基于实测数据与遥感影像的鄱阳湖水体光学分类

鄱阳湖是我国最大的淡水湖泊,受季风气候影响其水体空间动态变化大,且广阔的水域内部差异也较大,因此湖泊水体光学分类对反演湖泊水质参数及监测湖泊水质有着重要意义。以鄱阳湖为研究区,根据实测的反射光谱数据形态特征将鄱阳湖的水体分为4类:特别浑浊、中等浑浊、轻度浑浊和清水区,并分别对每一类结果进行分析。考虑到实测光谱数据局限于湖区某些离散点的情况,不足以观测整个鄱阳湖区域内所有不同水体类型的空间分布和动态变化,从而将该方法利用于遥感影像以便观测整个湖区水体类型。在Landsat OLI遥感影像上任意选取采样点,根据其波谱形态建立基于斜率的分类算法,并应用决策树模型把鄱阳湖水体分为5类:特别浑浊、中等浑浊、轻度浑浊、清水区和特别清澈,影像的分类结果图与实地考察的情况相一致。把模型应用于其他时期的遥感影像进行鄱阳湖水体分类,对比影像的分类结果图表明:2002、2005和2009年鄱阳湖区分别出现3种、4种和4种不同的水体类型,且水体浑浊范围呈现出动态变化。研究表明水体光学类型分类可以更好的监测湖泊水质的时空变异性。

OPTICAL CLASSIFICATION OF POYANG LAKE WATERS BASED ON IN SITU MEASUREMENTS AND REMOTE SENSING IMAGES

Poyang Lake is the largest freshwater lake in China. Affected by the monsoon climate, the seasonal and inter-annual changes of its water area are obvious, and the remarkable differences exist in the internal water area. Optical classification of lake water is of great significance for retrieval of lake water quality parameters and monitoring of lake water quality. Taking Poyang Lake as the study area, in this paper, the Poyang Lake waters were divided into four categories based on morphological characteristics of the measured spectral reflectance data, i.e., Special turbidi, Medium turbidi, Mild turbid, Generally Clear. We analyzeddd the results of each water body type. Considering the measured spectral data are limited to some discrete points of the lake area, which are not enough for the observation of the spatial distribution and dynamic change of all different water body types in Poyang Lake area, we use remote sensing image to observe the type of whole lake water body. Then, we selected some sampling points on the Landsat 8 OLI remote sensing images , and established a classification algorithm based on the slope according to their spectral shape. The Poyang Lake water bodies were finally divided into five types using a decision tree model: Extremely Turbid, Medium Turbid, Mild Turbid, Generally Clear, and Particularly Clear. Results from image classification were found to be consistent with the field observations. When the model is applied to other periods of remote sensing images for classification of Poyang Lake water body, comparison of image classification results showed that Poyang Lake presented three, four and four different types of water bodies in 2002, 2005 and 2009, respectively, and the water turbidity range showed dynamic changes. Turbid water areas were mainly distributed in the channel of Poyang Lake's water flow into the Yangtze River and near the Songmen Mountain at the end of September or at the beginning of October (in 2002, 2005, 20013), and turbid water areas were mainly distributed near the river estuary in Poyang Lake at the early May in 2009, the channel of Poyang Lake's water flow into the Yangtze River and near the Songmen Mountain failed to present a special turbidity. It was found that optical classification of lake water can help monitor the spatial and temporal variability of Poyang Lake water optical quality.