三极地区0.1º气溶胶光学厚度数据集（2000-2020）

2000-2020年三极地区0.1º气溶胶光学厚度数据集(也称为“Poles AOD Collection 1.0”气溶胶光学厚度（AOD）数据集），发展了冰雪上空的气溶胶卫星遥感反演算法（Aerosol Properties Retrieval over Snow，简称ARPS算法）, 结合Merra-2模式数据与MODIS卫星传感器AOD制作，数据覆盖时间从2000年到2020年，时间分辨率为逐日，覆盖范围为“三极”（南极、北极和青藏高原）地区，空间分辨率为0.1度。ARPS算法首先根据大气辐射传输方程建立地表反射率与天顶反射率之间的关系，建立气溶胶反演前向模型。之后，地表冰雪BRDF模型选择Kokhanovsky et al.（2012）的渐进辐射传输模型（ART）。最后假定不同观测角度获取的地表反射率的比值是不依赖波长的，利用多波长、多角度观测的卫星数据进行联合反演，求解气溶胶光学厚度。通过实测站点验证表明，数据相对偏差在35%以内，可有效提高极区气溶胶光学厚度的覆盖率和精度。

The 0.1° Aerosol Optical Depth (AOD) Dataset over the Three Poles Region from 2000 to 2020, also referred to as the "Poles AOD Collection 1.0" Aerosol Optical Depth Dataset, was developed by establishing the satellite remote sensing inversion algorithm for aerosols over snow and ice surfaces, namely the Aerosol Properties Retrieval over Snow (ARPS) algorithm. This dataset was constructed using MERRA-2 model data and AOD products from the MODIS satellite sensor. It covers the time period from 2000 to 2020, with a daily temporal resolution, and spans the Three Poles region including Antarctica, the Arctic, and the Tibetan Plateau, with a spatial resolution of 0.1 degrees. The ARPS algorithm first builds the forward model for aerosol retrieval by establishing the relationship between surface reflectance and zenith reflectance based on the atmospheric radiative transfer equation. Next, the asymptotic radiative transfer (ART) model proposed by Kokhanovsky et al. (2012) is selected as the surface snow and ice BRDF (Bidirectional Reflectance Distribution Function) model. Finally, assuming that the ratio of surface reflectance obtained from different observation angles is wavelength-independent, joint inversion is performed using multi-wavelength and multi-angle satellite observation data to solve for the aerosol optical depth. Validation using in-situ station measurements shows that the relative bias of this dataset is within 35%, which can effectively improve the coverage and accuracy of aerosol optical depth products in polar regions.