地表温度-归一化植被指数特征空间干旱监测方法的改进及应用研究  

An improvement of the land surface temperature-NDVI space drought monitoring method and its applications

作　　者：于敏[1] 程明虎[2] 刘辉[3] 

机构地区：[1]黑龙江省气象科学研究所,哈尔滨150001 [2]中国气象局气象探测中心,北京100081 [3]国家卫星气象中心,北京100081

出　　处：《气象学报》2011年第5期922-931,共10页Acta Meteorologica Sinica

基　　金：国家高技术研究发展计划(2007AA061901);国家自然科学基金项目(40905014)

摘　　要：地表温度-归一化植被指数(Ts-NDVI)特征空间干旱监测方法易受到监测期地表湿度、植被覆盖状况和卫星观测质量的影响,导致特征空间干、湿边不稳定,在基于长时间、大范围卫星观测资料的基础上对Ts-NDVI特征空间方法进行改进,将其称为通用特征空间方法。首先基于某观测时段,单独构建每年基于该单一年份单一时段卫星观测数据的Ts-NDVI特征空间,然后在已构建的各单一时段特征空间的干、湿边地表温度中,分别用最大值、最小值合成法提取各植被指数对应的多年最大、最小地表温度,分别构成各年通用的Ts-NDVI特征空间的干、湿边地表温度,最后通过线性拟合,获得通用特征空间的干、湿边方程。选择中分辨率成像光谱仪的归一化植被指数(NDVI)和地表温度(Ts)数据在中国主要粮食产区黑龙江省进行干旱监测试验,对改进前、后特征空间的植被温度状态指数进行对比分析,并用10、20cm土壤相对湿度和降水量进行检验。结果表明:通用特征空间方法合成的Ts-NDVI特征空间三角形结构特征显著,干、湿边明晰,比改进前更接近特征空间的理论边界,增强了特征空间干、湿边的稳定性,提高了干旱监测结果的精度,当单一时段特征空间稳定性较差时,改进效果最明显。通用特征空间对浅层土壤湿度更加敏感,与监测同期及前期总体降水都密切相关,且对降水的响应更敏感。The land surface temperature（Ts）-Normalized difference vegetation index（NDVI） space based drought monitoring is vulnerable to the soil moisture,vegetation cover and satellite observation state,which makes the dry edge and wet edge of Ts-NDVI space unstable,so it is likely to bring more errors in the drought monitoring.The paper develops a method named the general Ts-NDVI space using the multi-year satellite data in the same period to resolve the instability of the Ts-NDVI space.Firstly,the primary Ts-NDVI space based on the satellite data of some single period of some single year is respectively constructed.Then,from every NDVI,the maximum temperature is extracted from the dry edges of the primary spaces of all years to form the temperature of the general Ts-NDVI space dry edge.In the same way,the minimum temperature is extracted from the wet edges of the primary spaces of all years to form the temperature of the general Ts-NDVI space wet edge.Finally,by linear fitting,the equations of dry edge and wet edge of the general Ts-NDVI space are obtained.The composed general Ts-NDVI space can be applied to all years,in which the satellite data are used to compose the general space.With the MODIS 16-day NDVI（MOD13A2） and 8-day Ts（MOD11A2） data employed,the experiment is conducted over Heilongjiang Province,which is one of the main food production areas in China.The Vegetation Temperature Condition Index（VTCI） based on both the general space and the single space is worked out to demonstrate the difference between the primary Ts-NDVI space and the composed general Ts-NDVI space.The relative soil moisture in 10 cm and 20 cm depth,precipitation in the satellite observation period and the cumulative precipitation in the last satellite observation periods are used to validate the improvement of the general Ts-NDVI space.The results reveal that the composed general Ts-NDVI space using multi-year satellite data is clearly triangular in shape with the definite dry edge and wet edge,which are more approximate

关 键 词：中分辨率成像光谱仪 干旱监测 Ts-NDVI特征空间 方法改进 植被温度状态指数 

分 类 号：P407.4[天文地球—大气科学及气象学] P429