Prediction of Nitrate Concentration in Stream Water Based on Watershed Land Use and Stream Flow Rate

Abstract

Non-point source nitrate loading to stream water can be affected by land use, soil type, slope of discharging watershed, water flow rate, and storm events. This study considers the effects of land use type, water flow rate and seasonal and regional variation in order to understand and predict transport of nitrate in stream water discharging from watersheds along Virginia's Eastern Shore. A simple regression analysis was performed to check the relationship between nitrate concentration and land use type. Fertilized land was found to be the strongest contributor to nitrate concentration in stream water. Developed land was also a contributor but with a lower weighting factor. Forested land acted as a significant inhibitor. A predictive model was generated using multiple regression analysis. Besides land use type, nitrate concentration was also significantly influenced by stream flow rate. Other variables, including sampled area, seasonal and regional variation, may be treated as background information, in which they may have some effect on nitrate concentration, but not significantly enough to be included in the predictive model. Therefore, stream nitrate concentration levels can be predicted along the Eastern Shore of Virginia based on land use type and stream flow rate. Predictions may be improved by incorporating other geomorphic or hydrologic information.