What is GIS?

Geographical Information Systems (GIS) combines many things such as data analysis, data management and visualisation. Essentially though, GIS all about making data geographically aware. The power of GIS comes in being able to overlay lots of information associated with a location, and work with it in a computational way.

Words and phrases such as georeference, location based, geographic data and spatial all help describe data as being location aware.

We use GIS all the time, perhaps without realising, for example searching for businesses on Google Maps and navigating with a SatNav.

Installing QGIS

QGIS is currently on version 2.12. The tutorials in GISify have been tested with version 2.8.

To get you started with GIS, we are using QGIS. QGIS is free and open source GIS software that will do the majority of geospatial tasks very well. It is cross platform, so you can use it on any computer you like.

Installing QGIS is easy.

Simply follow the steps appropriate to the operating system you are using on the QGIS downloads page in order to install the software.

QGIS interface

Once you have downloaded QGIS, you can now open it for the first time and we can take you through some key features of the user interface.

When you first open QGIS, you will see this:

When we start loading in data, it will be mapped in the blank area on the right. This is known as the canvas. The data structure will be listed on the left hand side under Layers, and the Browser allows us to explore local files on our computers, and also any databases which we have linked to.

The Tutorials

GISify is made up of very small and quick tutorials. Each tutorial will introduce a concept, which will be reinforced through exercises using QGIS.

Tutorial datasets

Of course you will need some geographical datasets in order to experiment with QGIS.

We will be using a variety of types of data, mostly UK based. We’ve developed some great datasets based on aerial photography of Hampshire and satellite imagery of Cornwall.

All data is downloadable via links given in each tutorial.

Get organised

It’s normal when working with GIS that we require and generate large amounts of data. Think about naming conventions, and saving data in an organised, logical way.

Let’s begin

Now you have all the information you need to get started. First up, let’s learn about vector data.

Getting started

Download the following data layers.

• UK coastline
• UK woodland areas
• UK settlements

Unzip your downloaded layers and open a new QGIS project.

This data was sourced from the Ordnance Survey, and it part of their freely available Meridian2 dataset. Check out all the free geographic data released by Ordnance Survey here.

Adding vector data to a QGIS project

Vector data can be stored in a variety of vector data formats. QGIS supports many. In these tutorials we use ESRI Shapefiles, which are made up of multiple files with extensions .prj .shp, .shx and .dbf. This is why we’ve zipped them up for you.

Click on Add Vector Layer.

Navigate to the file called uk-coastline.shp which you have just downloaded and unzipped, and click Open.

You will notice that the data appears mapped in the canvas, and is listed in the Layers section on the left. Vectors can either be points, lines or polygons. This dataset is made up of line features.

Add the polygon data uk-woodlands.shp and the point data uk-settlements.shp in the same way.

Vector styling

QGIS allows us to style vector data. This can be helpful to enhance visualisation and presentation of data.

Right click on uk-coastline in the Layers section and choose Properties. Then choose the Style tab. Experiment changing the layers style using the options in this dialogue. You can do the same for the other layers too, which will have slightly different styling options.

Vector attributes

Vector data allows us to add information to individual features within the datasets, as attributes. We’ll explore this by using the uk-settlements layer.

Right click on uk-settlements and choose Open Attribute Table.

The attribute table displays vector data as a table. In this case, the uk-settlements data has 3 attributes, CODE, OSODR and NAME.

Each point feature is an entry in this table. If we wanted to, we could add further columns (attributes) allowing us to attach more information to each point. Perhaps we might want to add information on population size for example.

Vector labelling

It’s possible to label features using attributes. We’ve seen that the uk-settlements layer contains NAME information.

Open up the layer properties for uk-settlements and choose the Label tab. Set the options in this dialogue such that the layer is labelled using the NAME attribute.

There are a number of further options available to set label positioning, font and scaling rules. Experiment with these.

So far, we have just scratched the surface of what we use vector data for, but it’s enough for now! Moving on to raster data next. Time to look at some images.

Getting started

Download the following data layers.

Rasters

• Satellite image
• Winchester aerial photo
• Hurst Spit aerial photo

The satellite image has been sourced from the Landsat archive, the aerial photos from Hampshire Hub.

Vectors

• UK coastline
• Winchester minor roads
• UK B roads

Unzip the layers and start by adding the uk-coastline.shp to a new QGIS project.

Satellite imagery

Click on Add Raster Layer.

Navigate to the file called landsat-cornwall.tif which you have just downloaded.

