12

When using the term "feature" in GIS, what is often (if not always) meant is a geographical feature which "is a representation of real world phenomenon associated with a location relative to the Earth" (ISO 19125-2:2004). Are there non-geographical features and if so, are there any examples of such?

For example, could a compass or a scale bar be considered non-geographical features?

Marcus
  • 123
  • 1
  • 1
  • 6

3 Answers3

10

Features in the most general sense are analagous to records in computer science:

In computer science, records (also called tuples, structs, or compound data) are among the simplest data structures. A record is a value that contains other values, typically in fixed number and sequence and typically indexed by names. The elements of records are usually called fields or members.

So a record can be said to have "attributes", e.g. the records in an "attribute table". One of these attributes may be a spatial data structure describing the geographic (or non-geographic -- not every feature has to exist in the real world) coordinates of the entity represented by the record.

In a GIS, often a distinction is made that a feature represents a spatially-enabled record, however this is not always the case.

For example, in FME (the "Feature Manipulation Engine" by Safe Software), features may be spatial or non-spatial. Indeed, you will very often be working with non-spatial features in that environment, e.g. the rows of a CSV file would be considered features.

Regarding your examples of a compass or scale bar, yes, I think these could be considered non-geographical features, because they don't represent real world entities, but they are themselves entities. I would suggest that they are cartographic features, since they represent distances or directions on a map. As such they are features of a map, not features of the Earth. In another sense, one could argue that digital versions of these elements are indeed records in the computer memory/data structures that make up a map; for example in ArcGIS, these might be represented by elements on a page layout, and when you save the map, these records are persisted in structured storage in the form of a map document (MXD) file.

blah238
  • 35,793
  • 7
  • 94
  • 195
  • I am very content with this answer! Since you were so fast I will however give other people a chance to answer before marking this as accepted. Regarding compass and scale bar being cartographical features, this wiki says that cartographical features are "abstract geographical features". Would you say a scale bar is a cartographical feature, even though it does not have a geographical coordinate? – Marcus Aug 14 '13 at 12:43
  • 2
    My thought is that cartographic (map) features can be both geographic and non-geographic. Geographic map features include things like the equator and lines of latitude and longitude, while non-geographic map features would include things like the map surrounds. This may be a broader definition of "feature" than some would expect, but when working directly with map elements in the context of designing a map layout it is hard to avoid the comparison. Especially if you're doing it with FME :) – blah238 Aug 14 '13 at 13:02
3

According to the "OpenGIS Implementation Specification for Geographic information - Simple feature access - Part 1" (see http://www.opengeospatial.org/standards/sfa ) #4.11 a feature is an "abstraction of real world phenomena. (...) [adapted from ISO 19101]". This backs up the definition of "feature" by SS_Rebelious.

In order to avoid unwanted semantical constraints I prefer to say, that "a feature is an object - like in OO programming - which is an instance of a class, and has (in addition) at least one geometry attribute, which includes a dimension (2D/3D) and a coordinate reference system associated to it."

So a "non-geographical feature", to me, is just an object with no geometry attribute (value).

P.S. I would strictly distinguish "feature (or feature instance)" and "feature type (or feature class)", like "object" and "class" in mainstream IT, like in the OGC "Simple Feaurre Spec." mentioned above.

Stefan
  • 811
  • 7
  • 12
  • 1
    Thanks for the reference. It strikes me that their attempt at defining a feature only pushes back the question to the meaning of "real world phenomena." This is problematic, because I think many people--experts included--might not consider something like a proposed new political boundary or theoretically optimal route to be "real world" or a "phenomenon," yet such things are usually considered "features" in GISes. Another problematic issue is the role of raster-based data in your schema: is a raster a single feature, or are all cells individual features, or are groups of cells features? – whuber Aug 22 '13 at 21:55
  • @whuber, I would say "yes" to each of your questions. Raster data can be interpreted (by a person or by software) to be any of those representations. Each representation is useful in certain contexts. – blah238 Aug 23 '13 at 00:24
  • 1
    I think the OOP comparison is relevant, up to a point. In OOP, an object implies that there is a class that defines the common properties and behaviors of all instances (objects) of that class. Records on the other hand need not derive from any common class or structure (although they typically do in OOP languages and RDBMS's). This, in my mind, makes them more abstract and useful in the context of big data stores, e.g. NoSQL. – blah238 Aug 23 '13 at 00:36
1

To my mind the definition of geographical future as

a representation of real world phenomenon associated with a location relative to the Earth

is not quite correct. If we speculate about 'geographical future' we mean a feature that is the subject of study of the Geography. And a better explanation for the term can be found here:

Geographical features are the components of the Earth. There are two types of geographical features, namely natural geographical features and artificial geographical features. Natural geographical features include but are not limited to landforms and ecosystems. For example, terrain types, bodies of water, natural units (consisting of all plants, animals and micro-organisms in an area functioning together with all of the non-living physical factors of the environment) are natural geographical features. Meanwhile, human settlements, engineered constructs, etc. are types of artificial geographical features.

