Abstract: Based on the sea surface backscattering model, a space-borne lidar data screening method using CALIOP dual-wavelength sea surface signal for data quality control is proposed. The global sea surface wind speed is retrieved from CALIOP's Version 4.1.0 level 1 data from January to September 2018. In order to improve data availability, under the condition of detecting sea surface wind speed in cloudless sea area, the cloud-penetrating CALIOP data were added, and then wind speed inversion was carried out after atmospheric correction and dual-wavelength signal quality control. At last, quasi-synchronous AMSR-2 wind speed data were used as reference values for verification. The two empirical relationships of Cox- Munk model and Hu model were used to invert wind speed respectively. For the data from January to March, the standard deviations of the former model are 1.14, 1.18 and 1.16m /s, respectively, with correlation coefficients of 0.90, 0.91 and 0.91, while the standard deviations of the latter are 1.17, 1.19 and 1.17m /s, respectively, with correlation coefficients of 0.91, 0.91 and 0.91. In addition, compared with the inversion study in cloudless sea, the available data amount using the proposed method increases by about 45.5%. The results show that the method proposed can not only ensure the accuracy of inversion but also effectively improve the utilization rate of data, which can provide more data sources for the study of sea surface wind field.

Key words: remote sensing, spaceborne lidar, sea surface backscatter, sea surface wind speed, signal extraction