Correction method for cloud temperature retrieval combining rotational Raman, vibrational Raman, and elastic scattering lidar channels

Raman lidar is an important technique for atmospheric temperature detection. However, under thick-cloud conditions, the accuracy of in-cloud temperature retrieval is severely compromised by leaked elastic signals beyond the system’s conventional suppression capabilities. To address this problem, we propose a four-channel joint-retrieval cloud temperature correction method built on low- and high-quantum-number rotational Raman (RR), vibrational Raman (VR), and elastic scattering channels. Using the leakage-immune VR signal as a stable intermediate variable, the crosstalk coefficients for the two RR channels are calculated, thereby reducing biases in in-cloud temperature retrieval. The proposed method was validated through simulation experiments and further compared with other in-cloud temperature retrieval approaches using observational data. The results from multi-day observations indicate that, within cloudy regions, the mean absolute error (MAE) of Raman lidar temperature retrievals was reduced to within 1 K when compared with radiosonde data, and the root mean square error (RMSE) was lowered by over 86.34%. Compared with existing methods, it reduced RMSE by about 37.78% and provided more continuous spatiotemporal temperature evolution in 10 h of nighttime observations. The calibrated crosstalk coefficients can correct RR signals on different observation dates.