基于分位数映射法的黑河上游气候模式降水误差订正

区域气候模式降水弥补了高寒山区气象站点稀少的缺陷,是水文模拟的重要驱动变量。然而,高寒山区模式输出降水的总量和频率都存在较大不确定性。因此,改进了用于降水频率纠正的分位数映射法(Quantile Mapping,QM),对中尺度数值预报模式(Weather Research and Forecasting model,WRF)模拟的黑河上游日降水输出数据进行误差订正。选取第95分位和第98分位降水量为阈值,选择2004-2009年为建模时段,2010-2013年为验证时段,使用分段拟合的方法建立传递函数,侧重于对极端降水进行单独订正。研究结果表明:该方法不仅对降水空间分布有明显的改善,对极端降水也有很好的订正效果。订正前模式模拟日降水与台站之间的均方根误差为3.41 mm·d^-1,绝对偏差为115.67 mm·y^-1,订正后均方根误差减少为3.11 mm·d^-1,绝对偏差有明显改善,为60.3 mm·y^-1。订正后流域内年降水空间分布更加合理,年降水量也更接近于观测降水插值结果,其空间相关系数由0.74改善为0.94。春、夏季订正效果优于秋、冬季,其中夏季订正效果较为明显,订正前降水偏差百分比在-0.1~0.1以内的区域面积仅占流域总面积的28%,而订正后占比增加至66%。同时,该方法对极端降水有较好的订正效果,减小了日降水强度(SDII)和极强降水量(R99p)的模拟偏差,订正后的第95分位模拟降水与观测降水插值的相关系数由0.15提高到0.48。本研究为站点稀少的黑河上游提供了一种更有效的误差订正方案,有利于为寒区水文研究获取更精确的降水数据。

Bias Correction of Climate Model Precipitation in the Upper Heihe River Basin based on Quantile Mapping Method

Regional climate model precipitation makes up for the deficiency of scarce meteorological stations in the alpine and cold mountains,which is an important variable of hydrological simulation.However,there is great uncertainty of model outputs in alpine region,both in the total amount and frequency.In view of this,we have improved the existing quantile mapping method(QM)for precipitation frequency correction,and corrected the daily precipitation simulated by WRF model of the upper reaches of Heihe river.Precipitation at the 95th and 98th percentiles were selected as the threshold,and 2004-2009 as the modeling period and 2010-2013 as the validation period.The transfer function was established by piecewise fitting method,focusing on correct the simulated extreme precipitation separately.The results show that the method not only has a significant improvement on the spatial distribution of precipitation,but also has a great correction effect on extreme precipitation.Before the correction,the RMSE between the sinulated and the stations precipitation was 3.41 mm·d^-1,and the absolute deviation was 115.67 mm·y^-1.After correction,the RMSE was reduced to 3.11 mm·d^-1,and the absolute deviation was significantly improved to 60.3 mm·y^-1.The spatial distribution of annual precipitation in the basin improved obviously,and the annual precipitation amount is closer to the precipitation interpolated,the spation correlation coefficient increased from 0.74 to 0.94.The correction effect in spring and summer is better than that in autumn and winter,among which the correction effect in summer is the most obvious.Before the correction,precipitation deviation during-0.1~0.1 only accounts for 28%of the total area of the basin,while after the correction the proportion increased to 66%.At the same time,this method performs well in extreme precipitation correction,reduces the simulation deviation of simple daily intensity index(SDII)and very extremely precipitation(R99p),the correlation coefficient of simulated and interpolated precipitation at the 95th percentiles was improved from 0.15 to 0.48.This study provides a more effective bias correction method for the upper Heihe river with scarce stations,which is conducive to obtaining more accurate precipitation data for the hydrological research of the cold regions.