Ground‐mounted photovoltaic sites are often treated as impervious surfaces in stormwater permits. This ignores the pervious soils beneath and between solar arrays and leads to an overestimation of runoff. Our objective was to improve solar farm stormwater hydrology models by explicitly considering the disconnected impervious nature of solar design and site characteristics. Experimental sites established on utility scale solar farms in Colorado, Georgia, Minnesota, New York, and Oregon had perennial vegetative plantings with mean precipitation ranging from 40.6 to 124.5 cm, and soil texture ranging from loamy sand to clay. Soil moisture measurements were collected beneath arrays, under drip edges, and in the vegetated area between arrays at each site. Hydrus‐3D models for soil moisture and stormwater hydrology were developed that accounted for precipitation falling on solar panels, drip edge redistribution of rainfall, infiltration, and runoff in the pervious areas between solar arrays and beneath panels. Drip edge runoff averaged 3‐ to 10‐times incident precipitation at the New York and Minnesota sites, respectively. Root mean square error values between measured sub‐hourly soil moisture and predicted moisture for large measured single storm events averaged 0.029 across all five sites. Predicted runoff depths were strongly affected by precipitation depth, soil texture, soil profile depth, and soil bulk density. Runoff depths across the five experimental sites averaged 13%, 25%, and 45% of the 2‐, 10‐, and 100‐year design storm depths, clearly showing that these solar farms do not behave like impervious surfaces, but rather as disconnected impervious surfaces with substantial infiltration of runoff in the vegetated areas between and beneath solar arrays.

中文翻译：

---

使用断开的不透水表面方法测量和模拟太阳能发电场的土壤湿度和径流

在雨水许可中，地面安装的光伏发电站通常被视为不透水表面。这忽略了太阳能电池阵下方和之间的透水土壤，并导致径流的高估。我们的目标是通过明确考虑太阳能设计和场地特征的不连贯不渗透性质来改进太阳能发电场雨水水文模型。在科罗拉多州、佐治亚州、明尼苏达州、纽约州和俄勒冈州的公用事业规模太阳能发电厂建立的实验场地种植了多年生植物，平均降水量为 40.6 至 124.5 厘米，土壤质地为壤土到粘土。土壤湿度测量是在每个地点的阵列下方、滴水边缘下方以及阵列之间的植被区域收集的。开发了土壤湿度和雨水水文学的 Hydrus-3D 模型，该模型考虑了落在太阳能电池板上的降水、降雨滴水边缘重新分布、渗透以及太阳能电池板之间和电池板下方渗透区域的径流。纽约和明尼苏达地点的滴水边缘径流平均分别是事件降水量的 3 至 10 倍。在所有五个地点，测量的每小时土壤湿度与大型测量的单次风暴事件的预测湿度之间的均方根误差值平均为 0.029。预测径流深度受降水深度、土壤质地、土壤剖面深度和土壤容重的强烈影响。五个实验地点的径流深度平均为 2 年、10 年和 100 年设计风暴深度的 13%、25% 和 45%，清楚地表明这些太阳能发电厂的行为并不像不透水的表面，而是互不相连的。不透水表面，径流大量渗透到太阳能电池阵之间和下方的植被区域。