The statistical exploratory data analysis must be completed by exploring the spatial distribution of the data. In space, stationarity needs to be assumed in order to allow for statis tical inference.
The behavior of the variables in space needs to be understood in order to make the stationary decision and understand its potential drawbacks. On the other hand, we also need to define domains that will be used for esti mation or simulation.
These domains define the set of data used to infer the variable at unsampled locations within that same volume. Therefore, a representative distribution must be available within each domain to condition the inference process.
The statistical distribution of the variables must be characterized within each domain, but oftentimes, the sam ples are not spatially representative, biasing the statistics that can be inferred from them.
Download link: https:04-Geostatistics Spatial Data Analysis
Credits to Julián M. Ortiz
