RESUMEN
The prediction of suspended sediment load (SSL) within riverine systems is critical to understanding the watershed's hydrology. Therefore, the novelty of our research is developing an interpretable (explainable) model based on deep learning (DL) and Shapley Additive ExPlanations (SHAP) interpretation technique for prediction of SSL in the riverine systems. This paper investigates the abilities of four DL models, including dense deep neural networks (DDNN), long short-term memory (LSTM), gated recurrent unit (GRU), and simple recurrent neural network (RNN) models for the prediction of daily SSL using river discharge and rainfall data at a daily time scale in the Taleghan River watershed, northwestern Tehran, Iran. The performance of models was evaluated by using several quantitative and graphical criteria. The effect of parameter settings on the performance of deep models on SSL prediction was also investigated. The optimal optimization algorithms, maximum iteration (MI), and batch size (BC) were obtained for modeling daily SSL, and structure of the model impact on prediction remarkably. The comparison of prediction accuracy of the models illustrated that DDNN (with R2 = 0.96, RMSE = 333.46) outperformed LSTM (R2 = 0.75, RMSE = 786.20), GRU (R2 = 0.73, RMSE = 825.67), and simple RNN (R2 = 0.78, RMSE = 741.45). Furthermore, the Taylor diagram confirmed that DDNN has the highest performance among other models. Interpretation techniques can address the black-box nature of models, and here, SHAP was applied to develop an interpretable DL model to interpret of DL model's output. The results of SHAP showed that river discharge has the strongest impact on the model's output in estimating SSL. Overall, we conclude that DL models have great potential in watersheds to predict SSL. Therefore, different interpretation techniques as tools to interpret DL model's output (DL model is as black-box model) are recommended in future research.
RESUMEN
The geomorphometric analysis of watersheds provides useful quantitative information on stream hydrology and potential landscape change that can be used by soil conservation decision makers to determine areas prone to land degradation. In this study, we develop a methodology for the assessment of catchment-scale sensitivity to sediment yield using various topo-hydrological, vegetation, and climatic parameters using four multi-criteria decision making (MCDM) techniques: the Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS), Vlse Kriterijumska Optimizacija Kompromisno Resenje (VIKOR), weighted-sum analysis (WSA), and combined factor (CF). To identify the most important factors affecting sediment yield and soil erosion, a model incorporating principle component analysis with MCDM was devised, using infiltration number (IF), drainage density (Dd), length of overland flow (Lo), channel maintenance (C), stream frequency (Fs), and ruggedness number (Rn) as indices of sediment and erosion risk. Data from a previous study that employed the RUSLE3D model and sediment-yield field data were used to validate the results. The TOPSIS model achieved the highest correlation with the RUSLE3D results. The correlation of watershed activities to the experimental erosion and sediment prioritization results is 0.32. The TOPSIS results indicate that all 23 sub-watersheds yielded moderate amounts of sediment. Based on the VIKOR method, 17.39% (78.96 km2) of the region was classified as having very high erodibility, 26.08% (241.93 km2) high erodibility, 34.78% (225.95 km2) moderate erodibility, and 21.73% (105.05 km2) low erodibility. Considering the high sensitivity of Taleghan watershed to soil erosion, it is recommended that conservation efforts be implemented to minimize land degradation in the area. This methodology can be adapted to other regions that lack detailed topo-hydrological, vegetation, or climatic data.
