Abstract
The identification of the critical source of nonpoint source pollution (NPS) plays an important role in formulation of Best
Management Practices (BMPs). SWAT (Soil and Water Assessment Tool) was adopted for NPS nitrogen pollution modeling for
Qinhuangdao City. The model was calibrated and validated by Sequential Uncertainty Fitting Version 2 algorithm of SWATCUP
(SWAT Calibration and Uncertainty Programs) software. The model was used to quantify contributions of different
nitrogen sources to rivers total nitrogen (TN) load, and address the spatial-temporal distribution of NPS TN to river. The results
show that Changli county, Funing county and Lulong county have the most quantity of TN-to-river. The main sources of TN
pollution are livestock, nitrogen fertilizing and atmospheric deposition, which account for 49.47%, 26.14% and 15.31% of river
TN load, respectively. Rural residential land has the largest TN load (63.418kg/ha). The quantity of nitrogen from soil and
atmosphere is mostly affected by precipitation, and that from livestock, nitrogen fertilizing and rural life is affected by both
precipitation and sewage time. So to solve the nonpoint source nitrogen pollution issue, the plan development must take the
factors of time, region and type into account.