地球中高层大气是研究大气过程乃至气候变化的重要区域，目前对中高层大气的长时间 观测和数据分析仍然非常欠缺。太赫兹临边探测技术能够全天时、近乎全天候地获得较高垂直分 辨率（1~5 km）的大气廓线，特别是对部分臭氧损耗相关的卤素气体敏感，是测量地球中高层大气 参数的重要手段。围绕太赫兹临边探测技术，系统回顾了太赫兹载荷的技术发展历程和现状：太 赫兹临边探测已成功实现中高层大气中多种痕量气体的高垂直分辨率廓线测量，但现有载荷仍面 临系统体积庞大、噪声抑制能力不足等瓶颈问题；基于最新预研的载荷方案，下一代太赫兹探测系 统主要侧重于低噪声和小型化技术的发展；当前主流的物理反演算法计算效率较低，通过引入人 工智能技术，可在保证精度的前提下显著提升反演效率；未来亟需突破太赫兹低噪声接收机和高 分辨率数字谱仪等核心技术，进一步推动我国太赫兹临边探测技术的发展。

Abstract： Long-term observations and data analysis of the Earth’s middle and upper atmosphere, an important region for the study of atmospheric processes and even climate change for the study of human activities and climate change, are still sorely lacking. Terahertz limb-sounding technology can obtain atmospheric profiles all day and near all-weather with high vertical resolution (about 1~5 km), and is particularly sensitive to some of the halogen gases associated with ozone depletion which is an important method to measure the Earth's middle and upper atmosphere parameters. The Earth’s middle and upper atmosphere constitutes a crucial region for studying atmospheric processes and even climate change. However, there remains a significant lack of long-term observational data and comprehensive analysis for this atmospheric layer. Terahertz limb sounding technology has emerged as a vital tool for measuring atmospheric parameters in these regions, capable of obtaining atmospheric profiles with relatively high vertical resolution (1~5 km) under near-all-weather conditions with diurnal availability. Particularly sensitive to halogen gases associated with ozone depletion, this technology offers unique advantages. Focusing on terahertz limb sounding technology, this study systematically reviews the technological evolution and current status of terahertz payloads: While existing systems have successfully achieved high vertical resolution measurements of multiple trace gases in the middle and upper atmosphere, current payloads still face challenges such as bulky system configurations and inadequate noise suppression capabilities. Based on newly developed prototype payloads, next-generation terahertz detection systems primarily emphasize advancements in low-noise and miniaturization technologies. Regarding data processing, conventional physical retrieval algorithms suffer from low computational efficiency. The introduction of artificial intelligence technology demonstrates potential to significantly enhance retrieval efficiency while maintaining accuracy. Future development urgently requires breakthroughs in core technologies including terahertz low-noise receivers and high-resolution digital spectrometers, which will further propel the advancement of terahertz limb sounding technology in China.