气候模拟数据订正方法在作物气候生产潜力预估中的应用——以江苏冬小麦为例

Application of Bias Correction Method for Simulated Climate Data in Projection of Crop Climatic Potential Productivity——A Case Study of Winter Wheat in Jiangsu

利用全球气候模式BCC_CSM1.1(Beijing Climate Center Climate System Model version 1.1),耦合区域气候模式Reg CM4(Regional Climate Model version 4)输出的1961-1990年(基准时段)气候模拟数据,并根据同期实测资料,确定模拟值和实测值之间的非线性传递函数与方差订正参数,构建气候模拟数据的误差订正模型。利用1991-2005年(验证时段)模拟数据与实测资料验证该模型的有效性,并对RCP(Representative Concentration Pathway)情景下2021-2050年(未来时段)气候模拟数据进行订正,同时通过潜力衰减方法预估未来江苏冬小麦气候生产潜力格局。结果表明:将气候模拟数据订正方法应用到作物气候生产潜力预估是有效的。以均值传递函数和方差信息建立的模型可以较好订正江苏逐日气候模拟数据。订正后的秋冬季气温、辐射量、蒸散量和冬春季降水量模拟偏差明显减小。在此基础上研究发现,冬小麦的成熟期在RCP4.5和RCP8.5情景下介于153~175和153~174,较基准时段均明显提前。两种情景下冬小麦气候生产潜力分别介于10335~14368kg·hm^(-2)和9991~13708kg·hm^(-2),较基准时段呈下降趋势。其变异系数分别介于7.6%~14.6%和7.5%~13.6%,较基准时段呈增大趋势,表明江苏冬小麦气候生产潜力总体趋于不稳定。未来时段,徐州中北部、连云港东北部、宿迁西部以及盐城东南部冬小麦在RCP4.5和RCP8.5情景下可以保持相对较高的生产潜力(≥12501kg·hm^(-2)),该省应确保这些地区的冬小麦种植用地。研究建议,作物气候生产潜力预估应考虑利用研究区实测资料对气候模拟数据进行订正,以提高预估可信度。

A bias correction model for simulated climate data was constructed. The nonlinear transfer function between simulations and observations and the parameters for variation correction were determined based on historical simulations outputting from a regional climate model Reg CM4 coupled with a global model BCC_CSM1.1（Beijing Climate Center Climate System Model version 1.1） and the observations during baseline period from 1961-1990. The effectiveness of bias correction model was verified using simulated climate data and observed data during validation period from 1991-2005. This model was then used to correct climate data under RCP（Representative Concentration Pathway） scenarios during future period from 2021-2050. Meanwhile, the spatial patterns of winter wheat climatic potential productivity of Jiangsu were projected via productivity decay method under future scenario climates. The results indicated that it was effective to apply bias correction method for simulated climate data in projection of crop climatic potential productivity. The bias correction model with mean value and variation information was an excellent way of correcting simulated climate data at daily scale in Jiangsu. The bias of simulated temperature, radiation, evapotranspiration in autumn and winter as well as precipitation in winter and spring was reduced obviously after correction. Then, on the basis of bias correction, the maturity date of winter wheat in Jiangsu was projected between 153-175 and 153-174 respectively under RCP4.5 and RCP8.5 scenarios, and would advance obviously compared with baseline. The estimated climatic potential productivity of winter wheat was projected between 10335-14368kg·ha^（-1） and 9991-13708kg·ha^（-1） respectively, and would tend to be lower than that during baseline period. Accordingly, the coefficient of variation of climatic potential productivity was projected between 7.6%-14.6% and 7.5%-13.6% respectively, and would increase compared with baseline, which indicated a tendency towards unstable for climatic potential productivity of winter wheat in Jiangsu. Moreover, during future period, the climatic potential productivity of winter wheat would remain relatively high（ ≥12501kg·ha^（-1）） in central and northern Xuzhou, northeastern Lianyungang, western Suqian and southeastern Yancheng under RCP4.5 and RCP8.5 scenarios. Thus, the cultivated land for winter wheat in these regions should be guaranteed by the government of Jiangsu. Our results suggested a consideration of bias correction for simulated climate data using observations of study region before estimating crop climatic potential productivity, in order to enhance the credibility of the projections.