Objective Against the backdrop of increasingly frequent extreme weather events worldwide, traditional observation equipment can hardly meet the current demands, and there is an urgent need for fine vertical profiling of atmospheric temperature and humidity. To achieve precise detection and long-term, continuous, and stable lidar observation and to promote its operational application, this study designed and developed a Raman lidar system for simultaneous temperature and humidity detection. The system’s monitoring accuracy and data reliability have been verified to meet the requirements for operational applications.

Methods The Raman lidar system adopts a low-energy 355 nm laser as the emission source and a multi-stage spectroscopic receiving technology (Fig.1). The overall system is designed with a modular structure (Technical Parameters in Tab.1), enabling high-spatiotemporal-resolution continuous detection of temperature, relative humidity, and water vapor mixing ratio within the vertical range of 0.1-6 km. To conduct error analysis, the lidar data were compared with radiosonde data (Fig.2), and parameters including the absolute mean error, root mean square error, and correlation coefficient of temperature, humidity, and water vapor mixing ratio profiles were calculated. For case analysis, ERA5 reanalysis data were used for mechanism analysis and result verification.

Results and Discussions The evaluation of 62 concurrent lidar and radiosonde measurements confirmed that the relative errors of temperature (Fig.3), water vapor mixing ratio (Fig.6), and relative humidity (Fig.7) complied with standard meteorological observation requirements. All profiles displayed strong correlations and coherent temporal trends. The mean absolute errors were quantified across three vertical layers: 0.1-2 km, 2-4 km, and 4-6 km. The corresponding errors for temperature were 0.877 K, 0.922 K, and 1.551 K; for relative humidity, 5.528%, 8.024%, and 12.312%; and for water vapor mixing ratio, 0.415 g/kg, 0.456 g/kg, and 0.323 g/kg. A comprehensive statistical analysis of these errors across altitudes and monitoring periods was performed (Tab.2), which helped identify the dominant error sources in each layer. It was found that daytime observations exhibited greater susceptibility to background solar noise, whereas nighttime operations yielded significantly improved data consistency (Fig.4). A comparative assessment against systems reported in the literature (Tab.3) underscored the competitive advantages of the present system. Furthermore, the lidar was deployed to analyze the structure of a temperature inversion layer (Fig.8) and to track the evolution of temperature and humidity profiles preceding a rainfall event (Fig.12). The formation mechanisms of these weather phenomena were elucidated through supporting reanalysis data (Fig.9, Fig.13).

Conclusions The developed Raman lidar system for simultaneous temperature and humidity detection has demonstrated its capability for long-term, stable operation, successfully obtaining high-resolution vertical profiles of temperature, humidity, and water vapor mixing ratio. Validation against 62 radiosonde launches confirmed the system's accuracy and data consistency across different altitude layers, meeting operational meteorological monitoring requirements. Case studies of temperature inversion and pre-rainfall processes revealed detailed vertical structures and coupling mechanisms between temperature and water vapor. These observations, supported by reanalysis data, verified the formation mechanisms of the studied phenomena, underscoring the lidar's utility in analyzing complex weather dynamics and providing critical evidence for understanding regional severe weather. However, system performance is limited by a significantly degraded signal-to-noise ratio during daytime due to solar background noise, and insufficient detection accuracy above 6 km, which prevents effective coverage of the upper troposphere. Future work will focus on enhancing daytime performance, developing more robust signal processing algorithms, and advancing system automation with standardized quality control to achieve fully operational, long-term observational applications.