摘要
为实现高海拔区域湖泊蓄水量监测与预测,提出了一种融合ICESat/GLAS与Landsat数据监测并预测湖泊蓄水量季节性变化的方法。利用随机森林对GLAS激光脚点自动分类,解决湖面非水体与水体目标分类难的问题;结合水体指数和大津法自动分割Landsat数据中的水体,降低不规则边界的影响;联合GLAS和Landsat数据,构建高时间分辨率的水位、面积与蓄水量时序列变化关系,解决两类数据观测信息缺失及观测时间不匹配的问题;并利用LSTM模型,预测蓄水量变化。以西藏纳木错湖为例,对上述方法进行验证,结果显示2003-2009年纳木错湖水位上升1.15 m,面积增加25.58 km2,蓄水量增加2.30 km^(3);预测2010-2016年蓄水量增加1.16 km^(3),与实际情况相符。结果表明,该方法可为生态环境的动态监测提供数据支撑。
To address the challenges of monitoring water storage in high-altitude lakes,this study fused ICESat/GLAS and Landsat data to monitor and predict the seasonal variation of lake water storage.Based on Random Forest,GLAS laser foot points were classified into useful and useless foot points,which contributed to distinguish non-aquatic and aquatic objects on lake surfaces.The water bodies were segmented by combining water indices and Otsu's method in Landsat data,which could mitigate the impact of irregular boundaries.To resolve the issues of missing observations and temporal mismatches,a high-temporal-resolution relationship among water level,water surface area,and water storage was established by fusing GLAS and Landsat data.Lastly,the water storage was predicted by LSTM model.Using Nam Co Lake in Tibet as a case study,the feasibility of proposed methodology was tested.During 2003-2009,the water level,the water surface area,and water storage raised every year,which were 1.15 m,25.58 km2 and 2.30 km^(3) respectively.The predicted result of water storage increment was 1.16 km^(3) from 2010-2016,which was accorded with the actual situation.Collectively,the proposed methodology could provide crucial data support for dynamic monitoring of ecological environments.
作者
王振华
刘存锟
张熙
王凤香
刘向锋
WANG Zhenhua;LIU Cunkun;ZHANG Xi;WANG Fengxiang;LIU Xiangfeng(College of Information,Shanghai Ocean University,Shanghai 201306,China;Land Spatial Data and Remote Sensing Technology Institute of Shandong Province,Jinan 250002,China;Key Laboratory of Space Active Opto-Electronics Technology,Shanghai Institute of Technical Physics,Chinese Academy of Sciences,Shanghai 200083,China)
出处
《海洋测绘》
CSCD
北大核心
2024年第4期74-78,共5页
Hydrographic Surveying and Charting
基金
上海市科技创新行动计划自然科学基金面上项目(23ZR1473200)
上海市海洋局科研项目(沪海科2023-01)
中国科学院空间主动光电技术重点实验室基金项目(CXJJ-22S019)。
关键词
激光脚点分类
蓄水量预测
水位计算
水体分类
季节性监测
classification of laser foot points
water storage prediction
water level calculation
seasonal monitoring
water body segmentation
