Particulate materials arising from road-deposited sediments (RDS) are an essential target for the control and management of surface runoff pollution. However, the heterogeneity of urban spaces hinders the identification and quantification of particulate pollution, which is challenging when formulating precise control measures. To elucidate the factors that drive particulate pollution in heterogeneous urban spaces, the accumulation of RDS on dry days and the total suspended solids during six natural rainfall events were investigated across three urban–rural spatial units (central urban, central suburban, and remote suburban). The underlying surface type (asphalt or cement roads) and particle size composition jointly determined the spatial heterogeneity in the static accumulation and dynamic output loads of RDS during rainfall. These two factors explained 59.6% and 18.9% of the spatial heterogeneity, respectively, according to principal component analysis. A novel exponential wash-off equation that incorporates particle size composition and underlying surface type was applied. It precisely described the spatial heterogeneity of RDS wash-off loads, the estimated values exhibiting event mean concentration errors of 10.8–18.2%. When coupled with the curve, this exponential wash-off equation more precisely split the initial volume of runoff: a lower total volume (17.6–38.0%) was shown to carry a higher proportion of the load (70.0–93.7%) compared to the traditional coupled exponential wash-off equation (volume: 31.6–49.0%, load: 37–90%). This study provides a new approach to characterizing RDS wash-off processes and splitting initial runoff in heterogeneous spaces.

中文翻译：

---

异质城市空间中道路沉积物冲刷特征及初始径流污染的准确分割

道路沉积物（RDS）产生的颗粒物质是控制和管理地表径流污染的重要目标。然而，城市空间的异质性阻碍了颗粒物污染的识别和量化，这在制定精确的控制措施时具有挑战性。为了阐明异质城市空间颗粒物污染的驱动因素，我们对三个城乡空间单元（中心城市、中心郊区和偏远郊区）的干燥天 RDS 积累和六次自然降雨事件期间的总悬浮固体进行了调查。下垫面类型（沥青或水泥路面）和颗粒尺寸组成共同决定了降雨过程中RDS静态累积和动态输出荷载的空间异质性。根据主成分分析，这两个因素分别解释了空间异质性的59.6%和18.9%。应用了一种新颖的指数冲刷方程，该方程结合了颗粒尺寸组成和下垫面类型。它精确地描述了 RDS 冲刷负荷的空间异质性，估计值显示事件平均浓度误差为 10.8-18.2%。当与曲线结合时，这个指数冲刷方程更精确地分割了径流的初始体积：与传统耦合指数冲洗方程（体积：31.6–49.0%，负载：37–90%）。这项研究提供了一种表征 RDS 冲刷过程和在异质空间中分割初始径流的新方法。