Geospatial modelling revolves around the structures of data and the semantics of these structures. This is enough in simple cases, but becomes insufficient when the best structure and semantics is hard to find or the solution is too heterogeneous to fix and reuse. Field-based and objects-based geospatial models often share common GIS data structures interchangeably, but their all possible meanings are too many to define in an immutable manner. Less studied approach to geospatial modeling is using mutable structural properties and their semantic interpretation. This work shows that the functional aspect of geospatial models is just as important as the structural and semantic aspects. It also shows that semantic and even structural properties may change when functionality is integral part of the data model, and not exclusively separated at software implementation level. The paper uses this modelling paradigm to address the divide caused by field-based and object-based data models, and other challenges regarding synergy of geospatial systems that need to use both types of data models.