Inversion of Taipu River water quality parameters by UAV hyperspectral imaging technology

doi:10.13474/j.cnki.11-2246.2024.0406

Abstract

Abstract: In order to quickly and comprehensively grasp the water environment status of Taipu River,this study takes part of Taipu River as the test area,builds mathematical statistical models of three water quality parameters,including suspended matter concentration,turbidity and transparency,based on UAV hyperspectral data and measured water quality data,and then carries out accuracy evaluation. The model with the highest evaluation accuracy is selected to invert and analyze the water quality of Taipu River. The results show that: ①The hyperspectral band with the highest correlation with suspended matter concentration,turbidity and transparency is about 880～900 nm,and the correlation trend of suspended matter concentration and turbidity is consistent with the reflectance of each hyperspectral band,while the correlation trend of absolute value between transparency and reflectance of each hyperspectral band is consistent with the previous two. ②In the inversion model of suspended matter concentration,the ratio index model has the best effect (test set R²=0.91,validation set RMSE=27.04 mg/L,validation set MAPE=47.04%). And in the turbidity inversion model,the ratio index model has the best effect (test set R²=0.92,validation set RMSE=16.50 NTU,validation set MAPE=15.24%). The normalized index model has the best inversion effect among transparency inversion models (test set R²=0.85,validation set RMSE=2.43 cm,validation set MAPE=8.38%). By comparing the inversion effects of the three water quality parameters,we can see that the inversion effect of transparency is the best,followed by turbidity,and finally the concentration of suspended matter. ③The inverted suspended matter concentration of Taipu River is generally low,while the eastern part is high. Turbidity is generally high in the eastern part,followed by the western part,and lowest in the central part. Transparency is generally high in the western and central part,and lowest in the eastern part.

Key words: UAV, hyperspectral, ASD, Taipu River, water quality inversion

CLC Number:

P237

LIANG Wenguang, WU Yongfeng, SHI Yifan, WANG Dongmei, WANG Yihong. Inversion of Taipu River water quality parameters by UAV hyperspectral imaging technology[J]. Bulletin of Surveying and Mapping, 2024, 0(4): 29-34.

References

[1] 李雪清,郑航,刘悦忆,等. 基于多源数据机器学习的区域水质预测方法研究[J]. 水利水电技术(中英文),2021,52(11): 152-163.
[2] 彭令,梅军军,王娜,等. 工矿业城市区域水质参数高光谱定量反演[J]. 光谱学与光谱分析,2019,39(9): 2922-2928.
[3] 王思梦,秦伯强. 湖泊水质参数遥感监测研究进展[J]. 环境科学,2023,44(3): 1228-1243.
[4] 林剑远,张长兴. 航空高光谱遥感反演城市河网水质参数[J]. 遥感信息,2019,34(2): 23-29.
[5] 张兵,申茜,李俊生,等. 太湖水体3种典型水质参数的高光谱遥感反演[J]. 湖泊科学,2009,21(2): 182-192.
[6] 赵松. 基于多源遥感数据的邯郸市滏阳河水质参数反演[D]. 邯郸: 河北工程大学,2021.
[7] 黄宇,陈兴海,刘业林,等. 基于无人机高光谱成像技术的河湖水质参数反演[J]. 人民长江,2020,51(3): 205-212.
[8] 刘梅,马启良,原居林,等. 基于无人机高光谱遥感技术对内陆养殖池塘水质监测的研究[J]. 海洋与湖沼,2022,53(1): 195-205.
[9] 杨振,卢小平,武永斌,等. 无人机高光谱遥感的水质参数反演与模型构建[J]. 测绘科学,2020,45(9): 60-64.
[10] 臧传凯,沈芳,杨正东. 基于无人机高光谱遥感的河湖水环境探测[J]. 自然资源遥感,2021,33(3): 45-53.
[11] 林荷娟,甘月云,刘敏. 太浦河锑浓度异常预警指标研究[J]. 水文,2021,41(1): 95-101.
[12] 罗鹏程. 长三角一体化示范区太浦河的重金属多介质分布特征、风险评价及来源解析[D]. 上海: 上海师范大学,2022.
[13] 张桠婷,张馨用. 水污染的跨界协同治理机制研究: 以太浦河为例[J]. 国土资源情报,2021(5): 51-56.
[14] 周刚. 微塑料在淀山湖及太浦河流域的污染负荷、赋存特征和生态风险评价[D]. 上海: 东华大学,2022.
[15] 唐军武,田国良,汪小勇,等. 水体光谱测量与分析Ⅰ: 水面以上测量法[J]. 遥感学报,2004,8(1): 37-44.