Combination of Earth observation data with other types of geospatial data is highly recommended since existing information can be essential for improving the interpretation of remote sensing data. The automatic classification of agricultural crops from multispectral image data is such an example. Pixel-by-pixel classification usually gives many errors, owing to sensor noise and field heterogeneity. However, if parcel boundaries are known from a GIS database, one can classify all pixels within a field as a group, which reduces the number of misclassifications enormously, provided, of course, that the group as a whole is correctly classified.
Data can be integrated in an almost infinite number of ways. Results from data integration can, again, be combined with other geospatial data to produce yet other new information, and so on.
Explain and be able to apply basic vector and raster spatial data structures including selecting a suitable data structure for geographic phenomena (level 1, 2 and 3).
Explain the basic concepts of data retrieval (attribute and spatial queries) and formulate queries to make a selection on attributes and geospatial data from a spatial database.
Student is aware of the fact that workflows can be stored and executed in ModelBuilder (level 1).
Name different types of spatial process models and their characteristics including the coupling between GIS and models (level 1).