This study aims to assess the quality of retrieved stratospheric aerosol data from the Stratospheric Aerosol and Gas Experiment（SAGE） in China. Data from 10-year observation of ground Li DAR in Hefei, China were comparedwith the corresponding SAGE aerosol dataset. This work focused on the 20—40 km aerosol extinction profile and AOD to elucidate stratospheric aerosol. Two methods were used to form matching pair for Li DAR and SAGE because of the limited observation samples in Hefei for SAGE limb detection. The first method is singe profile matching, which matches the observed aerosol profile within the same day with the temporal and 10° spatialdistancetoobtain approximately 24 matching pairs. The second method is multi-profile matching, which matches the aerosol profiles with the 10° spatial distance within a certain period, such as per season; subsequently, differences in the averaged profiles were compared among the four seasons.Comprehensive comparisons show that aerosol variation trends are consistent when derived from Li DAR and SAGE in 10—30km. The positions of peaks and valleys of the aerosol extinction profile are similar. In addition, aerosol seasonal differences in 20—35km from Li DAR and SAGE, respectively, are relatively smaller, which reflects the stability of stratospheric aerosol. However, the detailed value distribution shows that the SAGE results are generally smaller than those of Li DAR, and their corresponding AODs differ within 20—40km. The quantitative results show that the aerosol AOD within 20—40 km as observed by Li DAR is approximately 0.004; the corresponding results retrieved by SAGE is approximately 0.002, which is only half of the former. The evident differences in Li DAR and SAGE results for stratospheric aerosol AOD and their extinction profiles are mainly attributed to their different measurement and retrieval methods. Furthermore,large temporal and spatial scales are required to match the Li DAR and SAGE observations, which may cause uncertainties in the compared results because of different detection technologies employed. Moreover, the SAGE aerosol profile was retrieved from Mie scattering theory.By contrast, the Li DAR aerosol profile was calculated using Fernald method, which requires the adjustment of several important constants,such as back scattering ratio and reference height. Artificial choices for these constants increase the possibility of errors for Li DAR aerosol retrieval. Therefore, different and complex retrieval methods can increase bias in the compared results. Comparisons of aerosol within20—40 km from SAGE and Li DAR in Hefei, China show that variation in stratospheric aerosol is generally reasonable as reflected by SAGE retrieval. The quantitative AOD value is systemically lower than half of that detected by ground Li DAR.

stratospheric aerosol;SAGE;ground LiDAR;extinction coefficient;optic depth