基于ETM植被指数和冠层温度差异遥感监测棉花冷害

Monitoring Frost Disaster of Cotton Based on Difference of Vegetation Index and Canopy Temperature by Remote Sensing

大范围地、及时地遥感监测棉花的冻害状况及损失对安排救灾、灾后评估有着现实的意义。利用2001年6月7日、8月10日和2000年8月7日ETM影像，结合农业灾害和农作物生长发育统计数据，通过植被指数变化和冠层温度差异对新疆沙湾2001年8月初棉花结桃时发生的冷害进行遥感监测。结果表明：与往年未遭受冷害的同期棉花植被指数相比，棉花植被指数NDVI绝对差值降低区域占67．8％，其中下降0～O．2占51％，下降大于0．2占17％，降低百分比处于0～20％。植被指数和温度图像散点图呈现显著负相关，相关系数-0．63。其中未受冷害影响，植被指数增加，长势较好的棉花冠层温度平均为26．4℃，植被指数未变化区域为27．6℃，植被指数降低较多，冷害程度较重区域冠层温度约为29．3℃，冠层温度差异显著。基于ETM遥感影像植被指数变化幅度和冠层温度差异可用于冷害程度区域划分。

Frostbite occurrence lies mainly on condition of weather and vegetauon, Frostbite can be monitored by remote sensing according to the difference of vegetation index,which has higher spatial and temporal resolution, and to the canopy temperature,which has higher precision of temperature retrieval. In addition, considering the aerosol and humidity effect on the temperature retrieval and the quality of the remote sensing image,monitoring the frostbite by remote sensing is rather complicated and difficult. North Xinjiang is the main production zone of cotton in China, where cotton is influenced seriously by later spring frost or early autumn low-temperature injury. Remote sensing is proved feasible in monitoring crop growth, especially after stresses. But monitoring frostbite of cotton with remote sensing has been rarely studied so far. So combining with the statistical data of agriculture disaster and crop development, according to the difference of NDVI and canopy temperature （CT）, the remote sensing was applied to three temporal Landsat ETM images in 2000-08-07,2001-06-07 and 2001-08-10, the remote sensing monitor the frostbite of cotton happened in Xinjiang Shawan and Shihezi county. The results showed the pixels of the absolute difference of NDVI fallen was 67.8%, the year fallen great than 0.2 was 17 %, the percent of difference NDVI decreased within 20%, apparently, the crop status of cotton in 2001 was worse than the year 2000s. The scatter plot between canopy temperature and NDVI showd the remarkably negative correlation,the correlation coefficient was -0.63. In the cotton zone,where didn＇t affected by frostbite, NDVI increased bigger, the canopy temperature was 26. 4℃ ;where affected faintly by frostbite, the canopy temperature was 27.6℃ ;where affected seriously by frostbite,the canopy temperature was 29.3℃. Based on the difference of NDVI and canopy temperature, frostbite and the degree of the disaster can be monitored.