基于CMIP5情景模拟的ETCCDI极端温度指数数据集

作者运用来自CMIP5（Coupled Model Intercomparison Project Phase 5）官方数据门户网站（https://pcmdi.llnl.gov/projects/cmip5/）的19个全球气候模式（GCM）输出的日最高气温（tasmax）和日最低气温（tasmin）数据，应用Yang和Xu（Atmospheric Science Letters, 2017, DOI: 10.1002/asl.724.）提出的半参数分位数回归方法（Quantile Regression Method，QRM），针对4种温室气体排放情景（历史、RCP2.6、RCP4.5和RCP8.5），重新计算了6个ETCCDI（Expert Team on Climate Change Detection and Indices， http://wcrp-climate.org/etccdi）百分位极端温度指数（TN10p、TN90p、TX10p、TX90p、CSDI和WSDI），得到基于CMIP5情景模拟的ETCCDI极端温度指数数据集。与CLIMDEX（http://www.climdex.org）指数计算结果比较，QRM计算结果具有消除了日温度序列自相关的影响，以及使用统一算法的优势。数据集包括19个模式在4种情景下的6个指数的年值数据，以及TN10p、TN90p、TX10p、TX90p四个指数的月值数据（CMCC-CM、CMCC-CMS和HadGEM2-CC模式的RCP2.6情景数据缺失）。历史情景时期从模式各自的起始年到2005年，RCP情景时期为2006-2100年。数据空间分辨率与各自模式输出的空间分辨率相一致。数据存储为*.nc（NetCDF）格式。原始数据量39GB（压缩后4.78GB）。（为便于比较，作者绘制了该数据集的全球平均时间序列和相应的CLIMDEX结果对比图）

The authors utilized daily maximum temperature (tasmax) and daily minimum temperature (tasmin) outputs from 19 Global Climate Models (GCMs) sourced from the official CMIP5 (Coupled Model Intercomparison Project Phase 5) data portal (https://pcmdi.llnl.gov/projects/cmip5/). Adopting the semi-parametric Quantile Regression Method (QRM) proposed by Yang and Xu (Atmospheric Science Letters, 2017, DOI: 10.1002/asl.724), they recalculated six percentile-based extreme temperature indices of the ETCCDI (Expert Team on Climate Change Detection and Indices, http://wcrp-climate.org/etccdi): TN10p, TN90p, TX10p, TX90p, CSDI, and WSDI for four greenhouse gas emission scenarios (historical, RCP2.6, RCP4.5, and RCP8.5), thereby developing the ETCCDI extreme temperature index dataset based on CMIP5 scenario simulations. Compared with the index calculation results from CLIMDEX (http://www.climdex.org), the QRM-derived results have the advantages of eliminating the impact of autocorrelation in daily temperature series and employing a unified algorithm. The dataset contains annual values of the six indices under the four scenarios for all 19 GCMs, as well as monthly values of the four indices TN10p, TN90p, TX10p, and TX90p, with RCP2.6 scenario data for CMCC-CM, CMCC-CMS, and HadGEM2-CC models missing. The historical scenario period spans from the respective start year of each model to 2005, while the RCP scenario periods cover 2006–2100. The spatial resolution of the dataset is consistent with that of the original model outputs. All data is stored in *.nc (NetCDF) format. The raw data totals 39 GB (4.78 GB after compression). For comparative convenience, the authors plotted global mean time series of this dataset and corresponding CLIMDEX results for comparison.

搜集汇总

背景与挑战

背景概述

该数据集是基于CMIP5的19个全球气候模型，采用半参数分位数回归方法（QRM）生成的ETCCDI极端温度指数数据集，覆盖历史、RCP2.6、RCP4.5和RCP8.5四种温室气体排放情景，包含六个极端温度指数（如TN10p、TX90p等）的年度和月度数据，时间跨度为2006-2100年，空间覆盖全球。数据集优势在于消除了日温度序列自相关影响，并使用统一算法提高了计算一致性，数据以NetCDF格式存储，原始大小为39GB。