SPARROW模型及其应用研究进展

SPARROW模型是由美国地质调查局开发的一个基于流域空间属性的估算污染物负荷、浓度等的非线性回归模型。由于模型通过质量守恒来约束污染物的传输,并以统计学的方法实现变量参数的校准,因而SPARROW模型在量化污染物的传输过程中具有足够高的精确度与合理性。总体来看,SPARROW模型在流域污染源及环境因子分析、水质评估与模拟、监测管理优化等方面发挥出了重要作用,并被广泛地应用于国内外的不同流域。针对SPARROW模型在不确定性分析中存在的自相关问题,贝叶斯分析的引入优化了模型在不确定性方面的评估。目前,SPARROW模型在国内流域中以估算总氮、总磷、COD等污染物负荷为主要应用。随着国内相关数据的积累以及共享程度的提高,其应用范围将会愈加广泛。

Research Progress of the SPARROW Model and Its Application

The SPARROW (SPAtially Referenced Regressions On Watershed attributes) model, which was developed by USGS, is a non-linear statistic modeling approach based on watershed attributes. Since of being constrained by mass balance, the SPARROW model has relative high accuracy and rationality in quantifying migration of contaminant. In general, the SPARROW model has been performed well on identifying contaminant sources, analyzing land-to-water delivery factors, evaluating water quality, simulating water quality and optimizing watershed environment management in various basins. Recently, the Bayesian analysis has been introduced to improve the uncertainty issue of this model. Although there are only a few applications of the SPARROW model in China, its application will be more extensive since of the rapid growth of accumulation and sharing of data.