基于多源卫星数据的小型水体蓝藻水华联合监测——以天津于桥水库为例

Remote monitoring of cyanobacterial blooms using multi-source satellite data: A case of Yuqiao Reservoir,Tianjin

近年来水体富营养化呈扩张趋势,蓝藻水华不仅在太湖等大型湖泊频发,水面面积较小的天津于桥水库等也形势严峻,亟需加强卫星遥感监测.但是,以往在太湖等业务化使用非常成功的MODIS等卫星数据(约500 m),由于空间分辨率较低,难以满足小型水体的监测要求;而Landsat-8等空间分辨率较高的卫星数据(>30 m),通常重返周期较长,无法满足水华高频监测需求.本文以天津市于桥水库(面积约80 km2)为研究区,针对常用的卫星数据,从空间、时间、光谱范围和数据可获取性共4个方面,评价不同卫星数据蓝藻水华监测能力和算法,同时对不同卫星监测结果一致性进行评估.结果表明:(1)筛选出国产HJ-1A\B CCD、GF-1 WFV和美国Landsat-8 OLI这3种卫星波段合适,空间分辨率较高,适用于桥水库蓝藻水华监测,但考虑到其重返周期较长,建议多星联合观测;(2)各个卫星监测结果与卫星影像目视解译结果基本一致,均方根误差和相对误差均分别控制在0.78 km2和4.9%以内;(3)不同卫星监测结果一致性良好,一致性精度达到99.5%;(4)根据历史影像结果,发现于桥水库2016年水质开始呈富营养化,藻华现象在夏、秋两季最为严重.研究表明,针对小型水面水体蓝藻水华监测,利用较高分辨率数据联合监测,是一种有效的替代策略,今后可在更多小型水域推广.

In recent years,there is a growing eutrophication of surface freshwaters. In many cases,this has resulted in an increase in the occurrence and persistence of cyanobacterial blooms,not only large lakes,but also in small reservoirs. There is an urgent need for new and better monitoring approaches. Satellite based remote sensing is one important and increasing accessible tool. However,the application of MODIS and other satellite data with low spatial resolution（ about 500 m） but high temporal frequency was limited to larger lakes. While with higher spatial resolution satellite data（ 30 m）,such as Landsat-8,have longer return periods,making them less useful for blooms monitoring. This study explores the usage of the Chinese HJ-1 A ／ B CCD and GF-1 WFV sensors,together with the United States Landsat-8 OLI and other high-resolution satellite data for joint observation of cyanobacteria blooms in a small reservoir（ Yuqiao Reservoir,Tianjin）. An algal extraction was developed for each satellite sensor,and the output of different satellite sensors was evaluated. The results indicate that：（ 1） the combined satellite monitoring provided consistent results when compared to the visual interpretation of multiband satellite images with a mean square root error and relative error of 0.78 km2 and 4.9%;（ 2） different satellite sensors provided consistent results,with an accuracy of 99.5%;（ 3） According to the research in 2016,the water quality of Yuqiao Reservoir was eutrophic,and algal bloom was the most serious in the two seasons of summer and autumn. The use of combined observations of high-resolution data was an effective,alternative strategy for monitoring cyanobacterial blooms in small waterbodies,opening up new possibilities to improve monitoring of these important freshwater environments.