Resumo

CAMPOS-ARANDA, Daniel F.. Regional fit of GEV distribution in 34 pluviometric stations of the Huasteca area in San Luis Potosi State, Mexico. Agrociencia [online]. 2008, vol.42, n.1, pp.57-70. ISSN 2521-9766.

The regional probabilistic analysis of annual maximum daily rainfall represents the base of the estimations of flooding when there is no hydrometric information. The regional methods define more precisely the right tail of the distribution of probabilities, which makes it possible rainfall predictions. The present study is based on information of 34 pluviometric stations of the Huasteca area in San Luis Potosí, which have over 30 years of records. The methodology consists in applying four procedures of regional fit of the General Extreme Values distribution (GEV), based on the method of the L moments: 1) stations-years method; 2) fit by moments of weighted heavy probability; 3) method of the median standardized values; 4) fit by regional L moments. It is concluded that the regional methods shown are reliable and that both procedures developed (3 and 4) make it possible to restrict the predictions.

Palavras-chave : Regional methods; L moments; maximum daily rainfall.