草型湖泊水质遥感监测中水生植物会造成“水体—水生植物”混合像元问题，针对因混合像元导致草型湖泊水生植物覆盖区域水质难以直接利用遥感监测的问题，本文以草型湖泊微山湖为研究对象，提出定性和定量相结合的总悬浮物浓度和浊度分区监测方法，实现微山湖水体总悬浮物浓度和浊度的时空变化监测。基于获取的2014年7月—2015年6月覆盖微山湖的多期高分一号（GF-1） WFV和HJ-1A/1B CCD影像，利用归一化水体指数将微山湖区分为水生植物覆盖区和水体区。针对水生植物覆盖区，利用时序MODIS NDVI数据获取微山湖主要水生植物的时谱曲线，识别不同水生植物的物候特征；基于不同物候期内的水生植物对总悬浮物浓度和浊度的指示作用，对微山湖水生植物覆盖区水体总悬浮物浓度和浊度进行定性监测。针对水体区，分别构建水体总悬浮物浓度和浊度的单波段/波段比值模型和偏最小二乘模型，定量反演微山湖水体区总悬浮物浓度和浊度。研究结果表明，微山湖中水生植物以光叶眼子菜、穗花狐尾藻和菹草等沉水植物为主，其中光叶眼子菜/穗花狐尾藻和菹草的空间分布和物候特征存在明显差异，不同水生植物在不同物候期内对水质具有不同的指示作用；微山湖水体总悬浮物浓度和浊度具有显著的空间变异性，基于定性和定量相结合的方法可以有效监测微山湖水体总悬浮物浓度和浊度的时空变化规律。本文提出的定性和定量相结合的监测方法为草型湖泊水质监测的业务化应用提供了新思路。

Remote sensing has been recognized as an effective tool for monitoring water quality in inland waters, especially in algal lakes. However, water quality in macrophytic lakes where aquatic vegetation grows because of mixed pixels is difficult to retrieve. In this study, a classification retrieval method and image data acquired between July 2014 and June 2015 using GF-1 WFV and HJ-1A/1B CCD sensors were proposed to monitor Total Suspended Matter (TSM) and turbidity in a macrophytic lake, namely, the Weishan Lake, considering aquatic vegetation phenology.
In the classification retrieval method, the Weishan Lake was divided into water overlying aquatic vegetation and water area using normalized difference water index. First, a qualitative method was proposed to retrieve TSM concentration and turbidity in water overlying aquatic vegetation considering aquatic vegetation phenology. In the qualitative method, time series MODIS NDVI data were used to obtain the time-spectrum curves of aquatic vegetation in the Weishan Lake for identifying different aquatic vegetation phenological periods. The characteristics of aquatic vegetation in different phenological periods were used to estimate the TSM concentration and turbidity. Second, single band, band ratio, and partial least squares models were applied to retrieve the TSM concentration and turbidity in the water area. Finally, the temporal and spatial variations in the TSM concentration and turbidity along with aquatic vegetation in the entire Weishan Lake were analyzed.
Results showed that the three main aquatic vegetation areas in the Weishan Lake, namely, Potamogeton lucens, Myriophyllum spicatum, and Potamogeton crispus, have different phenological periods. P. lucens and M. spicatum started to grow in spring, reached their peak at the end of summer, and gradually died in autumn. P. crispus started to grow in April, reached its peak at the end of spring, and quickly died in the early summer. Different aquatic vegetation areas had various indicators of TSM concentration and turbidity in different phenological periods. The TSM concentration was less than 15 mg/L in the water overlying three aquatic vegetation areas in the growth stage. The turbidity in the water overlying P. lucens, M. spicatum, and P. crispus in the growth stage were less than 30 and 15 NTU. The death of P. lucens and M. spicatum did not result in the deterioration of water quality because of their extensive death time. However, the death of P. crispus within a short time resulted in serious deteriorations of water quality. The TSM concentration and turbidity were 15-145 mg/L and 30-140 NTU, correspondingly, in the water overlying P. crispus in the death stage. The TSM concentration and turbidity in the Weishan Lake had a significant temporal-spatial variability. The TSM concentration and turbidity in the Southwestern Weishan Lake where P. lucens and M. spicatum grew had low levels in the seasons. By contrast, the TSM concentration and turbidity in the Northeastern Weishan Lake with P. crispus growth had low levels in spring. However, they became large in summer because of the quick death of P. crispus and gradually decreased in autumn and winter.
A classification retrieval method coupled with quantitative monitoring in the water overlying aquatic vegetation and qualitative monitoring in the water area was proposed to monitor the TSM concentration and turbidity in the Weishan Lake considering aquatic vegetation phenology. This method effectively monitored the temporal and spatial variations in water quality in the entire Weishan Lake. The findings indicated that the proposed method can be used to monitor the water quality in other macrophytic lakes.