Notice that in this case geographical feature is a consistent phenomenon (and consistency is present as an attribute for 'geographical feature' in Russian definition of this term)

See, if we will use the definition you quoted - a vehicle with the GPS tracking enabled would be valid geographical feature. But from geographical point of view a single moving vehicle is not a valid geographical feature. For Geography in this case only a spatial distribution of vehicles (e.g. aggregated information) may be a valid geographical feature.

So from geographical point of view a single vehicle (or a flock of birds) is a non-geographical feature. A compass and a scale bar on a map are just a map decorations and may not be considered as geographical or non-geographical features because they are not a real world objects.

EDIT: I see that many people have issues in distinguishing geographical and non-geographical features. It is completely Ok due to scientists have the same issues too))) I was asked to provide additional references to my statements. Unfortunately I have only Russian sources to refer to. There is one of my favourite books: А.Г. Исаченко "Теория и методология географической науки", 2004 (A.G. Isachenko 'Theory and Methodology of the Geography Science') ISBN 5-7695-1693-3. He discuss the term 'geographical feature' at page 27. He states that there is a mix of narrow-minded and scientific approaches to determination of 'geographical feature' definition. He continues that there is no pure scientific definition for 'geographic feature' term. Also he notes that identification of any object on Earth as a 'geographical feature' possible only from ultra-chorological point of view.

Conclusion: there is no clear scientific definition of what is 'geographical feature'. But identification of any object that can be mapped (especially if the map won't be representative few moments later) as a 'geographical future' are allowed only for minority of geographers who shares ultra-chorological point of view (and I have'n seen one of them despite I know many geographers) or for persons who are not familiar with the theory of Geography.

SS_Rebelious
  • 5,621
  • 3
  • 27
  • 62
  • 4
    I disagree with most of these points -- why wouldn't a vehicle or a flock of birds be able to be represented by geographic features? Certainly these are real world phenomena. Even if a particular instance of an entity doesn't exist in the real world, or we don't know precisely where it is, we can still represent it as a geographic feature by "making up" its location. – blah238 Aug 14 '13 at 10:58
  • Just reread my answer. The key word is CONSISTENCY. The AREAL of birds flock IS geographical feature, the flock itself - not. Dead vehicle remaining in place for years - is geographical feature, aggregated and visualised traffic statistics is geographical feature, moving vehicle - not. – SS_Rebelious Aug 14 '13 at 11:53
  • So point and linear features are not geographic features by your definition? Doesn't seem right to me. What exactly is meant by "consistency" in this context? The word has many different meanings in English. I disagree that geographic data equates to aggregate data. Where do you suppose aggregate data comes from? In fact most geographic data in my experience is non-aggregate, e.g. parcels, streets, rain gauges, etc. Why is a moving vehicle not a geographic feature? Perhaps linking to the definition you are referring to would be helpful, even if it is not in English. – blah238 Aug 14 '13 at 12:07
  • Flocks of birds and moving vehicles falls under the umbrealla of Temporal GIS, http://wiki.gis.com/wiki/index.php/Temporal_GIS – artwork21 Aug 14 '13 at 12:15
  • Here is definition in Russian. IDK why you concluded that point and linear features are not geographical objects, it is not what I wrote or meant. I encourage you to read some hand book on Geography (and try to find one that translated from Russian). The geographical feature in general is what can be studied by geographer. Birds are studied by biologists BTW. – SS_Rebelious Aug 14 '13 at 12:15
  • @artwork21, GIS != Geography – SS_Rebelious Aug 14 '13 at 12:22
  • I meant to say flocks of birds and moving vehicles falls under spatio-temporal data not a specific GIS system. – artwork21 Aug 14 '13 at 12:27
  • 3
    I am inclined to agree with @blah238. I do not see why an aircraft stops being a geographical feature as soon as it starts moving. This is the first time I have seen anyone mentioning that consistency is part of the definition (I am interpreting it as static versus dynamic objects that are both relative to the Earth). Do you have any other sources than the Russian wiki-page that support this claim? – Marcus Aug 14 '13 at 12:29
  • 1
    The context of the Russian wiki page definition is the legal use of geographic terms in Russian place-names. Not a very useful definition for geography in general, IMO. – blah238 Aug 14 '13 at 12:41
  • I made edits to my answer to clarify things some more. – SS_Rebelious Aug 14 '13 at 13:16
  • Thanks for the clarification -- it is good to note that there are probably as many interpretations of geography as there are geographers, and to be as inclusive as possible of others' viewpoints and reasoning. – blah238 Aug 14 '13 at 13:29