stmra_cov_nonstationary_R2_separable: Nonstationary spatiotemporal covariance function varying by...

View source: R/cov.R

stmra_cov_nonstationary_R2_separableR Documentation

Nonstationary spatiotemporal covariance function varying by latitude

Description

Compute pairwise covariances between two sets of locations using a nonstationary kernel convolution approach with standard deviation and anisotropy varying as weighted sum of Gaussian radial basis functions cenetered at different latitudes.

Usage

stmra_cov_nonstationary_R2_separable(locs1, locs2, theta)

Arguments

locs1

matrix with rows representing points and columns representing coordinates

locs2

matrix with rows representing points and columns representing coordinates

theta

vector of parameters (see Details)

Details

Parameter vector theta includes spatial nugget, temporal nugget, temporal range, standard deviation by latitude (9 values), spatial range in east-west direction (9 values), spatial range in south-north direction (9 values) (in this order).

References

Paciorek, C. J., & Schervish, M. J. (2006). Spatial modelling using a new class of nonstationary covariance functions. Environmetrics: The official journal of the International Environmetrics Society, 17(5), 483-506.


appelmar/stmra documentation built on April 12, 2022, 7:22 p.m.