Estimation of Monthly Potential Nitrogen Load from Agricultural and Forest Watersheds Using Quickbird Remote Sensing Imagery

Eutrophication of coastal oceans as a result of the delivery of excess nutrients by rivers is a critical environmental issue on local, regional, and global scale. Japanese agriculture has created high nitrogen surpluses in agricultural lands due to the increasing rate of chemical fertilizer application and the increase of livestock. In this study, to estimate the nitrogen load exported to Koise River, a major river in Lake Kasumigaura, from its catchment area, extensive field survey and water quality experiments have been conducted from 2007 to 2008. Remote sensing data of QuickBird imagery was applied to the dataset on water quality during two years (Year 2007 for model calibration and Year 2008 for model validation) at 11 streams. Water quality indicator of total nitrogen (TN) were examined and an efficient approach, which estimates monthly pollutant loads directly from satellite imagery and ground observed hydrological data, was investigated. Then the pollutant loads of TN were estimated and prediction results show that the proposed simulation technique can be used to predict the monthly TN pollutant loads of river basins in Japan. Seasonal variation of pollutant loads from 2007 to 2008 were also analysed from different sources according to land covers. This study presents usefulness of remote sensing technique for evaluation and interpretation of water quality datasets and apportionment of pollution sources with a view to get better information about the water quality for more effective management of water resources in river basins.