Uncertainty-based metal budget assessment at the watershed scale: Implications for environmental management practices

Heavy metal (HM) pollution is a serious and urgent issue in integrated watershed management in China and worldwide. Determining effective management strategies for pollution control requires quantification of input and output metal fluxes and the inherent uncertainties. Process-based metal models can simulate the metal movement on a watershed scale providing essential information for metal budget. However, there has been little effort to quantify the uncertainties in metal flux simulations. In this work, we modified the previously developed SWAT-HM (Soil and Water Assessment Tool - Heavy Metal) model by adding two external model inputs (atmospheric deposition and agricultural-source input) and one process-based module (plant uptake). We then linked the modified model with the SWAT Calibration and Uncertainty Programs SWAT-CUP for stochastic calibration and uncertainty analysis. The modified SWAT-HM was used to model the fluxes of primary inputs and outputs of zinc (Zn) in both uplands (soil) and channel (bed sediment) in the upper Liuyang River watershed in south-central China. To calibrate the model with uncertainty analysis, we used six-years of daily streamflow, daily sediment load, and daily Zn load at monthly frequency at the watershed outlet. In the upland phase, we identified Zn input from atmospheric deposition and Zn output through soil erosion as the most significant fluxes. In the channel phase, bed sediment was the critical Zn sink receiving 5,100 to 42,000 kg yr(-1) Zn. The method used in SWAT-HM calibration and uncertainty analysis is general with potential application to similar settings in the world.