Heinz Fill

Using regional regression within index flood procedures and an empirical Bayesian estimator

Studies have illustrated the performance of at-site and regional flood quantile estimators. For realistic generalized extreme value (GEV) distributions and short records, a simple index-flood quantile estimator performs better than two-parameter (2P) GEV quantile estimators with probability weighted moment (PWM) estimation using a regional shape parameter and at-site mean and L-coefficient of variation (L-CV), and full three-parameter at-site GEV/PWM quantile estimators. However, as regional heterogeneity or record lengths increase, the 2P-estimator quickly dominates. This paper generalizes the index flood procedure by employing regression with physiographic information to refine a normalized T-year flood estimator. A linear empirical Bayes estimator uses the normalized quantile regression estimator to define a prior distribution which is employed with the normalized 2P-quantile estimator. Monte Carlo simulations indicate that this empirical Bayes estimator does essentially as well as or better than the simpler normalized quantile regression estimator at sites with short records, and performs as well as or better than the 2P-estimator at sites with longer records or smaller L-CV.

Monthly rainfall–runoff modelling using artificial neural networks

Abstract Rainfall–runoff models usually present good results, but parameter calibration sometimes is tedious and subjective, and in many cases it depends on additional data surveys in the field. An alternative to the conceptual models is provided by empirical models, which relate input and output by means of an arbitrary mathematical function that bears no direct relationship to the physical characteristics of the rainfall–runoff process. This category includes the artificial neural networks (ANNs), whose implementation is the main focus of this paper. This study evaluated the capacity of ANNs to model with accuracy the monthly rainfall–runoff process. The case study was performed in the Jangada River basin, Paraná, Brazil. The results of the three ANNs that produced the best results were compared to those of a conceptual model at monthly time scale, IPHMEN. The ANNs presented the best results with highest correlation coefficients and Nash-Sutcliffe statistics and the smallest difference of volume. Citation Machado, F., Mine, M., Kaviski, E. & Fill, H. (2011) Monthly rainfall–runoff modelling using artificial neural networks. Hydrol. Sci. J. 56(3), 349–361.

Método de Regionalização para Avaliar a Energia Garantida Incremental de PCHs a Fio de Água Integradas na Região Sul do Brasil

There has been an increasing participation of small hydroplants (less than 30 MW) on the Brazilian energy production scene. This raises the need for the appropriate evaluation of their contribution to the firm energy output of the interconnected system. ANEEL (2001) proposed the computation of this contribution by censuring monthly streamflows at the plant capacity and computing their average over a 30 year period. Since most of these plants are not able to regulate monthly flows the censuring should be performed daily. Also using any 30 year period may not take critical hydrological conditions into account. FILL (1989) proposed a method based on stochastic reservoir theory to evaluate the energy contribution of hydroplants, integrated into an interconnected system. This paper proposes to generate synthetic daily streamflows at the plant site, using regional information, censuring these at plant capacity, computing the mean and standard deviation of the censured flows and then applying Fill’s method. A Second-Order Shot Noise model (Weiss, 1977) was used for streamflow generation. A case study is performed and the results are compared to the method proposed by ANEEL. Key-words: Guaranteed increment energy, Small Hydroplants, Regionalization Method, Synthetic series of daily mean flows.

Impact of climate change on hydropower production within the La Plata Basin

ABSTRACT This paper aims to estimate the variation of the combined dependable energy output of the set of major hydropower plants within the Brazilian part of the La Plata Basin due to possible climate changes during the twenty-first century. It uses and compares the predictions of two regional climate models, namely PROMES [Castro, M., Fernández, C., and Gaertner, M.A., 1993. Description of a mesoscale atmospheric numerical model. In: J.I. Díaz and J.L. Lions, eds. Mathematics, climate and environment. Rech. Math. Appl. Ser. Mason, 230–253; Gallardo, A., Galvan, C., and Mermejo, R., 2012. PROMES-MOSLEF: An atmosphere-ocean coupled regional model. Coupling and preliminary results over the Mediterranean basin. 4th HYMEX Workshop 2 2010] and RCA models [Rummukainem, M., 2010. State-of-the-art with regional climate models. WIREs Climate Change, 1, 82–96]. Rainfall and temperature predictions are converted into streamflow at key gauge stations using Variable Infiltration Capacity Model [Liang, X., Lettenmaier, D.P., Wood, E.F., and Burges, E.F., 1994. A simple hydrologically based model of land surface water and energy fluxes for general circulation models. Journal of Geophysical Research, 99, n. D7, 14,451–14,428]. The evaluation of the dependable energy output used the natural energy hydrograph method engineering consultants (Canambra Engineering Consultants, 1969. Power study of South Brazil. 13 v. Appendice XVII Final Report. Rio de Janeiro: Canambra Engineering Consultants), combined with the Monte Carlo simulation of synthetic series of natural energy. The main contribution of this paper is the consolidation of a methodology that provides estimates of the systems dependable energy as a function of the return period for both observed and future predicted streamflows. As a conclusion, a reduction of the dependable energy output of the hydropower plants within the La Plata Basin could be expected during the twenty-first century

Quantis de cheia GEV Regionais - Uma aplicação prática

This paper presents an application of f lood quanti le est imators using GEV (Generalized Extreme Value) distribution and PWM (Probabi l i ty Weighted Moments) estimation combined with regional analysis at Tibagi River in Brazil. A new flood quantile estimator called NQR (GLS) is used for the first time. This estimator may be combined with a 2P/AR estimator (Krüger, 1996; Krüger and Fill, 1996) resulting in a Bayesian Linear Estimator which is an extension of those proposed by Fill (1994), Krüger (1996) and Krüger and Fill (1996).