Streamflow is typically divided into two components that are direct runoff and baseflow, it is required to analyze and estimate behaviors of those two flow components to understand watershed characteristics so that watershed management plan can be effective in pollutant reductions. Since pollutant load behaviors in a stream or river are variable by flow component behaviors, best management practices need to be applied in a watershed based on the pollutant load behaviors varying with flow components. Thus, baseflow behaviors were analyzed separating baseflow from streamflow data collected from fifteen streamflow gaging stations in the 4 major river watersheds which are the Han river, Nakdong river, Guem river, and Yeongsan·Somjin river watersheds. Moreover, precipitation trends throughout the 4 River Systems were investigated, thus daily precipitation data were collected from sixty-five locations. The Hank river watershed displayed the largest precipitation (925.2 mm) in summer but the lowest precipitation (71.8 mm) in winter, indicating the watershed has the most fluctuating precipitation characteristic. While the precipitation trends in the Four River Systems varied, a distinct feature in baseflow trends was not found, moreover baseflow percentages to streamflow were typically greater than 50% in the Four River Systems. As shown in this study, it would be expected significant amount of pollutants could be contributed to the stream in the form of baseflow at the watershed.

Pollutant in watersheds comes from two major sources which are NPS (nonpoint source pollution) and PS (point source pollution). Most of the pollutant can be treated by wastewater treatment plants. However, wastewater treatment plants may not be an appropriate practice to improve water quality for the watersheds with large portion of NPS pollutant and NPS pollution from direct runoff and baseflow has different characteristics. Therefore the practices to improve water quality need to be comprehensive for pollutants by both direct runoff and baseflow. Riparian buffer, one of practices to manage pollutant in watershed, has been applied to reduce pollutant not only from direct runoff but also baseflow. In this study, the scenarios for pollutant reduction by wastewater treat plants and the nitrogen reduction by riparian buffer were simulated using SWAT-REMM to suggest an effective plan for pollutant reduction from baseflow. Riparian buffer provided nitrogen reduction of 0.2~75.0% in YbB watershed and 38.0~47.0% in GbA watershed. The result indicates that riparian buffer is effective to reduce the pollutant especially from baseflow, and it suggested as suitable for the a watershed which WWTP discharge is not capable to reduce enough pollutant.

It has been well known that it is not easy to quantify pollutant loads driven by non-point source pollution due to various factors affecting generation and transport mechanism of it. Especially pollutant loads through baseflow have been investigated by limited number of researchers. Thus in this study, the Web-based WAPLE (WHAT-Pollutant Load Estimation) system was developed and applied at study watersheds to quantify baseflow contribution of pollutant. In YbB watershed, baseflow contribution with WWTP discharge is responsible for 49.5% of total pollutant loads at the watershed. Among these, pollutant loads through baseflow (excluding any WWTP discharge) is responsible for 61.7% of it. In GbA watershed, it was found that 58.4% is contributed by baseflow with WWTP discharge 2.9% and 97.1% is by baseflow. For NbB watershed (without WWTP discharge), 52.3% of pollutant load is transported through baseflow. As shown in this study, it was found that over 50.0% of TN (Total Nitrogen) pollutant loads are contributed by non-direct runoff. Thus pollutant loads contributed by baseflow and WWTP discharge as well as direct runoff contribution should be quantified to develop and implement watershed-specific Best Management Practices during dry period.