We use surface soil moisture content as a proxy to assess the effect of drainage congestion due to structural barriers on the alluvial Fan of the Kosi River on the Himalayan Foreland. We used Sentinel-1 satellite images to evaluate the spatial distribution of soil moisture in the proximity of structural barriers (i.e., road network). We applied modified Dubois and a fully connected feed-forward artificial neural network (FC-FF-ANN) models to estimate soil moisture. We observed that the FC-FF-ANN predicts soil moisture more accurately (R = 0.85, RMSE = 0.05 ${\mathrm{m}}^3/{\mathrm{m}}^3$, and bias = 0) as compared to the modified Dubois model. Therefore, we have used the soil moisture from the FC-FF-ANN model for further analysis.

We identified the road network that traverses on the Kosi Fan horizontally, vertically, and with inclination. We create a buffer of 1 km along either side of the road. Within this, we assessed the spatial distribution of soil moisture. We observed a high concentration of soil moisture near the structural barrier, and decreases gradually as we move farther in either direction across the orientation of the road. The impact of structural barriers on the spatial distribution of soil moisture is prominent in a range between 300 to 750 m within the road buffer. This study is a step towards assessing the effect of structural interventions on drainage congestion and flood inundation.

我们使用地表土壤含水量作为代表来评估由于结构障碍对喜马拉雅前陆科西河冲积扇造成的排水堵塞的影响。我们使用 Sentinel-1 卫星图像评估结构障碍（即道路网络）附近土壤水分的空间分布。我们应用改进的 Dubois 和完全连接的前馈人工神经网络 (FC-FF-ANN) 模型来估计土壤水分。我们观察到 FC-FF-ANN 更准确地预测土壤水分（R = 0.85，RMSE = 0.05${\mathrm{米}}^3/{\mathrm{米}}^3$, 和偏差 = 0) 与修改后的 Dubois 模型相比。因此，我们使用 FC-FF-ANN 模型中的土壤水分进行进一步分析。

我们确定了在 Kosi Fan 上水平、垂直和倾斜的道路网络。我们在道路两侧创建了一个 1 公里的缓冲区。在此范围内，我们评估了土壤水分的空间分布。我们观察到结构屏障附近的土壤水分浓度很高，并且随着我们在道路方向的任一方向上移动更远，土壤水分逐渐减少。结构屏障对土壤水分空间分布的影响在道路缓冲区内300～750 m范围内较为显着。这项研究是评估结构干预对排水拥堵和洪水淹没影响的一步。