Urban forests are recognized as a nature-based solution for stormwater management. This study assessed the underlying processes and extent of runoff reduction due to street trees with a paired-catchment experiment conducted in two sewersheds of Fond du Lac, Wisconsin. Computer models are flexible, fast, and low-cost options to generalize and assess the hydrologic processes determined in field studies. A state-of-the-art, public-domain model, which explicitly simulates urban tree hydrology, i-Tree Hydro, was used to simulate the paired-catchment experiment, and results from field observations and simulation predictions were compared to assess model validity and suitability as per conditions in the broader Great Lakes basin. Model parameters were aligned with observed conditions using automatic and manual calibration. Model performance metrics were used to quantify the weekly performance of calibration and to validate predictions. Those calibration metrics differed substantially between the two periods simulated, but most calibration metrics remained positive, indicating the model was not fitting only the period used for calibration. Predicted avoided runoff for a five-month leaf-on period was 64 L/m² of canopy, 4 % lower than the field-estimated avoided runoff of 66 L/m² of canopy. Interception was the most directly comparable process between the model and field observations. Based on 5 storms sampled, field estimation of precipitation intercepted and retained on trees averaged 63 % and ranged from 22 % to 81 %, while model estimation averaged 61 % and ranged from 36 % to 99 %. This model was able to fit predictions to observed catchment discharge but required extensive manual calibration to do so. The i-Tree Hydro model predicted avoided runoff comparable with the field study and earlier assessments. Additional field studies in similar settings are needed to confirm findings and improve transferability to other tree species and environmental settings.

城市森林被认为是基于自然的雨水管理解决方案. 本研究通过在威斯康星州 Fond du Lac 的两个下水道进行的成对集水试验，评估了行道树造成的径流减少的基本过程和程度。计算机模型是灵活、快速和低成本的选择，可以概括和评估在实地研究中确定的水文过程。一个最先进的公共领域模型，明确模拟城市树木水文，i-Tree Hydro，用于模拟成对集水区实验，并比较现场观察和模拟预测的结果以评估模型的有效性根据更广泛的五大湖流域的条件和适用性。使用自动和手动校准将模型参数与观察到的条件对齐。模型性能指标用于量化校准的每周性能并验证预测。这些校准指标在模拟的两个时期之间存在显着差异，但大多数校准指标仍然是积极的，这表明该模型不仅适合用于校准的时期。五个月叶期的预计避免径流为 64 L/m²冠层，比现场估计的 66 L/m ²冠层避免径流低 4%。拦截是模型和实地观察之间最直接可比的过程。根据对 5 次风暴采样，截获和保留在树木上的降水的现场估计平均为 63%，范围为 22% 至 81%，而模型估计平均为 61%，范围为 36% 至 99%。该模型能够对观察到的集水区流量进行预测，但需要大量的手动校准才能做到这一点。i-Tree Hydro 模型预测的避免径流可与实地研究和早期评估相媲美。需要在类似环境中进行额外的实地研究，以确认发现并提高对其他树种和环境环境的可转移性。