Objective  Coherent lidar has the characteristics of small size, high detection efficiency and strong signal-to-noise ratio, and has been widely used in meteorology, aviation and wind power industries. But at present, most coherent wind lidar data products only use echo Doppler frequency shift information to retrieve wind field related products, and do not mine lidar echo signal strength information. Therefore, it is necessary to deeply process the echo signal data of lidar.

  Methods  Considering the insufficient development of the current coherent lidar data products, a new lidar detection technology based on coherent detection was introduced (Fig.1), which can detect various meteorological factors such as cloud height, extinction coefficient and visibility while detecting the wind field. Firstly, by analyzing the characteristics of the echo signal of the coherent wind lidar, the calculation method of the spectrum echo signal power of the coherent wind lidar was derived, and the echo signal strength information of each distance position in the detection range was obtained. Then the differential zero cross and threshold method were used to inverse the cloud bottom height, and the improved Klett method algorithm was used to inverse the atmospheric extinction coefficient, and the visibility measurement was realized. Finally, it was compared with horizontal visibility meter in weather station (Fig.3), laser altimeter (Fig.4) and 532 nm aerosol lidar (Fig.5) for verification.

  Results and Discussions   The results showed that the cloud height, visibility and the extinction coefficient were in good agreement, and lidar can continuously work in various meteorological environments. The cloud height detected by lidar and Vaisala showed that the correlation coefficient is 99.57%, which fully demonstrates the feasibility and accuracy of cloud height by lidar (Fig.3). The standard deviation and root mean square error of lidar are 5% and 7.2% respectively. The trend of visibility data measured by lidar and visibility meter is consistent, and the correlation coefficient is 82.06% (Fig.4). Within 30 km visibility, the accuracy of standard deviation was 12.2%, and root mean square error was 22.5%, which may be due to the following reasons. Firstly, the parameter selection in the algorithm was not accurate enough, and there is still room for optimization in the algorithm; Secondly, at present, the coherent wind radar system was a single polarization channel detection system, which still had signal detection defects. To solve this problem, the optimization algorithm and the integrated dual-polarization detection channel system were needed for further research and optimization. The measurement of extinction coefficient had experienced many meteorological environments such as sunny day, cloudy day, rainy day and foggy day. Extinction coefficient comparison in haze day showed that the results were basically consistent, including the position and thickness of clouds, extinction coefficient, aerosol content and some detailed characteristics (Fig.6). Extinction coefficient comparison in clear day showed that the detection value of coherent lidar increased steadily, while aerosol lidar increased speedily (Fig.7). The reason may be that there are differences between the two devices, which needs further study. Extinction coefficient comparison in overcast and rainy day showed that the detection values of the two devices were consistent with the actual weather conditions. When precipitation occurs, there was a lack of detection by coherent detection lidar. Although aerosol lidar had detection data, it was still unacceptable (Fig.8). Affected by heavy fog, the detection distance of the two lidars were limited, which was less than 500 m. In the morning, the aerosol lidar was missing data in heavy fog, while coherent lidar is still working (Fig.9). There were two main reasons for this phenomenon. First, the geometric factor for the separate antenna mode of transceiver and receiver. Second, in foggy day, the aerosol content was high, and the detection of aerosol lidar reached saturation.

  Conclusions  The results showed that the 1550 nm coherent lidar had the ability to detect atmospheric extinction coefficient, cloud height and visibility. The accuracy of cloud height comparison can reach 5.0%, and the accuracy of visibility comparison can reach 12.2%. The detection results were highly consistent with the detection results of aerosol lidar. It can be measured in sunny, cloudy, rainy and foggy weather continuously.