|
|||||
|
|
Multivariate geostatistical application for climate characterization of Minas Gerais State, Brazil |
|
| Authors: | Luiz G de Carvalho Marcelo de Carvalho Alves Marcelo S de Oliveira Rubens L Vianello Gilberto C Sediyama Luis M T de Carvalho |
| Institution: | 1. Engineering Department (DEG), Federal University of Lavras (UFLA), Cx. 3037, CEP 37200-000, Lavras, Minas Gerais, Brazil 2. UFMT/DSER, Federal University of Mato Grosso, Fernando Correa da Costa Avenue, 2367, Bairro Boa Esperan?a, Cuiabá, Brazil 3. Department of Exact Sciences, UFLA/DEX, Lavras, Brazil 4. National Institute of Meteorology—5th DISME, Contorno Avenue, n.8159, Santo Agostinho, Belo Horizonte, CEP 30110-051-MG, Brazil 5. Engineering Department, Federal University of Vi?osa, P. H. Holfs Avenue, CEP 36571000, Vi?osa, Minas Gerais, Brazil 6. Forest Science Department, UFLA/DCF, Lavras, Brazil |
| Abstract: | The objective of the present study was to assess for Minas Gerais the cokriging methodology, in order to characterize the spatial variability of Thornthwaite annual moisture index, annual rainfall, and average annual air temperature, based on geographical coordinates, altitude, latitude, and longitude. The climatic element data referred to 39 INMET climatic stations located in the state of Minas Gerais and in nearby areas and the covariables altitude, latitude, and longitude to the SRTM digital elevation model. Spatial dependence of data was observed through spherical cross semivariograms and cross covariance models. Box–Cox and log transformation were applied to the positive variables. In these situations, kriged predictions were back-transformed and returned to the same scale as the original data. Trend was removed using global polynomial interpolation. Universal simple cokriging best characterized the climate variables without tendentiousness and with high accuracy and precision when compared to simple cokriging. Considering the satisfactory implementation of universal simple cokriging for the monitoring of climatic elements, this methodology presents enormous potential for the characterization of climate change impact in Minas Gerais state. |
| Keywords: | |
| 本文献已被 SpringerLink 等数据库收录！ | |