基于不同标准化指标的农业干旱时空分析及影响评估
Spatiotemporal Analysis and Impacts Assessment of Agricultural Drought Based on Different Standardized Indices

降水亏缺以及土壤水分的降低导致干旱的发生。中国是世界面积第三的大国,同时也是个干旱频发的国家。近年来,气温升高和夏季降水日数减少导致不同气候区域和地理区域降水量的减少。土壤储水量的降低也引起土壤含水量的亏缺以及土壤水分的降低。由于中国人口的不断增加和经济的发展,粮食需求量也相应增大,降水和土壤水分的降低导致的气象和农业干旱会大大影响对农业粮食生产,因此,对干旱进行分析是十分重要的。本研究旨在分析中国不同气候区域降水和土壤水分的变化,以及利用标准化降水指数(SPI)和标准化土壤水分指数(SSI)对1月、3月、6月以及12个月等4个时间尺度的干旱趋势进行评价。本研究使用的数据是通过CLDAS-V1.0和ERA-Interim再分析模型得到的与我国气象站点相对应的模型网格数据集。为了验证模型的可靠性,在参照期内(2008-2016年)对实际观测数据和基于模型的数据进行了对比。随后,对整个研究期内(1986-2016)不同时间尺度的SPI和SSI进行改进的Mann-Kendall趋势检验。为了与改进的Mann-Kendall趋势检验作比较,应用泰森和线性回归方法对研究区域的数据进行趋势检验。最后,对不同时间尺度下SPI和SSI的相关性进行了评估。得出的主要研究结果有:(1)中国中部区域的旱情呈持续增加的趋势,SSI序列较SPI序列表现的干旱事件更少,这意味着降水量(气象干旱)少不一定立即导致土壤水分(农业干旱)的降低。(2)不同时间尺度下SPI和SSI的相关系数随着时间尺度的增大而减小,其中6区(青藏高原)的相关性较其他区域差。此外,SPI-6与SSI-1之间的相关性比其他时间尺度更为密切,尤其是在湿润区的生长季节内。(3)在湿润地区,SPI与SSI在1月尺度的相关性最强(Pearson相关系数为0.65),对于不同的分区,它们的相关性不同,这表明SPI与SSI的相关性是由区域差异和气候差异造成的。

Drought has been a significant resulted from shortage in precipitation and soil moisture.China is the third largest country in area in the world and has been frequently experienced drought.Recently,increasing temperature,less in rainfall days in summer season caused the different climate regions and geographical suffered by drought.Then the decreasing of storage water in soil layers led to shortage in soil water content.Since the increasing population and economic facilities in China,so the food demand is a major subject.Hence,it is vital to analyze drought considering the decreasing precipitation and soil moisture which influenced to agricultural production is a must.This study aims to assess the variation in precipitation and soil moisture in different climate regions over China,investigate the trends in drought by applying the Standardized Precipitation Index(SPI)and Standardized Soil Moisture Index for 4 timescales – from 1 month to 12 months.The model grid based-datasets that corresponds to meteorological stations distributed in China were obtained from CLDAS-V1.0 and ERA-Interim reanalysis.Observed and model-based data were compared for reference period(2008-2016)in order to examine the reliability of model.Later,modified Mann-Kendall trend test was applied on the SPI and SSI at different timescales for the entire period(1986-2016).The trends were also detected by Theil-Sen and linear regression for comparing with Mann-Kendall trend test over study areas.Afterwards,the correlation assessment analysis was used to investigate the correlation coefficient between meteorological drought(SPI)and agricultural drought(SSI).The results showed that:(1)In short,a persistent increasing drought trend was detected for partial of China such as central China.However,SSI indicated less drought events than SPI,which implied that a low precipitation condition(meteorological drought)may not absolutely lead to a low soil moisture condition(agricultural drought).(2)The correlation coefficients are decreasing with increasing time scale for most of the sub-regions and the correlations were poor in sub-region 6(Qinghai-Tibetan Plateau).Additionally,the relationship between SPI-6 and SSI-1 was very important especially in the humid areas during the growing season,which was more closely related to agricultural drought.(3)While the strongest relationship between SPI and SSI appeared at the 1-month timescale(Pearson ‘correlation coefficient being 0.65)in humid areas.Also,the correlations varied in different sub-regions of China.This confirms that the relationship between SPI and SSI can be explained by area properties and climatic conditions.

标准化降水指数; 标准化土壤水分指数; 趋势; 统计关系; 空间分布;

standardized precipitation index(SPI); standardized soil moisture index(SSI); trend; statistical relationship; spatial distribution;

国家十三五重点研发计划(编号:2017YFC0403303);

李毅;

S423

71656649K