Once the layer is added, right click on it in the Layers section and choose Zoom to layer.... Ensure that the uk-coastline layer is sitting above the landsat-cornwall layer. If it isn’t, you can drag it above.

The image we are using is a Landsat 7 image. As a joint venture between the USGS and NASA, the Landsat program represents the largest continuous archive of remote sensing imagery for land studies. With data from it’s 8 satellites, the Landsat imagery archive is a rich resource for researchers and interested people. It houses imagery of the world going back to the 70’s. Best of all, the imagery is free to use. Check out how Google have made use of this imagery for visualising how our Earth has changed since the 1980’s. Why not explore the Landsat Archive yourself and download the data.

You will see that the satellite image appears in Cornwall as it should. The file is saved in a format known as GeoTiff, which means it has location information associated with it.

Zoom in on the Raster dataset, using the zoom tools .

In this raster, each pixel represents a 30m x 30m square on the ground. So at high levels of zoom, you will see the image loosing detail.

Aerial Photography

Add the aerial photo called hurst-spit-aerialphoto.tif to your QGIS project. Zoom to Layer and you will see it will appear somewhere in the south of the UK.

If you zoom in, you will see the resolution of the aerial photo is much higher than the satellite image. The photo was captured from an airborne platform rather than high in orbit and such, pixel resolution is higher, at 25cm.

Load in winchester-aerialphoto.tif to your QGIS project, and winchester-minor-roads.shp. Make sure the vector layer is above the aerial photo layers in the Layers window. Zoom to Layer... and compare how the roads layer follows the actual roads displayed in the photo.

At high levels of zoom, you will notice that the road vector layers are quite generalised. They represent the shape of the roads when viewed on a large scale (zoomed out), but are not detailed enough to represent their shape when we zoom in closer.

You’ve learned what raster data is, and how to load it QGIS. You’ve also viewed vector data against a raster backdrop and gained an understanding of raster and vector resolution and detail.

Wouldn’t it be useful to have the ability to edit vector data?

Getting started

Download the following data layers

Rasters

• Winchester aerial photo

Vectors

• Winchester minor roads

It isn’t always better to have more detailed vector data. More detail equals greater data volumes. If it isn’t necessary to work with the data at small scales, then we can afford to generalise the data.

Add both these layers to the map, and zoom to the winchester-aerialphoto.

We may wish to edit vector data for a number of reasons. These can include adding features, changing the shape of line features, or adding attributes to the data.

In the previous tutorial, we showed that the roads layers are generalised, and the aerial photo actually shows more detail in the shape of the roads as it is a high resolution image. What we can do is edit the roads layer, such that it represents the features at a higher level of detail.

Editing line features

Zoom in to an area of the aerial photo, to roughly 1:2,500 scale. You can view the current scale of the canvas in the bottom right of the interface.

You may wish to change the styling of the vector layer to ensure it is easy to view over the aerial photo.

Select the winchester-minor-roads layer by clicking it in the Layers section. Right click on the layer and select Toggle Editing. This switches the layer into edit mode.

You will notice small red crosses appearing where ever a line changes direction. These are nodes.

Locate the edit tools on the QGIS interface.

From left to right, these are Save Layer Edits, Add Feature, Move Feature(s) and Node Tool.

Select the Node Tool. This tool allows you to select individual nodes, allowing you to move or delete them. Furthermore, if you double click somewhere on a line, this will add a new node which can then be moved to change the shape of the line.

Experiment with adding new nodes and moving or deleting existing ones. You can see how it is possible to add detail to the vector layer.

You may find it helpful to enable Snapping. Under the Settings menu, choose Snapping Options.... Here you can enable snapping for any layer you have loaded into the current project.

Once you are happy with your edits, save them using Save Layer Edits button and switch out of edit mode.

Creating new polygons

We will now create a new dataset from scratch by drawing around, or digitising the building outlines.

Locate and select the New Shapefile Layer tool.

Select Polygon and click ‘OK’. Then choose where you wish to save your layer, and give it an appropriate name. The data will appear in the Layers list automatically once it’s been created.

Toggle editing on this new layer, then select the Add Feature tool. Experiment tracing around the buildings in the aerial photo. When you have a completed polygon, right click to finish the feature.

Once you’ve completed a few, save edits and exit out of edit mode. Switch off winchester-aerialphoto layer. You have the beginnings of a map.

Getting started

Download the following dataset:

1. world-countries

What is a projection?

Projections exists to solve the problem that the earth is not flat, but yet we wish to display large swathes of it on flat media, like a map, or a computer screen. Therefore, projections aim to flatten the globe.

