红外大气探测干涉分光仪(Infrared Atmospheric Sounding Interferometer, IASI)能够获取平流层的大气温度廓线，而且由其 提取的大气温度扰动信息可用于平流层的重力波特征分析。基于神经网络方法，利用典型 大气廓线库和大气辐射传输模式建立了由IASI反演25 ～ 60 km高度范围内的平流层大气温度廓 线的算式，并结合再分析资料对反演结果进行了检验。模拟试验结果表明，平流层大气 温度反演的偏差主要在0 K附近且不超过1 K，均方根误差处在2 ～ 6 K之间，且50 ～ 5 hPa之间的均方 根误差在3 K以内。对比验证结果表明，本文的平流层大气温度反演与欧洲中期天气预报中心(European Centre for Medium-range Weather Forecasts, ECMWF)的再分析资料在总体上具有 可比性，而且反演速度更快，覆盖的区域更完整。

Infrared Atmospheric Sounding Interferometer (IASI) can be used to acquire stratospheric temperature profiles and analyze stratospheric gravity waves. On the basis of the neural network, a formula for IASI to retrieve the stratospheric temperature profiles in the altitude range from 25 km to 60 km is established by using the typical atmospheric profile database and the atmospheric radiation transfer mode. The retrieval results are verified according to the reanalysis results. The simulation test results show that the retrieval deviations of stratospheric temperatures are mainly close to 0 K and are not greater than 1 K. The root mean square error is between 2 K and 6 K and the root mean square error in the range from 50 hPa to 5 hPa is within 3 K. The comparison results show that the retrieval of temperature profiles from IASI_SFOV is comparable to the reanalysis data from ECMWF on the whole. The retrieval is faster and has a more complete coverage area.