Interpolation technique. An automated approach was applied
during the 78,840 interpolation procedures required in this study,
which were composed of each hour of the year multiplied by the
number of units plus the baseline. Hiemstra et al. (2009) developed
an automated kriging procedure for interpolating air pollution
fields, which they implemented in the R statistical programming
language (R Core Team, 2014). The automation occurs by fitting the
variogram automatically using a sill and range estimated from the
data. We extended this approach by testing for east-west and/or
north-south trends in the data with a linear regression model.
When at least one significant trend occurred, we applied universal
kriging (kriging with external drift) as the interpolation method
with the significant linear dependency for the drift component. If
more than one linear model was significant, we chose the model
with the highest R2
. We applied ordinary kriging when no linear
dependencies were significant
Interpolation technique. An automated approach was appliedduring the 78,840 interpolation procedures required in this study,which were composed of each hour of the year multiplied by thenumber of units plus the baseline. Hiemstra et al. (2009) developedan automated kriging procedure for interpolating air pollutionfields, which they implemented in the R statistical programminglanguage (R Core Team, 2014). The automation occurs by fitting thevariogram automatically using a sill and range estimated from thedata. We extended this approach by testing for east-west and/ornorth-south trends in the data with a linear regression model.When at least one significant trend occurred, we applied universalkriging (kriging with external drift) as the interpolation methodwith the significant linear dependency for the drift component. Ifmore than one linear model was significant, we chose the modelwith the highest R2. We applied ordinary kriging when no lineardependencies were significant
การแปล กรุณารอสักครู่..