The Four Fundamental Methods of Registration
Whether you are conducting an exploratory pilot study or the final, structured survey, you will use one of four fundamental methods to capture data about the world. These methods are universal. What changes between the pilot and the final registration is your goal, your tools, and your level of precision.
-
In the pilot study, your goal is exploration. You use unstructured tools (notebooks, paper maps ) to test your ontology and find its weaknesses.
-
In the final registration, your goal is production. You use structured tools (survey apps, pre-printed forms ) to apply your finalised ontology consistently and create clean data.
1. Inventorying: Capturing the State of Features
This is the act of creating a record for discrete features—be they points, lines, or areas—noting their location and properties at a specific point in time. The goal is to produce a static snapshot of what exists in a given location.
-
During the Pilot: You’ll use a notebook and map to list the features you find, like park benches. You might write, “Found a strange concrete bench, not in my schema. What is it?” This is a successful test of your ontology.
-
During Final Registration: You will use a survey app with a pre-defined form. For each bench, you’ll create a new point feature and fill in its attributes (
material,condition) from a dropdown list. For each bench, you’ll create a new point feature and fill in its attributes (material: 'wood',condition: 'good'). Similarly, for a park polygon, you might record attributes likearea: 5.2andbench_count: 15
2. Delineating: Defining Area Boundaries
This is the act of defining an area by drawing its boundary according to a set of classification rules. Unlike inventorying, which captures discrete, tangible objects, delimiting is used for conceptual areas where the boundary itself is a matter of interpretation. A classic example is defining the border between the “picnic area” and the “sports area” in a public park.
-
During the Pilot: You’ll use a pencil to sketch a boundary on a paper map. The goal isn’t a perfect line; it’s to find where your rules get fuzzy. If you can’t decide where the boundary between “park” and “wild area” goes, make a note! That ambiguity is a valuable discovery.
-
During Final Registration: You’ll use a GPS-enabled app or more typically, sit at a desktop GIS using aerial photos and field notes to create a precise polygon in what is called backdrop digitalisation. You are no longer questioning the rules but applying them consistently to produce a clean, final feature.
3. Tracing: Recording a Process or Event Over Time
This is the act of capturing the change in a feature’s location or extent over a period of time. In contrast to an inventory’s static snapshot, tracing records a dynamic process like movement, flow, or growth.
-
During the Pilot: You’ll sketch a path of a car, pedestrian or animal on your map. The sketch is a rough record used to understand the nature of the path itself.
-
During Final Registration: You monitor the entire area from a vantage point, record the tracks, walk the path tracking the movement or attach a GPS tracker to the object being tracked. Create, accurate LineString feature. The goal is a precise geometric representation of the event.
4. Monitoring: Recording Events Over Time
This is the act of repeatedly counting occurrences of an event at a fixed location to capture its frequency, rhythm, or flow. Unlike an inventory’s static snapshot, monitoring is designed to measure dynamics and change. The result is typically a table of counts corresponding to different time intervals.
-
During the Pilot: You’ll stand at a street corner with a tally counter for 15 minutes. You might note: “At 8:05 AM, a delivery truck blocked the intersection, stopping all cyclists. My count is biased.” You are testing the feasibility and potential problems of your monitoring protocol.
-
During Final Registration: A permanent traffic sensor automatically records the number of cars that pass each hour. The data is a clean, structured time-series (e.g.,
location_id: 101, timestamp: '2025-09-08T09:00:00', car_count: 254).
Representing observations: The Spatial Table
Once we have performed our observations—whether by counting, tracing, delineating, or mapping—we need a systematic way to store this information. The universal structure for organizing data is the table. In GIS, we use a special kind of table that has a “spatial awareness.”
At its core, all vector-based geospatial data is stored in tables where each row represents a single observed feature (like a specific tree, a single road, or one park) and each column represents an attribute (a characteristic of that feature).
The Anatomy of a Spatial Table
A spatial table has three key components:
-
Rows (or Records): Each row in the table corresponds to a single, discrete feature that was observed in the real world. If you have a dataset of trees, each row represents one individual tree.
-
Attribute Columns (or Fields): These are the standard columns you would find in any table. Each column describes a specific property of the features. For our trees, we might have columns for species, height_meters, and condition.
-
The Geometry Column: This is the special column that makes a table spatial. It stores the “where” information for each feature. For each row, the geometry column contains the coordinate data that defines the feature’s shape—whether it’s a point (for a tree), a line (for a road), or a polygon (for a park).
Let’s look at a simple example of a tree inventory:
| feature_id | species | height_meters | condition | geometry |
| 1 | Oak | 15.2 | Good | POINT (55.7 12.5) |
| 2 | Birch | 11.8 | Fair | POINT (55.8 12.4) |
| 3 | Oak | 18.5 | Good | POINT (55.7 12.6) |
In GIS software, this fundamental structure—a table containing a geometry column—is often referred to as a layer or a feature class. But it’s crucial to remember that underneath these labels is the simple, powerful concept of a spatial table.
The Power of the Table: The Relational Model
The true power of organizing our observations into a spatial table lies in its simple, predictable structure. This clean format isn’t just for storage; it’s a foundation for action. It allows us to perform powerful operations directly on our data, such as filtering the table to show only the trees in ‘Good’ condition or combining it with other tables to discover new insights.
These fundamental operations—selecting, transforming, and combining data—are not just ad-hoc tricks. They are formalised in a branch of mathematics known as relational algebra. The relational data model, which you’ll find at the heart of nearly all modern database systems and GIS software, is the practical implementation of these mathematical principles. 🧠
Therefore, by structuring our observations into a simple table, we unlock the ability to perform complex, repeatable, and logically sound analysis. It’s this robust, formalised approach that elevates GIS from a simple map-making tool to a true analytical science.