基于PIE-Engine Studio的西藏典型湖泊总悬浮物反演  

Inversion of total suspended matter in typical lakes in Tibet based on PIE-Engine Studio

作　　者：赵爽 柳锦宝[1] 耿蔚 陶星宇 汪菀蔚 ZHAO Shuang;LIU Jinbao;GENG Wei;TAO Xingyu;WANG Wanwei(School of Resources and Environment,Chengdu University of Information Technology,Chengdu 610225,China;Sichuan Provincial Meteorological Disaster Defense Technology Center,Chengdu 610072,China;Pingshan County Meteorological Office,Yibin 645350,China)

机构地区：[1]成都信息工程大学资源环境学院,四川成都610225 [2]四川省气象灾害防御技术中心,四川成都610072 [3]四川省宜宾市屏山县气象局,四川宜宾645350

出　　处：《人民长江》2024年第9期85-92,99,共9页Yangtze River

基　　金：国家自然科学基金项目(42065008);四川省科技厅重点研发项目(2021YFS0328);四川省科技厅重点研发项目(2020YFG0146);中国气象科技研究院科技合作项目(2019QZKK0105,2019QZKK0304);气候小型建设项目(2020年山洪地质灾害防治气象保障工程之水体生态质量气象监测评估系统);中国气象局创新发展专项(CXFZ2021J055)。

摘　　要：总悬浮物浓度是水环境评价的重要参数之一,并且能在各个方面影响水体其他参数。青藏高原湖泊环境正处于大幅变化的时期,而其分布与动态变化与湖泊的环境息息相关。基于西藏典型湖泊29个采样点水质实测数据和LandSat 8 Collection 2遥感数据,使用PIE-Engine Studio平台,对比分析了基于单波段和多波段组合的多种总悬浮物浓度反演模型,选出了最优的总悬浮物浓度反演模型,并运用该模型分析了色林错、扎日南木错以及塔若错总悬浮物2013~2023年月度时空变化特征。分析结果表明:①总悬浮物对绿光、红光波段最为敏感,以波段反射率组合(B3+B4)、(B2/B3)(B2为蓝光波段,B3为绿光波段,B4为红光波段)为自变量,总悬浮物浓度为因变量的三波段模型为西藏典型湖泊总悬浮物浓度遥感反演最佳模型;②总悬浮物浓度年内变化总体存在季节规律,夏季浓度高,而春秋季低。在色林错,浓度在北部较高,高值夏季主要分布在西北岸,而秋冬季则在东北岸,受风向影响较大;在扎日南木错的西部较高;在塔若错,夏季在南部较高。3个湖泊均在11月时出现东部浓度重新升高的现象。③年降水量与各湖泊的悬浮物浓度相关性较高。年际变化中,2013~2023年,悬浮物浓度的年均值先是逐年升高,在2018年后随着年降水量的减少而逐渐下降或稳定。另外,色林错、扎日南木错、塔若错湖面面积分别增长了2.87%(约68.14 km^(2))、3.01%(约30.31 km^(2))、1.01%(约4.87 km^(2)),色林错持续扩张,扎日南木错和塔若错先快速扩张,而后萎缩,湖面面积与悬浮物浓度均值表现较同步,均是受降水量和径流的影响。研究结果可为高原湖泊水质评价提供理论依据,为水土流失、水资源保护提供科学参考。The concentration of total suspended matter(TSM)is one of the key parameters for water environment assessment and can significantly influence other parameters of the water body in various aspects.The lakes on the Qinghai-Tibet Plateau are currently undergoing significant environmental changes,and the distribution and dynamic variation of TSM concentration are closely related to the environmental conditions of these lakes.Based on the in-situ water quality data from^(2)9 sampling points in typical lakes in Tibet and Landsat 8 Collection 2 remote sensing data,a comparative analysis was conducted using the PIE-Engine Studio remote sensing cloud computing platform to evaluate various TSM concentration retrieval models that based on single-band and multi-band combinations.The optimal TSM concentration retrieval model was selected and subsequently used to analyze the spatiotemporal variations of TSM in Siling Co,Zhari Namco,and Taro Co from^(2)013 to 2023.The results indicated that:①TSM was most sensitive to the green and red light bands.A three-band model using band reflectance combinations(B3+B4),(B2/B3)(where B2 represented the blue band,B3 represented the green band,and B4 represented the red band)as independent variables and TSM concentration as the dependent variable,was the optimal model for remote sensing retrieval of TSM concentration in typical lakes of Tibet.②The annual variation in TSM concentration generally followed a seasonal pattern,with higher concentrations in summer and lower concentrations in spring and autumn.In Siling Co,the concentration was higher in the northern region,with high values mainly distributed along the northwest shore in summer and shifting to the northeast shore in autumn and winter,significantly influenced by wind direction.In Zhari Namco,higher concentrations were observed in the western part,while in Taro Co,the southern region showed higher concentrations during summer.All three lakes exhibited a phenomenon that the TSM concentrations rose again in the eastern regions in No

关 键 词：总悬浮物 PIE-Engine Studio 遥感反演 云计算 高原湖泊 西藏 

分 类 号：X87[环境科学与工程—环境工程] X524