The original paradigm of urban hydrology identified links between urbanization and changes in hydrologic processes. One of the prominent relationships identified was increased impervious surface coverage decreased subsurface recharge and subsequently low flows. This does not consider watershed‐specific biophysical or anthropogenic characteristics that could increase low flows, such as direct discharge from wastewater treatment plants and indirect recharge to groundwater from leaking pipes and infiltration‐based green infrastructure. Here, in an effort to sharpen this original paradigm, we take a statistical approach to evaluate relationships between natural and human‐impacted watershed characteristics and a suite of low flow metrics that describe mean annual low flow magnitudes and interannual variability in 54 watersheds over 15 years in the Piedmont physiographic region of the southeastern United States. We find that annual low flow magnitudes are most strongly correlated to biophysical watershed factors, including positive relationships with topography and precipitation and negative relationships with poorly drained soils. In contrast, interannual variability is positively correlated to land use characteristics, including forest, and negatively related to urbanization. Similarities in best fit models across long‐term average, wet, and dry years indicate a robustness of hydrologic drivers across watershed wetness conditions. Our results also showcase that while the original paradigm of urban hydrology may hold across traditional low flow metrics, such as baseflow index, extreme low flow drivers are complex and counterintuitive. Understanding these first‐order controls on baseflow magnitude and variability across an urbanization gradient is needed to conceptualize the impacts that continued, rapid urbanization will have on low flow dynamics.

中文翻译：

---

美国东南部低流量水文的人为和生物物理控制

城市水文学的原始范式确定了城市化与水文过程变化之间的联系。确定的主要关系之一是不透水表面覆盖率增加，地下补给减少，随后流量减少。这不考虑会增加低流量的流域特定的生物物理或人为特征，例如废水处理厂的直接排放以及泄漏的管道和基于渗透的绿色基础设施对地下水的间接补给。在这里，为了强化这种原始范式，我们采用一种统计方法来评估自然和人类影响的流域特征之间的关系，以及一套低流量指标，这些指标描述了美国东南部皮埃蒙特（Piedmont）地理区域内15年中54个流域的年平均低流量幅度和年际变化。我们发现，每年的低流量大小与生物物理分水岭因素最相关，包括与地形和降水的正相关以及与排水不良的土壤的负相关。相反，年际变化与土地利用特征（包括森林）呈正相关，与城市化呈负相关。长期平均，湿润和干旱年份最佳拟合模型的相似性表明，在流域湿润条件下，水文驱动因素具有较强的鲁棒性。我们的结果还表明，尽管城市水文学的原始范式可能适用于传统的低流量指标（例如基流指数），但极端的低流量驱动因素却是复杂且违反直觉的。需要了解这些关于城市化梯度下的基础流量大小和变异性的一阶控制，以概念化持续快速的城市化将对低流量动态产生的影响。