This paper presents a comparison between the black box Nonlinear Auto-Regressive with eXogenous inputs-Neural Network (NARX-NN) and the conceptual Hydrologic Engineering Centre-Hydrologic Modelling System (HEC-HMS) Rainfall-Runoff models. The two models were applied on a small urban watershed to assess its response to fourteen hourly real storm events. The differences between the steps engaged in each model to reach the hydrograph were presented in detail. The estimation of the best parameters is carried out using a weighted average function during the calibration phase. A statistical evaluation was conducted to assess the model’s performance thereafter; a critical comparison was made to illustrate the differences and discuss the steps involved. The results indicate that both models successfully reflect the urban basin runoff. However, the NARX-NN outperforms in the testing phase owing to their strength generalization feature. The NARX-NN model has more strength to produce the shape bending of the hydrograph. Consequently, this model is better to highlight the curvatures resulting from the local peaks of rainfall.

Watershed climatic diversity poses a hard problem when it comes to finding suitable models to estimate inter-annual rainfall runoff (IARR). In this work, a hybrid model (dubbed MR-CART) is proposed, based on a combination of MR (multiple regression) and CART (classification and regression tree) machine-learning methods, applied to an IARR predicted data series obtained from a set of non-parametric and empirical water balance models in five climatic floors of northern Algeria between 1960 and 2020. A comparative analysis showed that the Yang, Sharif, and Zhang’s models were reliable for estimating input data of the hybrid model in all climatic classes. In addition, Schreiber’s model was more efficient in very humid, humid, and semi-humid areas. A set of performance and distribution statistical tests were applied to the estimated IARR data series to show the reliability and dynamicity of each model in all study areas. The results showed that our hybrid model provided the best performance and data distribution, where the R²_Adj and p-values obtained in each case were between (0.793, 0.989), and (0.773, 0.939), respectively. The MR model showed good data distribution compared to the CART method, where p-values obtained by signtest and WSR test were (0.773, 0.705), and (0.326, 0.335), respectively.