Abstract
One of the greatest challenges in water-quality management is to understand the many complex interactions between landscape structure and nutrient transport. This study analysed the relationship between the inter-annual vegetation phenologies and seasonal change of water quality in the Choptank Basin, a typical agricultural tributary of the Chesapeake Bay (along the east coast of USA), using blended images from Landsat Thematic Mapper and Moderate Resolution Imaging Spectroradiometer. Cross-Correlogram Spectral Matching (CCSM) was applied to explain the correlation between seasonal nitrate discharge and phenologies of different crop types as well as forest. Nitrate discharges from forest-dominated watersheds are reduced as the forest greens up with an increasing percentage above the predefined enhanced vegetation index (EVI) threshold values. The crop-dominant watersheds have higher correlations than the forest-dominant watersheds between seasonal nitrate discharge and vegetation phenologies. The correlations between seasonal water quality and vegetation phenology in forest-dominant watershed have less variance than those of crop-dominant watersheds. The highest correlations were found between EVI and water quality at one or two months ‘later’, e.g. between the ‘earlier’ green-up and the ‘later’ nutrient discharge. This work could be applied to guide farmland planning to effectively control the non-point source pollution.
Acknowledgements
We acknowledge the contribution of Tom Fisher from the University of Maryland Center for Environmental Science for providing water-quality data and valuable comments.