Open AccessConstruction of Spatial local C-vine model and application of air temperature interpolation
Author(s) -
Lin Chao Dai,
Jianxiu Liu
Publication year - 2020
Publication title -
iop conference series. earth and environmental science
Language(s) - English
Resource type - Journals
eISSN - 1755-1307
pISSN - 1755-1315
DOI - 10.1088/1755-1315/585/1/012102
Subject(s) - akaike information criterion , inverse distance weighting , bayesian information criterion , kriging , interpolation (computer graphics) , multivariate interpolation , spatial analysis , statistics , mathematics , computer science , artificial intelligence , bilinear interpolation , motion (physics)
Based on the daily average temperature of 73 meteorological stations and the geographical location information of each station from 1980 to 2010 in Yunnan Province, spatial interpolation method of temperature data suitable for Yunnan province was explored. Based on the definition of spatial local C-Vine, the full model, horizontal distance model and vertical distance model were constructed. Using maximum likelihood, Akaike information criterion (AIC)and Bayesian information criterion (BIC) select the three models to obtain full model as the optimal model. 6 stations in the region were randomly selected and temperature interpolation was made using full model, Ordinary Kriging and Inverse Distance Weighting method. The results showed that full model based on Spatial local C-vine had better prediction accuracy.