Learning GIS takes time—hands-on work, experimentation, and a clear sense of what your spatial data can tell you. While you’ll build much of your expertise through practice, this page is designed to guide you through the first steps of spatial analysis. While most tutorials use QGIS as software, the information in the articles apply to entirety of the GIS world. Think of it as an interactive map of options: you are guided by questions about your data, and the page helps you find the best GIS tools and workflows for your demands. If your data is purely attribute-based (no spatial component), this might not be the right place. In this case you might want try out the statistics landscape instead. But if you have spatial features (points, lines, polygons, raster, coordinates etc.),you’ve come to the right place.

This decision tree is an introduction for GIS and serves as an orientation to steer the space of spatial sciences. Feel encouraged to go deeper by yourself using this as a basis and the various provided information as stepping stone. To help you start your spatial journey, it is recommended to read the introductory article about Geographical Information Systems.

In case you wanto know how to how to install QGIS and start working with it, click here.

Go through the images step by step, click on the answers that apply to your data, and let the page guide you.

Simple Map[edit]

Start here with your data! This is your first question.

Types of spatial data[edit]

Spatial data is the foundation of Geographic Information Systems (GIS), linking information to specific locations on the Earth’s surface so patterns and relationships can be visualized and analyzed. It comes in two main forms: vector data, which represents discrete features like roads, buildings, and boundaries using points, lines, and polygons; and raster data, which represents continuous surfaces such as elevation, temperature, or imagery through grids of cells (Read article linked below for further info). By combining spatial data with descriptive attribute information, GIS allows users to understand spatial patterns, model real-world processes, and support informed decision-making across many fields.

Which type of spatial data do you have?

How do i know whether my data is vector or raster?

• While vector data mostly are composed of different shapes of lines, points or polygons and contain an attribute table, raster data is a grid of cells, each having only one value. Here is an article describing both data formats.

Beyond this, you might want to know more about Projections.

Vector data[edit]

You have vector data. Is your vector data (e.g. sample coordinates) in a table format, such as excel or csv?

Good to know: Vector data can have attribute tables similar to an excel sheet. Variables are in columns. Rows represent each object of your data, also called feature. Here are articles about attributes and about filtering.

Shapes[edit]

In GIS, vector data is represented through three basic shapes—points, lines, and polygons—each used to model different types of real-world features with precision. What shapes does your vector data have?

Polygons[edit]

Do you want to work with polygon layers to combine, compare, or extract spatial data based on their shapes and locations?

Why would i use polygons for ground truth?

Polygons are used for ground truth in QGIS because they capture the full extent and shape of real-world areas, making them ideal for training and validating land-cover classifications, collecting accurate area-based statistics, and performing spatial analyses such as overlays or zonal statistics. Unlike points or lines, polygons allow algorithms to sample all pixels within an area, improving classification accuracy and providing richer, more reliable data for GIS workflows.

Processing polygons[edit]

Do you want to combine multiple polygon data sets to analyse attributes for intersecting or shared areas? There are multiple options depending on what you want to achieve. Do you want analyse attributes for shared areas or do you want to stack layers or do you simple want to limit your data to a specific subset using another polygon (area)?

How do i know which tool to use

To choose the right tool in QGIS, first decide what you want to accomplish—editing data, analyzing spatial relationships, cleaning geometry, converting formats, or doing raster analysis—because each goal points to specific tools like Buffer, Clip, Dissolve, Spatial Join, or Raster Calculator. The fastest way to find tools is by using the Processing Toolbox search bar, and you can check each tool’s built-in help for explanations and examples. Also make sure you know whether you're working with vector or raster data, since different tools apply to each. If unsure, rely on a simple task-to-tool cheat sheet: for example, use Select by Location for spatial selections, Join Attributes by Location for summaries, Fix Geometries for errors, and Reproject Layer for coordinate conversions.

Joins & Intersection[edit]

Do you want to statistically analyse features and attributes of your spatial data?

Extraction[edit]

Extraction of data in GIS involves isolating specific spatial features or attribute information from larger datasets, allowing users to focus on the most relevant geographic elements for analysis and decision-making. Do you want to extract raster values using the shape or boundary of a vector layer?

Lines[edit]

Do you want to measure distances or apply more complex analysis?

What are lines in spatial data?

Points[edit]

Do you want to show simple locations or further process them?

Point processing[edit]

Do you want to group points based on attributes, values or locations or do you rather want to estimate spatial values between and outside known data points?

Extra- & Interpolation[edit]

It seems you want use your point data to extra- or interpolate from them? If not, you may want to use them as validation points, e.g. to test how accuracte a (raster) classification represents the real world.

Area statistics[edit]

Do you want to calculate simple statistics or group and categorise data depending on attribute values?

Raster & Vector data[edit]

In GIS, both raster and vector data can be analyzed together using a wide range of functions, from spatial queries and overlay operations for vector features to map algebra and surface analysis for raster datasets. This enabe powerful exploration and interpretation of geographic patterns. The raster and vector data can also be analysed as a combined unit for extracting certain information.

Does your (combined vector and raster) data represent topographic height values or attributes?

Raster data[edit]

Does your raster data represent topographic height values or attributes?

Imagery[edit]

Do you have aerial photos or paper maps, that are not in a geoinformation data format or do not have an assigned coordinate reference system?

Paper info[edit]

Do you want to align a scanned map or image to real world coordinates so it can be used with other GIS data?

Satellite data[edit]

What is satellite data? Here is an article describing the fundamentals of Satellite Data.
Do you have finer or coarser spatial resolution?

How do i know which resolution i have or which is relevant?

• Check out Spatial, Spectral, and Temporal Resolution as the main properties of satellite and remote sensing data.

Finer resolution[edit]

Which processing path do you want to follow for working with finer grained raster imagery?

Classification[edit]

Do you want to perform classification using supervised training samples or through unsupervised, automatic class assignment?

How do I know which one is better suited for my task?

• Supervised classification usually prodices higher accuracy and you have clear control over class definitions, due to the utilization of training data with known, labeled samples. Choose unsupervised classification when you need an exploratory approach that automatically groups pixels by their spectral patterns, especially when class boundaries are unknown or training data is lacking.

Accuracy assessment of raster data in GIS involves evaluating how closely the mapped values represent real-world conditions, typically by comparing them to reference or ground-truth data. Both methods require Accuracy assessment and Ground truth validation.

Coarser resolution[edit]

Do you want to analyze changes over time?

Changes over time[edit]

Do you only have few dates?

Change detection & Comparison maps[edit]

Do you want to analyze different dates to detect and measure changes in features over time or just make a simple comparison of two maps?

Time series[edit]

You might want to try out a Time line. Do you want to animate your timeline or analyze the statistics of temporal changes?

Statistics of time series[edit]

Do you want to statistically analyze how geographic data changes over time or use the data to predict future trends?

Digital elevation models[edit]

Why and when would I need a DEM?

A Digital Elevation Model (DEM) is a digital representation of the Earth's surface topography, stored as a grid of elevation values. It captures the height of terrain features without vegetation or buildings. DEMs are widely used in geographic information systems (GIS) for applications such as hydrological modeling, terrain analysis, infrastructure planning, and environmental studies. They help visualize landscapes, analyze slopes, and simulate water flow patterns accurately.

Good to know

• While a DEM is a general elevation model, a Digital Surface Model (DSM) captures the heights of the land plus objects on it (trees, buildings), in other words all surfaces. A Digital Topographic Model (DTM) represents only the bare-earth terrain with all surface objects removed. DTM is often used interchangeably with DEM.