Title:	Model for multiple imputation to estimate daily rainfall data and filling of faults.
Authors:	CARVALHO, J. R. P. de MONTEIRO, J. E. B. de A. NAKAI, A. M. ASSAD, E. D.
Affiliation:	JOSE RUY PORTO DE CARVALHO, CNPTIA; JOSE EDUARDO B DE ALMEIDA MONTEIRO, CNPTIA; ALAN MASSARU NAKAI, CNPTIA; EDUARDO DELGADO ASSAD, CNPTIA.
Date Issued:	2017
Citation:	Revista Brasileira de Meteorologia, São Paulo, v. 32, n. 4, 575-583, Oct./Dec. 2017.
Description:	Modeling by multiple enchained imputation is an area of growing importance. However, its models and methods are frequently developed for specific applications. In this study the model for multiple imputation was used to estimate daily rainfall data. Daily precipitation records from several meteorological stations were used, obtained from system AGRITEMPO for two homogenous climatic zones. The precipitation values obtained for two dates (Jan. 20th 2005 and May 2nd 2005) using the multiple imputation model were compared with geo-statistics techniques ordinary Kriging and Co-kriging with the altitude as an auxiliary variable. The multiple imputation model was 16% better for the first zone and over 23% for the second one, compared to the rainfall estimation obtained by geo-statistical techniques. The model proved to be a versatile technique, presenting coherent results with the conditions of different zones and times.
NAL Thesaurus:	Meteorological data Geostatistics
Keywords:	Geoestatística Sistema Agritempo Precipitação Modelagem por imputação múltipla Model by multiple imputation Precipitation Ordinary kriging Ordinary Co-kriging Homogeneous zones Missing data Statistical inference
DOI:	http://dx.doi.org/10.1590/0102-7786324006
Notes:	Na publicação: José Eduardo Boffinho Almeida Monteiro.
Type of Material:	Artigo de periódico
Access:	openAccess
Appears in Collections:	Artigo em periódico indexado (CNPTIA)