基于遗传算法综合Terra/Aqua MODIS热红外数据反演地表组分温度

混合像元组分温度相对来说更有应用价值,而多角度热红外遥感的发展推动了混合像元组分温度反演基础和方法的发展.根据前期数值模拟得到Terra和Aqua卫星上的MODIS测量可以认为是同一卫星在两个不同观测时间和观测角度上的测量,综合利用Terra和Aqua卫星上的MODIS数据反演混合像元内土壤和植被组分温度.根据混合像元热红外辐射模型,利用遗传算法,分别模拟Terra卫星MODIS的32和33通道,以及Terra和Aqua卫星上MODIS的32通道辐射反演了河北怀来试验区范围内植被覆盖率、土壤组分温度和比辐射率、植被组分温度和比辐射率等表面参数.通过与实测数据进行比较,综合利用上午Terra和下午Aqua卫星32通道数据反演的上午植被组分温度与地面同步测量温度偏差在1℃内,而利用上午Terra卫星32和33通道数据反演的上午植被组分温度与地面同步测量值偏差在1.4℃内.尽管利用双星数据反演的组分温度精度相对较高,但针对同一个像元,两个方案反演的结果有一定偏差.

Genetic algorithm based surface component temperatures retrieval by integrating MODIS TIR DATA from Terra and Aqua satellites

The component temperatures are key parameters in the environment problems. The multi-angle thermal infrared data are important for the retrieval of component temperature. MODIS (Moderate Resolution Imaging Spectroradiometer) is one of the sensors onboard EOS-Terra and EOS-Aqua, two sun-synchronous polar orbiting satellites. The sensitivity analysis shows that the same MODIS bands of Terra and Aqua satellites measures in the same behavior under the same variations of atmospheric water vapor, solar and viewing geometries. These MODIS bands may be considered to be the observations of a satellite at two viewing angles and two viewing times. In the study, the foliage and soil component temperature were retrieved by integrating these MODIS data. Based on linear thermal infrared radiation model, the bands 32 and 33 in Terra-MODIS, or the band 32 in Terra-MODIS and Aqua-MODIS, were separately used to retrieve the vegetation cover ratio, soil component emissivity and temperature, foliage component emissivity and temperature by employing genetic algorithm in Huailai area, Hebei Province. By comparing with the synchronous field measurements, the maximum error in the component temperatures is less than 1℃ for the integrated Terra/ Aqua MODIS data, 0.4℃ smaller than that by just Terra-MODIS data. However, for a pixel, its component temperatures are different by such two group data.