| 摘要: |
| 湖泊透明度是衡量水质变化和湖泊水生生态系统健康重要的指标。一般通过塞氏盘（Secchi Disk）现场观测获得湖泊透明度信息，其变化主要是由水体中非藻类颗粒物、藻类和黄色物质（CDOM）浓度共同决定，因此可以通过光学遥感进行反演。本研究基于2004-2009年8次查干湖实地采样获得的132个透明度数据，通过分析其同步Landsat天顶反射率产品数据(TOA)不同波段组合与透明度之间的相关性，发现以B3/B1比值和1-3波段均值AV(B3~B1)组合建立的透明度反演模型精度最高(R2 = 0.88)，误差最小(RMSE = 3.49, MAPE = 14.37%)，进而基于此模型反演了1984-2019年查干湖以及周边湖泊透明度。结果表明：查干湖透明度范围在1.0 - 63.2 cm之间，年均值为16.5 ± 39.2 cm；查干湖透明度年均值变化分为二个阶段: 1984-2001年间透明度波动范围较大，没有明显趋势；2002-2019年期间呈上升趋势（2.3 cm/10yr）；通过分析查干湖透明度与风速、降雨、水面积和植被指数的关系，发现其分别与透明度具有一定的相关性，R2分别为0.31, 0.11和0.27。为查干湖流域生态治理提供科学依据。 |
| 关键词: 查干湖 透明度 Landsat 遥感 |
| DOI： |
| 分类号: |
| 基金项目:吉林省科技发展计划项目 (20220203024SF)，国家自然科学基金（42171374，42371390） |
|
| Remote estimation of water clarity variation in the past 36 years (1984-2019) for Chagan Lake |
|
Zhao Fangrui,Wang Qiang,Mu Chunsheng,Liu Ge,Wen Zhidan,Tao Hui,Song Kaishan
|
|
Northeast Institute of Geography and Agroecology, Chinese Academy of Sciences
|
| Abstract: |
| water clarity is a crucial proxy for evaluation eutrophication of lakes or reservoirs. Traditionally, water clarity is determined with Secchi disc depth (SDD) in the field, which is time and labor consuming and not suitable for water clarity assessment at large scale. Water clarity is mainly governed by non-algal particulate matter, algal abundance and colored dissolved organic matter (CDOM) in water column, which can be monitored with optical remote sensing. In this study, eight cruise surveys were carried out over the Chagan Lake during 2004-2009, and in situ measured SDD were determined. In order to determine the best band or band combinations, correlation analysis between in situ measured SDD and Landsat calibrated top-of-atmosphere (TOA) reflectance were carried out. Based on in situ measured SDD and TOA pairs, we developed an algorithm based on the B3/B1 ratio and average of B1-B3, e.g., AV (B3, B1) to estimate SDD in Chagan Lake. A high model calibration accuracy was achieved (R2 = 0.88), with lower RMSE (3.49), and MAPE (14.37%) for model validation. The algorithm was applied to Landsat images for deriving SDD distribution maps from 1984 to 2019. The Landsat-derived results indicate that the SDD in Chagan ranges 1.0-63.2 cm and an annual mean of 16.5 ± 39.2 cm, a significant increasing trend during 1984-2019 is encountered with an increasing trend of 2.3 cm/10 years. The annual mean values of SDD variation dynamics in Chagan Lake underwent two phases. From 1984 to 2001, the SDD showed no obvious trend, and then it demonstrated a significant increasing trend. Our analysis indicated that the SDD in Chagan Lake has some connection with wind speed, precipitation, and vegetation coverage (NDVI), which explained the variation of 0.31, 0.11 and 0.27 of the increased SDD in Chagan Lake, respectively. |
| Key words: Chagan Lake Secchi disc depth Landsat remote sensing |
