Modeling the frequency distribution of precipitation for Agreste Meridional in the State of Pernambuco

• Autor(a) principal: Andrade, Antonio Ricardo Santos de
• Data de Publicação: 2020
• Outros Autores: Souza, Luciano, Silva, Edijailson Gonçalves da, Andrade, Emylle Kerolayne Palmeira de, Costa, Claudia Machado, Silva, Maria Gorete dos Santos, Silva, Jéssica Dayana de Souza, Cruz, Adiel Felipe da Silva, Santos, Wllias Mendonça dos, Ferreira, Maria Emanuely da Silva
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Research, Society and Development
• Texto Completo: https://rsdjournal.org/index.php/rsd/article/view/9899
• topic: Modelagem; Distribuição de probabilidade; Precipitação.; Modelado; Distribución de probabilidade; Precipitación.; Modeling; Probability distribution; Precipitation.

The characterization of climate change in a region allows determining planning actions for future agricultural activities. Therefore, the objective was to model the rainfall in 2017 for the Agreste Meridional of Pernambuco, Brazil, from probabilistic distributions. The quality of fit and adherence of different probability functions (Cos-Weibull, Weibull-Exponential, Kumaraswamy Weibull and Kumaraswamy Weibull Poisson and Gumbel) were analyzed. To verify the adjustments, the criteria were determined the statisticians of AIC, BIC and HQIC, in addition to the tests of Anderson Darling (AD) and Cramér-von Misses (CVM). The study area consists of 71 municipalities distributed in six regions of Agreste Pernambucano and is inserted in the coverage area called "areas subject to drought", which has a drought period lower than the sertão. For the elaboration of this work, we used average annual rainfall data of 2017 of the 71 meteorological stations (municipalities), acquired from the Water and Climate Agency of Pernambucana (APAC) and the National Water Agency (ANA). The five probability functions resulted in suitable and good adjustments, except for Kumaraswamy Weibull Poisson and Gumbel. However, the results indicated that the two-parameter Cos-Weibull distribution was more appropriately adjusted to the rainfall conditions of the six regions of Agreste Pernambucano, followed by the Weibull-Exponential, Kumweibul and Kumwpoisson distributions. For the data in question, the probability function that presented the most accurate result to the rainfall conditions of the six regions of Agreste Pernambucano was that of Cos-Weibull with two parameters, followed by the Weibull-Exponential, Kumweibul and, finally, Kumaraswamy Weibull Poisson.