QM和QDM方法对中国极端气候的高分辨率气候变化模拟的误差订正对比

采用分位数映射（Quantile Mapping, QM）和delta分位数映射（Quantile Delta Mapping, QDM）两种误差订正方法对区域气候模式RegCM4在中国区域内模拟的逐日气温和降水数据进行订正。模式数据是5种不同全球气候模式驱动下的区域模式气候变化模拟结果。计算订正前后的极端气候指数进行对比分析，包括日最高气温极大值（TXx）、日最低气温极小值（TNn）、连续干旱日数（CDD）和最大日降水量（RX1day）。结果表明，5组模拟结果和其集合平均（ensR）都显示气温指数的模拟效果高于降水指数，其中对TXx模拟最好，对CDD的模拟最差；经过订正后，针对不同模式的两种订正结果都能够有效地减小模式与观测的偏差并提高了空间相关系数，且两种方法的订正效果无明显差别。对RCP4.5情景下未来变化的分析中，QM在一定程度上改变了模式模拟的未来变化幅度和空间分布特征，QDM则能够有效地保留所有极端指数的气候变化信号。从全国平均来看，除CDD外，所有指数未来都呈现增加趋势，且QDM订正结果与订正前模式模拟的变化趋势更为接近。建议在气候变化模拟的误差订正中采用QDM方法。

Bias Correction in Climate Extremes over China for High-Resolution Climate Change RegCM4 Simulations Using QM and QDM Methods

QM and QDM methods are used to correct simulated temperature and precipitation over China from a set of regional climate model (RegCM4) projection simulations in which the RegCM4 is driven by five different general circulation models. Extreme climate indices are derived separately based on both original-simulation and bias-corrected results, and a comparative analysis is conducted. Both the individual simulations and ensemble RCM (ensR) show higher simulation skill scores in the temperature indices compared to those in the precipitation indices in which the annual maximummax daily maximum temperature (TXx) is the best and the annual maximum consecutive dry days (CDD) is the worst. Biases can be effectively reduced using both the QM and QDM methods for all the five simulations and ensR. After bias corrections, the spatial correlations between the simulations and observations are certainly increased. No significant differences exist in the correction effects of the two methods. For future changes under RCP4.5, the QM modifies the originally projected change magnitudes and spatial patterns of changes, whereas the QDM effectively preserves the climate change signals in the extreme indices. Changes in the national average show that all the indices except the CDD will continuously increase, and the results corrected through the QDM are closer to the original future projection simulations. Therefore, employing the QDM to correct climate change simulations is recommended.