ABSTRACT Ballona Creek watershed in Los Angeles, California provides a unique combination of heterogeneous urban land cover, a semi-arid environment, and a large outdoor water-use flux that presents a challenge for physically-based models. We ran simulations using the Noah Land Surface Model and Parflow-Community Land Model and compared to observations of evapotranspiration (ET), runoff, and land surface temperature (LST) for the entire 11-year study period. Both models were systematically adjusted to test the impact of land cover and urban irrigation on simulation results. Monthly total runoff and ET results are greatly improved when compared to an in-situ stream gauge and meteorological tower data: from 0.64 to 0.81 for the Nash–Sutcliffe efficiency (NSE) for runoff and from a negative NSE to 0.82 for ET. The inclusion of urban irrigation in semi-arid urban environments is found to be vital, but not sufficient, for the accurate simulation of variables in the studied models.

中文翻译：

---

半干旱城市环境建模中的城市灌溉：加利福尼亚州洛杉矶 Ballona Creek 流域

摘要加利福尼亚州洛杉矶的 Ballona Creek 流域提供了一种独特的组合，将异质的城市土地覆盖、半干旱环境和大型室外用水通量结合在一起，这对基于物理的模型提出了挑战。我们使用 Noah Land Surface Model 和 Parflow-Community Land Model 进行了模拟，并与整个 11 年研究期间的蒸散量 (ET)、径流和地表温度 (LST) 的观测值进行了比较。两种模型都经过系统调整，以测试土地覆盖和城市灌溉对模拟结果的影响。与原位流量测量仪和气象塔数据相比，每月总径流和 ET 结果大大改善：Nash-Sutcliffe 效率 (NSE) 从 0.64 到 0.81 径流，从负 NSE 到 0.82 ET。