Início do conteúdo

The tool referred to as Genetic Algorithm Methodology for IDF (GAM-IDF) was developed for members of the Research Group on Hydrology and Hydrological Modeling in Watersheds from the Federal University of Pelotas, namely: João Pedro de Morais da Silveira, Maíra Martim de Moura, Marcelle Martins Vargas, Samuel Beskow, Tamara Leitzke Caldeira Beskow, and Zandra Almeida da Cunha. 

The main objective is to make available for users a tool characterized by a friendly and intuitive interface, which is founded on state-of-the-art methodologies for the fit of Intensity-Duration-Frequency (IDF) equations. GAM-IDF was fully implemented in the R programming language and consists of two main modules (IDF and Design rainfall) and two informative modules (Technical information and About).

Users can import both rainfall data from the main hydrological database in Brazil (Hidroweb/ANA – Agência Nacional de Águas, file in the .csv format)  and previously defined annual maximum daily rainfall (AMDR) series (.csv, .txt or .xls) into GAM-IDF.

The major strengths of GAM-IDF and its novelties in relation to other existing tools are listed below.

i) users can import rainfall files in the Hidroweb/ANA format into GAM-IDF with no need for previous processing; 

ii) the tools executes a trend analysis in the AMDR series according to the Mann-Kendall test (Mann, 1945; Kendall, 1975);

iii) GAM-IDF enables users to apply multiparameter probability density functions (PDFs) as, for instance, GEV, Kappa and Wakeby, in addition to other PDFs that are widely used in hydrology;

iv) the L-Moments method is employed to estimate the parameters of the PDFs;

v) the fit of the PDFs to the AMDR series is evaluated in accordance with the Anderson-Darling test (Anderson & Darling, 1954), which is more robust than those commonly used for the same type of hydrological analysis (e.g. Kolmogorov-Smirnov and Chi-Square);

vi) GAM-IDF is founded on the Duration Relation Method to disaggregate daily rainfall, allowing users to define either the disaggregation constants proposed by CETESB (1979), which are disseminated in Brazil, or those recommended for other sites; and

vii) the tool makes use of the genetic algorithm named Shuffled Complex Evolution – University of Arizona (SCE-UA) (Duan et al., 1994), which is broadly used for other purposes in hydrology, in order to seek adequate values for the parameters of the IDF equation. 

In addition to the IDF module intended for the fit of parameters of the IDF equation, GAM-IDF also provides a module called Design rainfall that was structured to offer practicality to users. This module makes it available two options: i) visualization of tables containing design rainfall values derived from the fitted IDF for different previously defined combinations of return period and duration; and ii) an IDF calculator where the parameters that were earlier fitted are displayed directly on the IDF equation, which is then able to to compute the design rainfall associated with return period and duration required by the user.

GAM-IDF provides a friendly and intuitive interface and does not require installation. In addition, this tool can be used in computers, tablets, and smartphones. If you wish to use GAM-IDF, please click here.


Click here and check out the places that have used GAM-IDF around the world!