Surface water temperature observations of large lakes by optimal estimation
Surface water temperature observations of large lakes by optimal estimation
复制标题
DOI：
10.5589/m12-010
复制
发表时间：
2012-02-01
影响因子：
2.6
通讯作者：
Merchant, Christopher J.
中科院分区：
文献类型：
--
作者：
MacCallum, Stuart N.;Merchant, Christopher J.
开发了最佳估计 (OE) 和概率云筛选，以通过一系列（高级）沿轨扫描辐射计 (ATSR) 提供湖面水温 (LSWT) 估计。通过对观测到的辐射率进行正向建模，可以解释海拔、盐度和大气条件等物理特性的变化。因此，开发的 OE 检索方案是通用的（即适用于所有湖泊）。从 1995 年到 2009 年，通过 ATSR-2 和 AATSR 图像获得了地球上 258 个最大湖泊的 LSWT。与几个湖泊的原位观测结果进行比较，得出卫星原位白天观测值差异为 -0.2 +/- 0.7 K，夜间观测值差异为 -0.1 +/- 0.5 K（平均值 +/- 标准差）。相比之下，之前基于操作海面温度算法的方法的白天观测值为 -0.05 +/- 0.8 K，夜间观测值为 -0.1 +/- 0.9 K。新方法还增加了覆盖范围（减少将晴空误分类为云），并在使用不同通道视图组合的检索之间表现出更大的一致性。将经验正交函数 (EOF) 技术应用于 LSWT 检索（其中包含由于云层覆盖而产生的间隙），以重建 LSWT 的空间和时间完整时间序列。新的 LSWT 观测和基于 EOF 的重建有利于数值天气预报、湖泊模型验证，并提高我们对全球湖泊气候学的了解。观察结果和重建结果均可从 http://hdl.handle.net/10283/88 公开获取。
Optimal estimation (OE) and probabilistic cloud screening were developed to provide lake surface water temperature (LSWT) estimates from the series of (advanced) along-track scanning radiometers (ATSRs). Variations in physical properties such as elevation, salinity, and atmospheric conditions are accounted for through the forward modelling of observed radiances. Therefore, the OE retrieval scheme developed is generic (i.e., applicable to all lakes). LSWTs were obtained for 258 of Earth's largest lakes from ATSR-2 and AATSR imagery from 1995 to 2009. Comparison to in situ observations from several lakes yields satellite in situ differences of -0.2 +/- 0.7 K for daytime and -0.1 +/- 0.5 K for nighttime observations (mean +/- standard deviation). This compares with -0.05 +/- 0.8 K for daytime and -0.1 +/- 0.9 K for nighttime observations for previous methods based on operational sea surface temperature algorithms. The new approach also increases coverage (reducing misclassification of clear sky as cloud) and exhibits greater consistency between retrievals using different channel-view combinations. Empirical orthogonal function (EOF) techniques were applied to the LSWT retrievals (which contain gaps due to cloud cover) to reconstruct spatially and temporally complete time series of LSWT. The new LSWT observations and the EOF-based reconstructions offer benefits to numerical weather prediction, lake model validation, and improve our knowledge of the climatology of lakes globally. Both observations and reconstructions are publically available from http://hdl.handle.net/10283/88.