A paper map answers the question you had already thought of. A geographic information system answers the ones you haven't — and then draws the answer for you. That, more or less, is the whole idea.

GIS stands for geographic information system, and the name is the least helpful thing about it. It doesn't describe one piece of software you can buy and install. It describes a way of handling information that has a place attached to it. Postcodes, bus stops, drainage pipes, footpaths, school catchments, reported fly-tipping: give each record a position on the Earth, and you can start asking questions a spreadsheet alone would struggle with.

Think in layers, not pictures

Take away the colours and labels from any decent town map and you find a stack of separate layers. Roads sit on one. Building footprints sit on another. Then a river, an administrative boundary, a patch of tree canopy, a scatter of bus stops. Each layer is stored independently, with its own set of rules about how it looks and what it means.

The old analogy is tracing paper: sheet over sheet over sheet, each one adding something to the picture. The analogy holds because it captures the useful part — you can lift a sheet off, swap it for another, or draw a new one without redrawing the rest. In software, layers can also carry data behind them. A polygon representing a field isn't just a shape; it can have a name, an area, a land-use class and a last-surveyed date stored alongside it.

Two kinds of spatial data

Nearly everything you load into a GIS arrives in one of two forms, and knowing which you're dealing with saves a lot of confusion.

• Vector — discrete features drawn as points, lines or polygons. A bench, a lane, a parish boundary. Each feature has a row in an attribute table, which is where the descriptive information lives.
• Raster — a grid of cells, like a photograph or a spreadsheet of squares. Elevation models, satellite imagery and rainfall surfaces are usually raster. Each cell holds a value rather than a shape.

Both can be styled, queried and combined. Vector is better for things with edges and identities; raster is better for things that vary continuously across a surface.

The questions a GIS is built for

Under all the toolbars, a GIS does five or six jobs over and over. If you can name which one you need, the software becomes far less intimidating.

1. Where is it? Plot a list of addresses and look at the pattern.
2. What is here? Click a point and read everything that overlaps it.
3. How many are within…? Count features inside a boundary, or within 500 metres of a school gate.
4. What is nearby? Find the closest pharmacy to each of forty villages.
5. What has changed? Compare two dates of the same layer and measure the difference.
6. What is the best route? Work out a path along a network with rules attached, such as one-way streets or weight limits.

The tools behind those questions have dull names and useful behaviour. A buffer draws a zone at a fixed distance around a feature. A clip trims one layer to the outline of another. An overlay compares two layers and keeps only what they share. A spatial join attaches the attributes of one layer to the features of another based on where they sit. Learn those four and you can do a surprising amount of real work.

Where the data comes from

Most beginners stall here, assuming they need to buy something. They usually don't. OpenStreetMap is a global, crowd-built dataset you can download by region. National mapping agencies in the UK publish open datasets covering boundaries, roads and place names. Local councils release bins, parking, planning and air-quality layers. Satellite and aerial imagery is available at no cost from several public programmes, and elevation data is often free too.

Two habits are worth building early. First, read the licence before you publish anything — some data is free to use, some is free with attribution, some is not free at all. Second, read the metadata: who collected it, when, at what scale, and with what accuracy. A boundary drawn for a national overview will look wrong when you zoom to a single street, and that isn't a bug.

Coordinates trip people up as well. Data can be stored in different coordinate reference systems, and layers in mismatched systems will refuse to line up. If two layers land in the wrong place, the projection is the first thing to check, not the last.

GIS is not the same as cartography

They overlap heavily, but they aren't twins. A GIS stores, questions and analyses spatial data. Cartography is the craft of making the finished map readable — hierarchy, colour, label placement, honest scales. You can be excellent at one and shaky at the other.

A sound analysis displayed badly will mislead people just as effectively as a wrong analysis displayed beautifully.

That's why the design stage matters. Decide the message, choose the layers that support it, and leave the rest out. A map showing everything usually says nothing.

A sensible first project

Skip the tutorials that ask you to map a fictional country. Pick somewhere you know on foot and a question you actually care about: which streets are more than ten minutes' walk from a park, or where new tree planting would shade the most pavement.

1. Write the question down in one sentence. Be specific about the area and the measure.
2. Install QGIS. It is free, open source and runs on ordinary laptops.
3. Download two layers that help answer it — a boundary and a set of features, for example.
4. Load them, check they align, and style them simply: grey background, one strong colour for the thing you care about.
5. Run one operation. A buffer, or a count within an area. Nothing more.
6. Export a map as an image, then look at it again the next morning before you show anyone.

Expect the first attempt to be scrappy. The second will be better, and by the third you'll be thinking in layers without noticing.

Start with a question, not a map

The most common beginner's mistake is opening the software and wondering what to draw. Work the other way round. Decide what you want to know, then ask which layers could answer it, then check whether those layers exist and how trustworthy they are. The map is the last step, not the first.

It also helps to keep one small, unglamorous rule in mind: write down each step as you go. Spatial work involves so many small decisions — filters, distances, dates, projections — that a project assembled over a fortnight becomes impossible to explain a month later. A few lines in a notebook, or a text file beside your data, will make your work repeatable and your conclusions defensible. That habit, more than any particular tool, is what separates a map enthusiast from someone who genuinely works with GIS.

Photo: Brett Jordan / Pexels