Landsat 8衛星影像應用於桃園地區埤塘總磷水質監測

Total Phosphorus for Water Quality Monitoring of Irrigation Ponds Using Landsat 8 Imagery in Taoyuan, Taiwan

蘇柏聞(Po-Wen Su)

指導教授 ： 駱尚廉

國立臺灣大學/工學院/環境工程學研究所/博士(2022年)

https://doi.org/10.6342/NTU202201618

摘要

氣候變遷調適為因應全球暖化的對策，旨在降低社會和生物系統的脆弱性，表示透過預測氣候變遷帶來的不利影響，隨即採取合適的行動預防或減少損害。優養化為自然水體必然會面臨的環境問題，而造成水質變化的因子很多，且不同因子對水體的影響程度尚不清楚，尤其是與氣候變遷有關的因子。以時間和空間高解析度監測水質之參數至關重要，其原因為減輕污染的費用通常較早期預防及干預措施昂貴。因此，本研究將Landsat 8衛星影像應用於埤塘監測總磷濃度。
在桃園地區埤塘監測部分，本研究利用Landsat 8衛星影像，以多元迴歸模型建立衛星波段反射率與總磷濃度之關係。大氣校正採用ACOLITE軟體，並選取短紅外光波段對衛星影像進行處理。本研究之方法通過p值和變異數膨脹因素閾值對參數作正向選擇，預測結果由多元迴歸分析推導得出方程式，其R2值為0.67，且在近紅外光波段最具相關性，即Landsat 8衛星的波段5。於研究範圍內繞37B池、繞32B池和社子1號池三口埤塘中，僅繞32B池和社子1號池取得些許的像元，其原因為部分埤塘的水表面被太陽能光電板覆蓋所致。總磷濃度之結果顯示，將本研究之模型套用至整個桃園地區的埤塘，使用衛星遙測評估埤塘水質具有可行性，而推導之關係方程式可能適用於擴展埤塘的時間和空間有效水質數據。

關鍵字

優養化 ； 埤塘 ； 多元回歸分析 ； 遙感探測 ； 水庫 ； 水質

並列摘要

Climate change adaptation is a response to global warming that aims to reduce the vulnerability of social and biological systems. This means anticipating the adverse impacts of climate change and then taking appropriate actions to prevent or reduce damage. Eutrophication is an inevitable environmental problem for natural water bodies. However, there are many factors that cause water quality changes, and the influence of different factors on water bodies is not clear, especially for those factors related to climate change. Monitoring water quality parameters at high spatial and temporal resolution is essential because pollution mitigation is usually more expensive than early prevention and intervention. Therefore, this study applied Landsat 8 satellite imagery to monitor the total phosphorus (TP) concentration in irrigation ponds.
For irrigation pond monitoring in the Taoyuan area, this study used Landsat 8 satellite imagery to establish the relationship between satellite reflectance and the concentration of TP using a multiple regression model. ACOLITE software was used for atmospheric correction, and shortwave infrared (SWIR) bands were selected to process the satellite imagery. P values and variation inflation factor (VIF) thresholds were used for forward selection of variables, and prediction variables in the multiple regression equation were derived. The derived equation yielded a coefficient of determination (R2) of 0.67. The near-infrared band (band 5 of Landsat 8) was found to be the most significant band. The imagery retrieved for two of the three studied ponds, Shetzu No. 1 and Jao No. 37B, contained only a few pixels of the irrigation ponds because parts of the surfaces of the irrigation ponds are covered by floating photovoltaic plants. The TP concentrations results show that it is feasible to apply the model to the whole area of Taoyuan and use satellite remote sensing to assess water quality, and the derived equation could be used to temporally and spatially extend the available water quality data of these irrigation ponds.

並列關鍵字

Eutrophication ； Irrigation pond ； Multiple regression analysis ； Remote sensing ； Reservoir ； Water quality

參考文獻

Anderegg, W. R., Prall, J. W., Harold, J., Schneider, S. H. (2010). Expert credibility in climate change. Proceedings of the National Academy of Sciences, 107(27), 12107-12109.

Arhonditsis, G. B., Brett, M. T., DeGasperi, C. L., Schindler, D. E. (2004). Effects of climatic variability on the thermal properties of Lake Washington. Limnology and oceanography, 49(1), 256-270.

Austin, R., Petzold, T. J. (1981). The determination of the diffuse attenuation coefficient of sea water using the Coastal Zone Color Scanner Oceanography from space (pp. 239-256): Springer.

Austin, R. W., Petzold, T. J. (1986). Spectral dependence of the diffuse attenuation coefficient of light in ocean waters. Optical Engineering, 25(3), 253471.

Belokon, K., Emmons, M., Fowler, W., Gilson, B., Hernandez, G., Johnson, A., . . . Tullos, E. (1997). Multispectral Imagery Reference Guide: LOGICON Geodynamics. Inc. Fairfax, Virginia.