Development of Weighted Curve Number Approach and Watershed Quality Index Tool Using AnnAGNPS: an Obed River Watershed Case Study

Alamdari, Nasrin

ABSTRACT: Land use and land cover (LULC) change have significant impacts on watershed responses, such as changes in water quality and hydrology. Understanding these positive or negative impacts on water quality and hydrology is important for watershed planning, management and restoration. To evaluate the effects of LULC change, several computer models have been developed to quantify watershed responses, and one such model is the Annualized Agricultural Non-Point Source (AnnAGNPS) model. Though AnnAGNPS has been widely used in recent decades, the model simplifies the computations data to derive curve numbers (CN), by using a dominant soil type and land use, which is then used to determine runoff volumes and nutrient loads in a watershed. In this study, the limitations of this simplification are highlighted and an alternate approach called a Weighted CN method is proposed to overcome those limitations. In addition, a Watershed Quality Index (WQI) tool is developed for evaluating the effects of LULC change on the watershed quality at multiple hydrologic scales. Both WQI and Weighted CN method are demonstrated on the Obed River watershed with a drainage area of 520 mi2, located in the Cumberland Plateau region in Tennessee. Comparing simulated runoff to observed flows, results indicate that the weighted CN method is somewhat able to better represent surface runoff by providing detailed information of soil and land use data. Weighted average CN method showed improvement in R2 by 1.4% and NSE by 1.6% compared to dominant CN method. However, the improvements in NSE and R2 were not statistically significant which can be due to the fact that Obed River watershed is primarily forested. In addition, weighted CN and dominant CN methods were compared at a smaller scale. The results showed improvement in R2 and NSE by 2.1% and 1.9%, respectively. Though the improvements were still not significant, the model performance is improved greater in smaller scales. This is mainly because of the presence of variety of land uses including urban and forest within the smaller scale watershed. Moreover, WQI tool is used for evaluating the effects of LULC change on water quality. It is found that there is a strong positive relationship between LULC change and spatial variability of river water quality. Our results showed that the LULC change from deciduous forest to urban area led to increasing of NPS pollution and the deciduous forest contributed the least to soil erosion, sediment transport and nutrient load considering its change in area. In addition, the WQI has been applied on smaller scale to evaluate the applicability of the tool to show the changes in water quality and it is found that the distance over which land use change scenario influences on water quality depends on the size of the sub-watersheds and by changing the small portion of the sub-watersheds, smaller scales are more sensitive than larger scales. Application of the weighted CN method might provide detailed information on each AnnAGNPS cell if applied on watersheds with variety of land uses and could efficiently help planners and decision makers to better understand the limitations and simplifications in watershed models. Moreover, it reveals that WQI is a very useful and reliable tool to study the impacts of LULC change on water quality and stream hydrology. This tool can provide guidelines for resource managers, local planners, and policy makers in evaluating alternate land management practices and implementing appropriate strategies.