Abstract This study undertakes a comprehensive assessment of dynamical downscaling of summer monsoon (June–September; JJAS) rainfall over heterogeneous regions namely the Western Ghats (WG), Central India (CI) and North-Eastern Region (NER) for long term mean, excess and deficit episodes for the historical period from 1951 to 2005. This downscaling assessment is based on six Coordinated Regional Climate Downscaling Experiments (CORDEX) for South Asia (SAS) region, their five driving Global Climate Models (GCM) simulations along with observations from India Meteorological Department (IMD) and Asian Precipitation Highly Resolved Observational Integrated Towards Evaluation for Water Resources (APHRODITE). The analysis reveals an overall reduction of dry bias in rainfall across the regions of Indian sub-continent in most of the downscaled CORDEX-SAS models and in their ensemble mean as compared to that of driving GCMs. The interannual variabilities during historical period are reasonably captured by the ensemble means of CORDEX-SAS simulations with an underestimation of 0.43%, 38% and 52% for the WG, CI and NER, respectively. Upon careful examination of the CORDEX-SAS models and their driving GCMs revealed considerable improvement in the regionally downscaled rainfall. The value addition of dynamical downscaling is apparent over the WG in Regional Climate Model (RCM) simulations with an improvement of more than 30% for the long term mean, excess and deficit episodes from their driving GCMs. In the case of NER, the improvement in the downscaled rainfall product is more than 10% for all the episodes. However, the value addition in the CORDEX-SAS simulations for CI region, dominantly influenced by synoptic scale processes, is not clear. Nevertheless, the reduction of dry bias in the complex topographical regions is remarkable. The relative performance of dynamical downscaling of rainfall over complex topography in response to local forcing and orographic lifting depict the value addition (30% over WG and 10% over NER, with a statistical significance of more than 5% level), when compared with the synoptic scale system induced rainfall over the plains of central-India.

中文翻译：

---

对长期平均、过剩和赤字印度夏季风的区域降尺度模拟的评估

分析表明，与驱动 GCM 相比，在大多数缩减的 CORDEX-SAS 模型及其整体平均值中，印度次大陆地区降雨的干偏差总体减少。历史时期的年际变化被 CORDEX-SAS 模拟的集合方法合理捕获，WG、CI 和 NER 的低估分别为 0.43%、38% 和 52%。仔细检查 CORDEX-SAS 模型及其驱动 GCM 后发现，区域降尺度降雨量有相当大的改善。在区域气候模型 (RCM) 模拟中，动态降尺度的附加值在 WG 上很明显，其驱动 GCM 的长期平均、过剩和赤字事件改善了 30% 以上。在 NER 的情况下，对于所有事件，降尺度降雨产品的改进超过 10%。然而，主要受天气尺度过程影响的 CI 区域 CORDEX-SAS 模拟中的附加值尚不清楚。然而，复杂地形区域的干偏压的减少是显着的。响应局部强迫和地形抬升的复杂地形上降雨动态降尺度的相对表现描述了增值（WG 30% 和NER 10%，统计显着性超过5% 水平），与天气尺度系统在印度中部平原引起降雨。然而，复杂地形区域的干偏压的减少是显着的。响应局部强迫和地形抬升的复杂地形上降雨动态降尺度的相对表现描述了增值（WG 30% 和NER 10%，统计显着性超过5% 水平），与天气尺度系统在印度中部平原引起降雨。然而，复杂地形区域的干偏压的减少是显着的。响应局部强迫和地形抬升的复杂地形上降雨动态降尺度的相对表现描述了增值（WG 30% 和 NER 10%，统计显着性超过 5% 水平），与天气尺度系统在印度中部平原引起降雨。