Abstract

Understanding how precipitation fluctuates geographically and temporally over a specific place due to climate change is critical. Generally, simulations of general circulation models (GCM) under different scenarios are downscaled to the local scale to study the impact of climate change on precipitation. However, selecting suitable GCMs for the given study area is one of the most hectic tasks, as the performance of GCMs may vary with respect to space and timescale. Therefore, the current study ranks twenty-seven CMIP 6 (Coupled Modelled Intercomparison Project Phase 6) GCMs in simulating precipitation over India for nine times series, including daily, monthly, yearly, and six extreme series extracted with annual maximum and peak over threshold methods. The gridded daily rainfall data provided by the India Meteorological Department (IMD) are used as the observed data. The GCMs' outputs are corrected for the systematic bias using the linear scaling method. The performance of a GCM is assessed with three statistical performance metrics, namely NSE, RMSE, and R². The GCMs' ranks are determined using a multi-criterion decision-making technique named the modified technique of order preference by similarity to an ideal solution (mTOPSIS) for every grid point and nine timescales (i.e., daily, monthly, yearly, and six extreme series). From the results, for the entire India, the top ten recommended CMIP 6 GCMs are FGOALS-g3, HadGEM3-GC31-MM, EC-Earth3, BCC-CSM2-MR, CNRM-CM6-1-HR, CanESM5, AWI-ESM-1-1-LR, MPI-ESM-1-2-HR, IITM-ESM, and INM-CM5-0. The identified best-performing models provide insightful information for better regional climate projections and underscore the necessity of considering multiple model outputs for reliable climate change impact assessments and adaptation strategies in the region.

中文翻译：

---

CMIP 6 气候模型模拟印度降水的排名

摘要

了解气候变化导致特定地点降水量如何在地理和时间上波动至关重要。通常，不同情景下的大气环流模型（GCM）模拟都会缩小到局地尺度，以研究气候变化对降水的影响。然而，为给定的研究区域选择合适的 GCM 是最繁忙的任务之一，因为 GCM 的性能可能会随空间和时间尺度的变化而变化。因此，本研究对模拟印度降水量的 27 个 CMIP 6（耦合模型比对项目第 6 阶段）GCM 进行了九个时间序列的排名，包括日、月、年和用年最大值和峰值超过阈值方法提取的六个极端序列。观测数据采用印度气象局（IMD）提供的网格化日降雨量数据。使用线性缩放方法对 GCM 的输出进行系统偏差校正。GCM 的性能通过三个统计性能指标进行评估，即 NSE、RMSE 和R ²。GCM 的排名是使用多准则决策技术确定的，该技术称为顺序偏好修改技术，通过与每个网格点和九个时间尺度（即每日、每月、每年和六个极端）的理想解决方案 (mTOPSIS) 的相似性来确定。系列）。从结果来看，对于整个印度来说，推荐的前十名 CMIP 6 GCM 分别是 FGOALS-g3、HadGEM3-GC31-MM、EC-Earth3、BCC-CSM2-MR、CNRM-CM6-1-HR、CanESM5、AWI-ESM -1-1-LR、MPI-ESM-1-2-HR、IITM-ESM 和 INM-CM5-0。确定的最佳表现模型为更好的区域气候预测提供了富有洞察力的信息，并强调了考虑多个模型输出以实现可靠的气候变化影响评估和该区域适应战略的必要性。