OE-6.5 Process of GIS & Geospatial Analysis

What this is and why it exists

Overlay analysis is the operation that made this field famous. Stack layers — soil, slope, land use — and the intersection answers a question no single layer could.

Network analysis is the machinery behind every routing application you have used. It is the most familiar idea here, arriving under an unfamiliar name.

And the integration section is the point of the whole course. Remote sensing supplies the layers; this supplies the questions.

The vocabulary

• Encoding — getting data into the system in a usable form.
• Query — selecting records by attribute, by location, or by both.
• Geospatial measurement — distance, area, perimeter computed on the ground rather than on paper.
• Overlay — combining layers so that their intersections can be examined.
• Buffer — a zone at a fixed distance from a feature.
• Network analysis — routing and reachability over a connected set of lines.
• Surface analysis — deriving slope, aspect and visibility from elevation.
• Integration — using sensed imagery as a layer alongside recorded data.

The mental model

Encoding comes first and is where most of the effort actually goes. Raster data arrives as grids, vector data as coordinates, and attributes as tables. All three must agree about position, projection and units before anything else is possible. Most analysis failures in this field are encoding failures wearing a different hat.

Query is then the ordinary database operation with a spatial half added. Select by attribute, by location, or both together. Parcels zoned residential and within two hundred metres of a river is one query rather than a project.

Measurement is not as simple as it looks, and it is worth one caution. Distances and areas depend on the projection you are using, because a curved surface cannot be flattened without distorting something. Two answers that disagree are usually two projections rather than an error.

Overlay is the famous one. Stack layers and examine where they intersect. Suitability analysis is overlay. Where could this be built, given slope, soil, flood risk and existing use? The answer exists in no single layer. Buffers extend the same idea to proximity questions.

Network analysis treats lines as a connected graph and asks for shortest paths, service areas and reachability. That is the graph material from the Programming area, applied to roads. Recognising it as the same thing is worth more than learning it twice.

Surface analysis derives new information from elevation: slope, aspect, and what is visible from where. All three are computed from the same grid, which is a good illustration of why elevation is stored as a raster.

And integration closes the course. Sensed imagery becomes another layer alongside recorded data, so a classification derived from a satellite can be overlaid on cadastral boundaries. The two halves of the syllabus are one workflow, and this is where that becomes visible.

What you should now be able to explain or do

Encode raster, vector and attribute data and get them to agree on position and units. Write a query with an attribute half and a spatial half. Say why two area measurements can legitimately disagree. Perform an overlay and say why the answer exists in no single layer. Recognise network analysis as graph work you have already met. Derive slope and visibility from elevation. Combine sensed imagery with recorded data.

Check yourself