Knowledge of hydrodynamic behaviour is imperative for the water resource management and the sustainable development of basins. Hydraulic and hydrologic modelling are essential tools for the simulation of this behaviour. However, lack of basin assessment makes modelling input data scarce. This is the case for the Juqueriquere River basin, which comprises the major floodplains of the northern coastline of Sao Paulo State, Brazil. This region is susceptible to floods and is under pressure from real estate speculation, gas exploitation, and infrastructure development for port facilities. The purpose of this study was to estimate and calibrate Manning’s roughness coefficients of the Juqueriquere River basin, as they are mandatory data for hydraulic modelling. The estimation of Manning's n values was based on field observations. These data were inserted into the Hydrologic Engineering Center's River Analysis System (HEC-RAS) model for the calibration of the roughness coefficients in a trial and error process. The findings of this study proved that the estimation and calibration methods were efficient, with mean absolute deviations lower than 10% in the studied sections.

Located in the southeast region of Brazil, the Cantareira System consists of six interconnected reservoirs and supplies around 14 million people in the state of São Paulo. Built in the 1970s, when extensive fluviometric series were not available in the region, the system underwent several operating rules that culminated in the water crisis caused by the 2014/2015 drought. This article makes a brief critical account of what has been experienced in these almost 50 years of operating the system, the factors that influenced the water crisis, and what has been learned.

The Soil Conservation Service Curve Number (SCS-CN) model is widely used to estimate the watershed runoff. However, the model may result in inaccurate estimations when assuming a tabulated CN value and an initial abstraction (λ) of 0.20. Therefore, in this paper, we evaluated the suitability of four CN estimation methods for an urban watershed: Tabulated CN, National Engineering Handbook 4, Least Squares and Asymptotic Fitting. Also, the effect of ordering the data and different λ values (0.20 or 0.05) were analyzed. Our findings suggest that the transient approach of Asymptotic Fitting method, a variance from the original Asymptotic Fitting method, provides the best runoff estimations using rank ordered data and λ equal to 0.05 (R2 = 0.79, PBIAS = 0.01% and NSE = 0.78). It is to be noted that similar results were also obtained through a simple linear model.

This article demonstrates the successful application of Long Short-Term Memory (LSTM) recurrent neural networks to simulate streamflow in the Aquidauana River basin, located in the Brazilian Pantanal. The LSTM network used daily precipitation data as input to predict future streamflow in the region. The results obtained from this research show a coefficient of determination (R2) of 0.82, indicating a strong fit of the model to the observed data. Additionally, the Root Mean Squared Error (RMSE) was found to be 0.53, indicating the model's accuracy in predicting streamflow compared to actual data. These findings highlight the effectiveness of LSTM networks in hydrological modeling for the Pantanal region, which is crucial for water resource planning and sustainable management in this ecologically significant area. This study is expected to serve as a catalyst for further research and make a substantial contribution to the advancement of streamflow prediction techniques in complex watersheds such as the Aquidauana River basin.