Principles of Stochastic Generation of Hydrologic Time Series for Reservoir Planning and Design: Case Study

Abstract

Abstract:Simulation has been an important tool for planners in many fields of knowledge. In the field of water resources the uncertainties due to unknown data population and the short length of the records work together to make the simulation especially important. The major utilization of water resources at the level needed in modern society makes water storage essential for satisfying the demand. Therefore, the need to reduce the uncertainty in the design of water storage capacity is an important problem in the field of water resources utilization. This problem can only be satisfactorily solved with the aid of simulation. On the other hand, the implementation of adequate exploration politics also needs to use simulations in order to obtain results with low uncertainty. In simulation of water resources systems the use of synthetic time series is a current practice. In this study the number of generated time series to use in the problems described are analyzed. Annual and monthly synthetic flows were generated preserving the relevant statistics of the available historical data and then the number of time series to generate was determined. The results of the case studies indicate that the proposed methodology is a plausible approach to solve the analyzed problem and lead to the conclusion that the number of time series to generate should be 1200.

The integrated effects of climate and hydrologic uncertainty on future flood risk assessments

“…Without parallel computing capabilities, the number of climate simulations for each climate change must be limited to manage the computational burden of the entire vulnerability assessment. However, many weather generator simulations are needed to adequately explore the possible climate fluctuations that may emerge due to chance (Guimarães and Santos, ). To circumvent this issue, we initially generate 500 different 62‐year, daily simulations of climate and then choose ten of those simulations to be used for the vulnerability assessment.…”

A hiddenMarkov model combined with climate indices for multidecadal streamflow simulation

“…Following Guimarães and Santos [], we generated 1200 ensembles from both models, HMG and HMC, each of length 105 (the same length of the data). Figure shows the boxplots of basic distributional statistics—mean, standard deviation, skewness, lag‐1 autocorrelation, minimum, and maximum—along with observed value as horizontal line.…”

Um desenvolvimento adicional do método dos fragmentos. Aplicação à desagregação de escoamentos anuais em escoamento diários

“…A selecção aleatória de fragmentos é feita sem reposição até que a classe fique vazia. Quando tal acontece, procede-se à reposição dos fragmentos originalmente contidos nessa classe [19,20,34].…”