Abstract. In this paper, we introduce a spectral surface reflectivity climatology based on observations made by TROPOMI on board the Sentinel-5P satellite. The database contains the directionally dependent Lambertian-equivalent reflectivity (DLER) of the Earth's surface for 21 wavelength bands ranging from 328 to 2314 nm and for each calendar month. The spatial resolution of the database grid is 0.125° × 0.125°. A recently developed cloud shadow detection technique is implemented to avoid dark scenes due to cloud shadow. In the database, the anisotropy of the surface reflection is described using a third-order parameterisation of the viewing angle dependence. The viewing angle dependence of the DLER is analysed globally and for a selection of surface type regions. The dependence is found to agree with the viewing angle dependence found in the GOME-2 surface DLER database. Differences exist, related to the actual solar position. On average, the viewing angle dependence in TROPOMI DLER is weaker than for GOME-2 DLER, but still important. Validation of the new database was first performed by comparison of the non-directional TROPOMI surface LER with heritage LER databases based on GOME-1, OMI, SCIAMACHY, and GOME-2 data. Agreement was found within 0.002–0.02 in the UV-VIS (below 500 nm), up to 0.003 in the NIR (670–772 nm), and below 0.001 in the short-wave infrared (SWIR) (2314 nm). These performance numbers are dominated by the performance over ocean, but they are in most cases also representative for land surfaces. For the validation of the directional TROPOMI surface DLER, we made use of comparison with the MODIS surface bi-directional reflectance distribution function (BRDF) for a selection of surface type regions. In all cases the DLER performed significantly better than the traditional LER, and we found good agreement with the MODIS surface BRDF. The TROPOMI surface DLER database is a clear improvement on previous surface albedo databases and can be used as input not only for satellite retrievals from TROPOMI observations, but also for retrievals from observations from other polar-orbiting satellite instruments provided that their equator crossing time is close to that of TROPOMI. The algorithm that is introduced in this paper can be used for the retrieval of surface reflectivity climatologies from other polar satellite missions as well, including Ocean and Land Colour Instrument (OLCI) on the Sentinel-3 satellites, Sentinel-5, and Multi-viewing Multi-channel Multi-polarisation imager (3MI) on the MetOp-SG-A1 satellite to be launched in 2025, as well as the future CO2M mission.

中文翻译：

---

根据 TROPOMI 观测确定的定向表面反射气候学

摘要。在本文中，我们介绍了基于 TROPOMI 在 Sentinel-5P 卫星上进行的观测的光谱表面反射率气候学。该数据库包含地球表面每个日历月 328 至 2314 nm 21 个波段的方向相关朗伯等效反射率 (DLER)。数据库网格的空间分辨率为0.125°×0.125°。采用最近开发的云阴影检测技术来避免由于云阴影而导致的黑暗场景。在数据库中，表面反射的各向异性是使用视角依赖性的三阶参数化来描述的。对 DLER 的视角依赖性进行全局分析并选择表面类型区域。发现该依赖性与 GOME-2 表面 DLER 数据库中发现的视角依赖性一致。存在差异，与实际的太阳位置有关。平均而言，TROPOMI DLER 中的视角依赖性弱于 GOME-2 DLER，但仍然很重要。首先通过将非定向 TROPOMI 表面 LER 与基于 GOME-1、OMI、SCIAMACHY 和 GOME-2 数据的传统 LER 数据库进行比较来验证新数据库。在 UV-VIS（低于 500 nm）中，一致性在 0.002-0.02 范围内，在 NIR（670-772 nm）中一致性高达 0.003，在短波红外（SWIR）（2314 nm）中一致性低于 0.001。这些性能数据主要由海洋上的性能决定，但在大多数情况下它们也代表陆地表面。为了验证定向 TROPOMI 表面 DLER，我们利用与 MODIS 表面双向反射分布函数 (BRDF) 的比较来选择表面类型区域。在所有情况下，DLER 的性能都明显优于传统的 LER，并且我们发现与 MODIS 表面 BRDF 非常一致。 TROPOMI 表面 DLER 数据库是对以前的表面反照率数据库的明显改进，不仅可以用作 TROPOMI 观测数据的卫星反演输入，还可以用作其他极轨卫星仪器观测数据的反演，只要它们的赤道穿越时间接近即可。到特罗波米的。本文介绍的算法也可用于从其他极地卫星任务中检索表面反射率气候，包括 Sentinel-3 卫星、Sentinel-5 和多视图卫星上的海洋和陆地颜色仪器 (OLCI)将于 2025 年发射的 MetOp-SG-A1 卫星上的多通道多偏振成像仪 (3MI) 以及未来的 CO2M 任务。