SciELO - Scientific Electronic Library Online

 
vol.52 issue7Meteorological drought forecasting using discrete Kalman filter in the Fuerte river watershed, MexicoSucrosesucrose and citric acid in preservative solutions with nano silver particles in the vase life of rosa author indexsubject indexsearch form
Home Pagealphabetic serial listing  

Services on Demand

Journal

Article

Indicators

Related links

  • Have no similar articlesSimilars in SciELO

Share


Agrociencia

On-line version ISSN 2521-9766Print version ISSN 1405-3195

Abstract

CAMPOS-ARANDA, Daniel F.. Fitting with mobile L moments of the GEV distribution with variable location and scale parameters. Agrociencia [online]. 2018, vol.52, n.7, pp.933-950. ISSN 2521-9766.

The frequency analysis of annual maximum hydrological data as floods, intensity of rainfall, sea level, wind speeds and daily maximum precipitation, considers that their records are integrated by independent values generated by a stationary random process; because of this, their properties do not change over time. The construction of reservoirs, the effects of urbanization in the basins and the impact of regional climate change result in annual maximum hydrological data series with trends and non-constant variability that make them non-stationary. The objective of this study was to expose the method of mobile L-moments, to estimate the parameters of location (u) and scale (a) variables with time, used as a covariate in the probability distribution function General Extreme Values of type non-stationary (GVE11), with constant shape parameter (km). Through linear or exponential functions, the variation in time of the parameters u and a was plotted to make predictions that are associated with certain probabilities of non-exceedance in the future (years 2050 or 2100). Based on the standard error of fit, the GVE11 distribution was accepted or rejected as a probabilistic model of large series of extreme hydrological data that exhibit non-constant trend and variability. By means of two numerical applications, the practical approach of the mobile L-moments method was shown and, through predictions for the future, its importance and usefulness were highlighted in the probabilistic analyses of large, non-stationary records of the aforementioned type (with non-constant tendency and variability).

Keywords : stationary and non-stationary GVE distributions; linear trend; variability with respect to the mean; mobile L-moments; linear regression; standard error of fit.

        · abstract in Spanish     · text in English | Spanish     · English ( pdf ) | Spanish ( pdf )