Produces the same shape datamatch::accessEnvDat() returns, so the pipeline
runs identically against mock and Copernicus data. Covariates carry the same
latitudinal and seasonal structure planted in the mock abundances, so the model
has something real to learn.
Arguments
- config
a config list, as returned by
load_config()- years
years to generate
- months
months to generate
- resolution
grid spacing in degrees
- seed
random seed
Value
an sf POINT object matching fetch_covariates()'s contract