measurement) (IMERG). We designed nine experiments with a unique pair of SW (Dudhia, Rapid Radiative Transfer Model (RRTM), Rapid Radiative Transfer Model for GCM (RRTMG)), LW (RRTMG, New Goddard, RRTM) radiation schemes, analysed the multiple cloudburst events of 19th–20th August 2022 in parts of Uttarakhand, Himachal Pradesh. We inferred that [RRTMG, RRTM] ([SW, LW]) has an intrinsic potential to resolve cloudbursts. The combination showed high accuracy in location, evident from the False Alarm Ratio (FAR), Probability of Detection (POD) ranging from 0.07 to 0.17, 0.89 to 0.97, respectively, for the low-elevation regions of Sarkhet village, Dhanoulti, Maldevta in Uttarakhand. The time series analysis showed that its peak rainfall led IMERG by only 30 min. The statistical tests revealed that it has a small Root Mean Square Error (RMSE) (in the range 4.50–6.00 mm), Mean Absolute Error (MAE) (in the range 2.40–3.30 mm), bias (in the range − 1.30 to 0.28 mm), correlates with IMERG rainfall at 5% significant level, particularly in Uttarakhand. The categorical metrics for intensity also revealed a relatively higher proportion of ‘True-Hits’. We substantiated the performance of [RRTMG, RRTM] further by observing a sharp drop in the Outgoing Longwave Radiation (OLR) ~ 9–10 h before the cloudburst in Sarkhet village, Dhanoulti in Uttarakhand, which was also prominent in ERA5 reanalysis data. We also inferred that the model has a regional bias, wherein its performance is relatively poor in Himachal Pradesh. We attributed this to the poor representation of local factors in the model.