Evaluation of MODIS and Himawari‐8 Low Clouds Retrievals Over the Southern Ocean With In Situ Measurements From the SOCRATES Campaign
Evaluation of MODIS and Himawari‐8 Low Clouds Retrievals Over the Southern Ocean With In Situ Measurements From the SOCRATES Campaign
复制标题
DOI：
10.1029/2020ea001397
复制
发表时间：
2021-01
影响因子：
3.1
通讯作者：
Litai Kang;R. Marchand;W. Smith
中科院分区：
文献类型：
--
作者：
Litai Kang;R. Marchand;W. Smith
在2018年1月至2月的南大洋云辐射气溶胶传输实验研究期间收集的飞机观测数据用于评估来自三个卫星成像仪数据集的云特性：（1）中等分辨率成像光谱仪2级（收集6.1）云产品，（2）CERES-MODIS第4版云产品，以及（3）NASA SatCORPS Himawari-8云产品。总的来说，卫星检索比较好，与现场观测，小偏差和适度良好的相关系数时，考虑到所有的飞机配置文件，有重合的中分辨率成像光谱仪观测。然而，当应用于具有一致Himawari-8观测结果的较大轮廓集时，Himawari-8产品确实显示出统计学显著的平均偏倚，有效半径（re）约为1.2 μm，光学深度（τ）约为2.6。重新检索的总体平均偏差较低，部分原因是补偿了非降水或轻度降水病例与降水较重病例之间的误差。当降水出现在云顶附近时，无降水和轻微降水的情况下，re略微偏高（约0.5-1.0 μm），而当降水出现在云顶附近时，严重降水的情况下，re偏低约3-4 μm。非和轻度降水条件下的偏差是由于（至少部分）在检索中假设的液滴尺寸分布太宽。这些偏差最终传播到回收的液态水路径和数量集中。
Aircraft observations collected during the Southern Ocean Cloud Radiation Aerosol Transport Experimental Study in January‐February of 2018 are used to evaluate cloud properties from three satellite‐imager datasets: (1) the Moderate Resolution Imaging Spectroradiometer level 2 (collection 6.1) cloud product, (2) the CERES‐MODIS Edition 4 cloud product, and (3) the NASA SatCORPS Himawari‐8 cloud product. Overall the satellite retrievals compare well with the in situ observations, with little bias and modest to good correlation coefficients when considering all aircraft profiles for which there are coincident MODIS observations. The Himawari‐8 product does, however, show a statistically significant mean bias of about 1.2 μm for effective radius (re) and 2.6 for optical depth (τ) when applied to a larger set of profiles with coincident Himawari‐8 observations. The low overall mean‐bias in the re retrievals is due in part to compensating errors between cases that are non‐ or lightly precipitating, with cases that have heavier precipitation. re is slightly biased high (by about 0.5–1.0 μm) for non‐ and lightly precipitating cases and biased low by about 3–4 μm for heavily precipitating cases when precipitation exits near cloud top. The bias in non‐ and lightly precipitating conditions is due to (at least in part) having assumed a drop size distribution in the retrieval that is too broad. These biases in the re ultimately propagate into the retrieved liquid water path and number concentration.