Serviços Personalizados
Journal
Artigo
Indicadores
- Citado por SciELO
- Acessos
Links relacionados
- Similares em SciELO
Compartilhar
Revista mexicana de ciencias agrícolas
versão impressa ISSN 2007-0934
Resumo
ROMANTCHIK KRIUCHKOVA, Eugenio; ARTEAGA-RAMIREZ, Ramón; CERVANTES OSORNIO, Rocío e VAZQUEZ-PENA, Mario A.. Models to predict probabilistic precipitation in Tabasco, Mexico generated with published information. Rev. Mex. Cienc. Agríc [online]. 2018, vol.9, n.spe21, pp.4341-4354. ISSN 2007-0934. https://doi.org/10.29312/remexca.v0i21.1535.
The weather stations usually present data lost in their records, which complicates probabilistic studies in this case of precipitation. But there is published information obtained for this purpose, so the objective was to generate models to predict probabilistic precipitation in the state of Tabasco with published information. There was information published graphically of 19 stations in the state of Tabasco, of these the average precipitation was taken and the probabilistic precipitation was generated at levels of 80, 60, 40 and 20%, the simple linear model was used and four models were generated to estimate the probabilistic precipitation at the indicated levels based on the average rainfall with data from 17 stations, the other two were used in the validation of the models. To define the predictive goodness of these, the square root of the mean square of the error (RCCME) was used. The four generated models presented good adjustment, since their coefficients of determination were 0.959, 0.985, 0.991 and 0.97, in the probability levels of 80, 60, 40 and 20% respectively. The values of the RCCME varied from 4.6 to 27.7 mm which indicates that the models are good predictors.
Palavras-chave : precipitation estimation; linear model; square root of the mean error square and crop zoning.