Rainfall-runoff modeling in ungauged basins continues to be a great hydrological research challenge. A novel approach is the Long-Short-Term-Memory neural network (LSTM) from the Deep Learning toolbox, which few works have addressed its use for rainfall-runoff regionalization. This work aims to discuss the application of LSTM as a regional method against traditional neural network (FFNN) and conceptual models in a practical framework with adverse conditions: reduced data availability, shallow soil catchments with semiarid climate, and monthly time step. For this, the watersheds chosen were located on State of Ceará, Northeast Brazil. For streamflow regionalization, both LSTM and FFNN were better than the hydrological model used as benchmark, however, the FFNN were quite superior. The neural network methods also showed the ability to aggregate process understanding from different watersheds as the performance of the neural networks trained with the regionalization data were better with the neural networks trained for single catchments.

This research evaluates the partial-area effect and its relationship with the rainfall intensity–duration–frequency (IDF) equations. In the Rational Method, if the critical rainfall duration is shorter than the time of concentration, the partial-area effect occurs. We proved that the partial area could exist for the general ID equation $i=a/{\left(b+t_d\right)}^c$, only when $c>1$. For these equations, in the application of the Rational Method, the maximum discharge at basin outlet occurs for rainfall duration ($t_d$) equal to b/(c−1). Nevertheless, for that case, the Depth Duration Frequency (DDF) has a maximum at that rainfall duration. These situations are present in engineering practice and will be discussed in this paper. Research was done to look for IDF equations with $c>1$ in hydrologic engineering practice. It found 640 inconsistent IDF equations ($c>1$) in four countries (Brazil, Mexico, India, and USA), which means that a fundamental principle for building consistent IDF equations (i.e., $c>1$), published in the scientific literature since 1998, did not reach the hydrologic engineering practice fully. We provided some analysis regarding this gap between theory and engineering practice.

Socio-hydrology has recently garnered attention as a promising research paradigm among hydrologists. In this paper, we argue that viewing the socio-hydrological systems as a network of causal factors can assist in these efforts by revealing emergent properties of socio-hydrological systems. These factors, which are neither actors nor affiliations but symbolic representations of miscellaneous phenomena, represent the key variables affecting socio-hydrological system behaviour. Despite being crucial for understanding system dynamics, these factors are not captured by most equations of coupled human–water coevolutionary models. To demonstrate our method, we use qualitative case study research to analyse the Jaguaribe Basin in Ceará State-Brazil as a network of factors, i.e. mental conceptions, and representations of water resource management experts. We show that most results are unsurprising, exhibiting expected causal relationships. However, there was at least one unexpected result, which showed that the variable “conflict” was the single most sensitive variable to systemic evolution.

This paper analyzes decadal climate phenomena and the phases of the Atlantic Multi-Decadal Oscillation (AMO) and Pacific Decadal Oscillation (PDO), and assesses the climate-water-energy nexus at the reservoir system of the São Francisco River Basin (SFRB), Brazil. This study brings a non-traditional approach focused on the link between climatic patterns and hydrological processes, and their social-economic consequences, such as energy production. The methodology used data of the precipitation, natural streamflow, AMO and PDO series. These variables were evaluated by Wavelet Transform Analysis (WTA), Flow Duration Curve (FDC), streamflow that is equaled or exceeded 90% of the time (Q90), the Reliability Index (RI) for different demands and the hydropower generated. The results suggest that the WTA for the SFRB was more significant on the interannual scale, and the period with opposite phases of AMO(-) and PDO(+) showed the highest values of FDC, Q90, RI, and hydropower generated simulated for the SFRB.

Streamflow forecasting plays an important role in ensuring the reliable supply of electricity in countries heavily reliant on hydropower. This paper proposes a novel framework that integrates various hydrological models, climate models, and observational data to develop a comprehensive forecasting system. Three families of models were employed: seasonal forecasting climate models integrated with hydrological rainfall-runoff models; stochastic or machine learning models utilizing endogenous variables, and stochastic or machine learning models that consider exogenous variables. The hyper-multimodel framework could successfully increase the overall performance of the scenarios generated through the use of the individual models. The quality of the final scenarios generated was directly connected to the performance of the individual models. Therefore, the proposed framework has potential to improve hydrological forecast for the Brazilian electricity sector with the use of more refined and calibrated individual models.

This work proposes a methodology for post-flood analysis in ungauged basins with low data availability located in semi-arid regions. The methodology combines social perception with recorded data. Social perception can be a useful tool to enhance the modeling process in cases where official records are nonexistent or unsatisfactory. For this aim, we structured a four-step methodology. First, we create a repository with the information that reconstructs the analyzed event. Photos and news of the flood event are collected from social media platforms. The next step is to consult official government agencies to obtain documented information about the disaster. Then, semi-structured interviews are carried out with residents to obtain the extension and depth of the flooded spot. This social information creates an overview of the flood event that can be used to evaluate the hydraulic/hydrological modeling of the flood event and the quality of the recorded data. We analyzed a flood event in a city in semi-arid Brazil. The event caused several damages such as the breaking of dams and about 40% of the population was somehow impacted although the official rain data pointed to non-extreme precipitation.