3.2.3 "Knowledge Hierarchy": Features, Interpretations, Representations , and Properties

Topic Version1Published09/11/2015
For StandardRESQML v2.0.1

RESQML has now expanded the concept of "data" to include the relationships between data objects, which allows a more precise classification. The current design now supports the transfer of abstract subsurface features, human interpretations of those features, the data representations of those interpretations, and the properties indexed onto those representations, which results in a well-defined knowledge hierarchy of feature/interpretation/representation/properties (informally referred to as "FIRP")—a key organizing concept in RESQML. Additionally geometry is integral to RESQML organization. (Geometry is also an important part of the RESQML data model.)

The table below defines these terms; each of these is a type of RESQML data object.




Something that has physical existence at some point during the exploration, development, production, or abandonment of a reservoir. For example: It can be a boundary, a rock volume, a basin area, but also extends to a drilled well, a drilling rig, an injected or produced fluid, or a 2D, 3D, or 4D seismic survey.

Features are divided into these categories:

  • Geological, for objects that exist a priori, in the natural world, for example, rock formations and how they are positioned
  • Technical, for objects that exist by the actions of humans, for example, wells.


RESQML uses the definition of David Gawith, which explains an interpretation as a single consistent description of a feature. An interpretation is subjective and very strongly tied to the intellectual activity of the project team members. The initial curiosity and reasoning of the people on the project team initiates the early pre-screening campaign (remote sensing, surveys). They make hypotheses that consist of as many interpretations as necessary to describe the features. (Gawith and Gutteridge 2007; for citation, see 5.5 References .

NOTE: The RESQML formal name is actually "feature-interpretation" and some of the class names use this full term. For conciseness of documentation, we use simply "Interpretation" where this usage is not confusing.


A digital description of a feature or an interpretation. For example, currently in RESQML, a horizon interpretation may be represented by a point set, a set of triangulated surfaces, or a set of orthogonal grids.

A representation contains the topology and the geometry of the data objects. These terms are defined in Section 3.2.4 Topology and Geometry

NOTE: Like "feature-interpretation" in the previous table row, this term is formally "feature-interpretation-representation", but we use "representation" for conciseness.


A property can be attached to any indexable element of any representation.

Properties refer to semantic variables (for example, porosity, permeability, etc.) and the corresponding data values, which are recorded in arrays, which may be stored in HDF5 datasets.

For more information about properties, see 8 Properties .