Exploiting weather forecast data for cloud detection
Exploiting weather forecast data for cloud detection
复制标题
利用天气预报数据进行云检测
DOI：
10.1016/j.rse.2005.11.001
复制
发表时间：
2009
影响因子：
13.5
通讯作者：
S. Mackie
中科院分区：
文献类型：
--
作者：
S. Mackie
准确、快速地检测卫星图像中的云有许多应用，例如数值天气预报（NWP）和大气和地球表面温度的气候研究。大多数云检测的实用技术依赖于云和晴空观测值之间的差异，这些差异在空间和时间上或多或少是恒定的。事实上，不同的云有不同的光谱特性，不同的云类型或多或少可能在不同的地方和不同的时间，取决于大气条件和地球表面的属性。对晴空的观测在空间和时间上也各不相同，这取决于大气和地面条件，以及是否存在气溶胶粒子。在这个项目中采用的贝叶斯方法允许像素特定的物理信息（例如从数值预报）被用来预测晴空的像素特定的观测。然后计算每个像素包含云观测的基于物理的、空间和时间特定的概率。该方法的优点在于，与通常由操作技术产生的二进制结果相比，从概率结果识别模糊分类的像素是直接的。该项目开发并验证了云检测的贝叶斯方法，并扩展了其适用的应用范围，在每种情况下都达到了与操作方法相匹配或超过操作方法的技能分数。高温度梯度可以使海洋锋面周围的晴空观测，特别是在热波长上，看起来与云观测相似。为了增加这种模糊云检测结果的潜在来源，选择了AATSR传感器获得的图像区域，该区域被注意到包含一些海洋锋。区域内的像素根据其光谱特性进行聚类，目的是分离对应于不同海洋热状况的像素。然后使用贝叶斯云检测技术处理每个聚类中像素的平均光谱特性，并将所得到的后验概率分配给各个像素。几种聚类方法，我们重新调查，和最合适的，它允许像素与多个集群，与归一化向量的“成员强度”，被用来进行案例研究。当包括聚类时，最终计算的清晰概率的分布变得明显更加双峰，这表明模糊的分类更少，但代价是错过了如此多的单个像素云。虽然进一步的调查可以提供一个解决方案，聚类方法的计算费用使这不切实际的，包括在本项目的工作。这种新的贝叶斯云检测方法已经被这个项目成功地开发出来，并在公共许可证下发布。该项目最初是作为一种工具设计的，以帮助从夜间图像检索海表面温度，
Accurate, fast detection of clouds in satellite imagery has many applications, f r example Numerical Weather Prediction (NWP) and climate studies of both the atmosphere an d of the Earth’s surface temperature. Most operational techniques for cloud detection r ely on the differences between observations of cloud and of clear-sky being more or less cons tant in space and in time. In reality, this is not the case different clouds have different spectral p roperties, and different cloud types are more or less likely in different places and at different times, dep ending on atmospheric conditions and on the Earth’s surface properties. Observations of clea r sky also vary in space and time, depending on atmospheric and surface conditions, and on the pre sence or absence of aerosol particles. The Bayesian approach adopted in this project allows pixel-specific physical information (for example from NWP) to be used to predict pixel-specific obs ervations of clear sky. A physically-based, spatiallyand temporally-specific probability that each pixel contains a cloud observation is then calculated. An advantage of this approach is tha t identification of ambiguously classed pixels from a probabilistic result is straightforward, in contrast to the binary result generally produced by operational techniques. This project ha s developed and validated the Bayesian approach to cloud detection, and has extended the range of ap plications for which it is suitable, achieving skills scores that match or exceed those achieved by op erati nal methods in every case. High temperature gradients can make observations of clear sky around oc ean fronts, particularly at thermal wavelengths, appear similar to cloud observations. To add ress this potential source of ambiguous cloud detection results, a region of imagery acquired by the AATSR sensor which was noted to contain some ocean fronts, was selected. Pixels in the reg ion were clustered according to their spectral properties with the aim of separating pixels that c orrespond to different thermal regimes of the ocean. The mean spectral properties of pixels in eac h cluster were then processed using the Bayesian cloud detection technique and the resulting p osterior probability of clear then assigned to individual pixels. Several clustering methods we re investigated, and the most appropriate, which allowed pixels to be associated with multiple clusters, with a normalized vector of ‘membership strengths’, was used to conduct a case study. The distribution of final calculated probabilities of clear became markedly more bimodal when c lustering was included, indicating fewer ambiguous classifications, but at the cost of so me single pixel clouds being missed. While further investigations could provide a solution to this , the computational expense of the clustering method made this impractical to include in the work of this project. This new Bayesian approach to cloud detection has been successfully de veloped by this project to a point where it has been released under public license. Initially designed as a tool to aid retrieval of sea surface temperature from night-time imagery, this projec t has extended