There are several ways to do this, and a lot of different projections to choose from. However, the main point to be aware of is that a projection will generally aim to retain a key aspect of true geometry, at the expense of all others. For example:

• Equal area projections will retain the measure of area to conform across the whole projection.
• Equal distant projections will retain the measure of distance.
• Conformal projections will retain the true shape of the feature on the earth.

We will use both the term projection and coordinate reference system (CRS). For the purposes of this tutorial, these can be taken to mean the same thing.

Whatever projection we use, it is still important to remember that distortions will exists. Distortions are given in a in a measure of scale factor. True scale with no distortions will have a scale factor of 1. Any deviation from this will indicate distortion. Scale factor distortions can vary across a single projection, such that the scale factor of a projection is not uniform. We visualise distortion using Tissot’s indicatrix.

So far in the tutorials, we have been using data which is all on the same projection, British National Grid.

However, we will now experiment changing the projection of our QGIS project. This will affect how QGIS renders the data within our project, regardless of the datasets native projection. As well as this, We’ll also show you how to change the native projection of a dataset.

Changing how our data is displayed

To illustrate how a choice of projection can change the look and geometry of the geographic data, we will use a worldwide vector dataset of countries.

Open a new QGIS project.

Load in the vector layer world-countries.shp

Notice the bottom right of the screen.

This section of the interface tells us what projection the QGIS project is using to display the data. This is indicated with an EPSG code. The current EPSG code is 4326 which tells us the projection is WGS84.

The International Association of Oil & Gas Producers (OGP) maintains the EPSG dataset. The primary dataset is maintained in an online registry. It has become something of a standard to refer to coordinate systems using EPSG codes, as it is an internationally known system. All projections supported by QGIS reference their EPSG code.

The projection associated with the first dataset loaded into a new QGIS project determines it’s projection. Open the Properties of the world-countries layer and view the coordinate reference system information, visible in the General tab.

‘On the fly’ projection makes it possible to load in datasets with multiple different coordinate reference systems, and all will be rendered using the coordinate reference system of the QGIS project. This should be used with caution! If a datasets coordinate reference system is declared wrongly, this will ultimately result in geolocation errors.

Exit out of Layer Properties.

Click the symbol at the bottom right of the screen. This brings up a dialogue allowing us to change the projection of the QGIS project.

You will first need to check the box to Enable 'on the fly' CRS transformation. This means that we can now change the coordinate reference system, or projection of the QGIS project, which will render the datasets in the project to this new projection ‘on the fly’.

Note! This is only changing how the data is displayed. Geographic datasets will have a native coordinate reference system, or projection associated with them, which can be different to this.

In the Filter box, type 900913. This will highlight Google Mercator. Select this projection and click OK.

Notice how the look and geometry of the dataset has changed.

The Mercator projection was originally developed for marine navigation, as if you plot a straight line on the map, it will have a constant bearing. However, it distorts area. Notice the size of Greenland compared to Africa, they look roughly the same size. In reality Greenland is smaller in area than it appears here.

Let’s change the display to an equal area projection for contrast. Change to World_Eckert_V, EPSG: 54011. Keep your eyes on Greenland and Africa.

Try changing the projection of the QGIS project to the following projections.

• World_Two_Point_Equidistant, EPSG: 54031
• World_Equidistant_Conic, EPSG: 54027

Changing the projection of a dataset

Explore some more map projections. What’s your favourite?

So far, we’ve been changing the projection of a QGIS project, such that the data is displayed in that projection. Now, we will change the projection of the actual data.

Right click on the world-countries layer in the Layer list.

Choose Save As.... This will open a dialogue in which we can specify an output CRS (coordinate reference system) for the dataset. Browse to EPSG: 54021. Choose a name for your new dataset, such as world-countries-54021.shp. Ensure you are saving the type as ESRI Shapefile. Click Save.

Open a new QGIS project.

Load in your new dataset world-countries-54021.shp. Notice how the projection of the QGIS project is now set itself to EPSG: 54021. Access the Layer Properties of your new dataset and view the coordinate reference system. Notice here also, that the dataset is in EPSG: 54021.

Got this far?

That was a very quick tour around projections and coordinate reference systems. We hope you can see how important they are to consider when working with geographic data.

You may enjoy reading some further details on projections.

That’s it for now. You’ve learnt some fundamental GIS concepts, and a great piece of software too.

What next?

Let @lo_ise know on twitter if you want more tutorials.

What about learning how to compile and print a map? Integrating your GPS unit with QGIS? Exploring datasets showing land surface elevations? There are plenty of possibilities!