A reprocessing for climate of sea surface temperature from the along-track scanning radiometers: Initial validation, accounting for skin and diurnal variability effects
A reprocessing for climate of sea surface temperature from the along-track scanning radiometers: Initial validation, accounting for skin and diurnal variability effects
复制标题
DOI：
10.1016/j.rse.2011.02.028
复制
发表时间：
2012-01
影响因子：
13.5
通讯作者：
Owen Embury;C. Merchant;G. Corlett
中科院分区：
文献类型：
--
作者：
Owen Embury;C. Merchant;G. Corlett
对沿着迹扫描辐射计（ATSR）反演海表温度（SST）的气候再处理（ARC）进行了初步验证。ATSR-2和高级ATSR（AATSR）SST估计值与1995年至2008年期间的漂流浮标和系泊浮标观测值进行了比较。ATSR的主要估计是表层SST，而浮标测量表层以下的SST。因此，对趋肤效应、日分层和卫星观测时间的差异进行了调整。经过这样的调整，在~0.01K的范围内，卫星-原位差异在昼夜之间是一致的。由于海洋的日变暖和日变冷，卫星现场差异与观测时间的差异有关。这些数据被用来验证平均行为的物理和经验模型的升温/降温速率。除了最极端的情况（TCWV <5 kgm-2，TCWV > 60 kgm-2）外，调整后的AATSR和现场SST对纬度，总柱水汽（TCWV）和风速的系统差异小于0.1K。对于所有类型的检索，除了最低点只有两个通道（N2），区域偏差小于0.1K的80%的海洋。与漂流浮标的全球对比表明，夜间双视双通道（D2）SST偏暖0.06±0.23K，双视三通道（D3）SST偏暖0.06±0.21K（白天D2：0.07±0.23K）。仅最低点结果为N2：0.03± 0.33 K和N3：0.03± 0.19 K，表明算法间一致性提高到约0.02 K。这代表了对现有操作检索算法的显著改进，其中算法间不一致性>0.5K。与比漂流浮标更精确的热带系泊浮标相比，给出了较低的误差估计（N3：0.02±0.13K，D2：0.03±0.18K）。对于ATSR-2获得了可比较的结果，除了ATSR-2 SST与AATSR相比约为0.1K暖。
An initial validation of the Along Track Scanning Radiometer (ATSR) Reprocessing for Climate (ARC) retrievals of sea surface temperature (SST) is presented. ATSR-2 and Advanced ATSR (AATSR) SST estimates are compared to drifting buoy and moored buoy observations over the period 1995 to 2008. The primary ATSR estimates are of skin SST, whereas buoys measure SST below the surface. Adjustment is therefore made for the skin effect, for diurnal stratification and for differences in buoy–satellite observation time. With such adjustments, satellite-in situ differences are consistent between day and night within ~0.01K. Satellite-in situ differences are correlated with differences in observation time, because of the diurnal warming and cooling of the ocean. The data are used to verify the average behaviour of physical and empirical models of the warming/cooling rates. Systematic differences between adjusted AATSR and in-situ SSTs against latitude, total column water vapour (TCWV), and wind speed are less than 0.1K, for all except the most extreme cases (TCWV <5kgm–2, TCWV >60kgm–2). For all types of retrieval except the nadir-only two-channel (N2), regional biases are less than 0.1K for 80% of the ocean. Global comparison against drifting buoys shows night time dual-view two-channel (D2) SSTs are warm by 0.06±0.23K and dual-view three-channel (D3) SSTs are warm by 0.06±0.21K (day-time D2: 0.07±0.23K). Nadir-only results are N2: 0.03±0.33K and N3: 0.03±0.19K showing the improved inter-algorithm consistency to ~0.02K. This represents a marked improvement from the existing operational retrieval algorithms for which inter-algorithm inconsistency is >0.5K. Comparison against tropical moored buoys, which are more accurate than drifting buoys, gives lower error estimates (N3: 0.02±0.13K, D2: 0.03±0.18K). Comparable results are obtained for ATSR-2, except that the ATSR-2 SSTs are around 0.1K warm compared to AATSR.