Creating a spatio-temporal model using raster data as both dependent and independent data

[Kurt McLaren]

Hi all,

Has anyone ever used or is aware of someone using raster data/raster map as the dependent data and other raster data/raster maps as the predictors in a spatio-temporal inla model? If so, could you direct me to an example of where it was done? I am trying to avoid having to convert the raster data to points, because the dependent variable for each time period (year) would have > 14k points (and I have over 19 years).

Thanks,

Kurt