Abstract: Taking erectophile type continuous vegetation as a example, authors had done many Monte Carlo simulations, and established empirical analytic expressions of radiances with component temperature, soil emissivity and leaf area index. Empirical analytic expressions were used to construct objective function and genetic algorithm was employed to synchronously retrieve 5 parameters, such as component temperature, soil emissivity and LAI, from thermal infrared 2 channels and 2 angles data. Many experiments of genetic algorithm inversion from simulated data were conducted, results show that it is very robust to retrieve component temperature using genetic algorithm；genetic algorithm can cope with uncertainty inversion problem pretty well if full advantage of priori knowledge was taken.Comparison between inversion results and ground-truth data has been done. This paper offers a new example to retrieve component temperature from multi-channel, multi-angle thermal infrared data based on the model of directionality of thermal radiance.

Key words: genetic algorithm, inversion of component temperature, multi-channel, multi-angle thermal infrared remote sensing, continuous vegetation

摘要： 在热辐射方向性规律的基础上，以喜直型连续植被为例，进行了大量的Monte Carlo模拟，建立了辐射亮度和组分温度、植被叶面积指数及土壤比辐射率之间的经验函数关系。采用遗传算法，从热红外2个波段2个角度数据中，同时反演混合像元组分温度、土壤比辐射率和叶面积指数等5个参数。通过对模拟的观测数据进行大量的遗传算法反演试验，结果表明，遗传算法反演组分温度非常稳健，在宽松的先验知识条件下，遗传算法可以解决不确定性反演问题。遗传算法反演结果和野外实测数据作了比较，证实了反演方法的正确性。为基于热红外方向性辐射模型反演地表组分温度，提供了新的实例。

关键词: 遗传算法, 组分温度反演, 多波段多角度热红外遥感, 连续植被