../../_images/1_0001.png

Chapter 6: Working with Raster Data

Learning Objectives

  • Load Raster Data
  • Change Raster Symbology
  • Terrain Analysis

Thus far we’ve worked mostly with vector data, which consists of features, and these features are themselves made up of points and lines. In this chapter we will learn about raster data. Remember when you were editing OpenStreetMap in JOSM? The points, lines and shapes that you drew were vector data. But when you loaded Bing aerial imagery in the background, that was raster data. So what’s the difference?

Raster data essentially comes in the form of an image. It is made up of pixels, like a photograph, and a raster image will always be some number of pixels wide and some number of pixels high. If you zoom in far enough on a raster image, it will start to become blurry, just as if you opened a photo on your computer and zoomed in very close. As we’ll see in this chapter, however, a raster image can mean more than just a photograph from the sky. Follow along and we’ll learn all about rasters!

6.1 Loading Raster Data

Not all raster data consist of aerial photographs. There are many other forms of raster data, and in many of those cases, it’s essential to symbolise the data so that it becomes properly visible and useful. In this section we’ll add a new kind of raster and see how to change its symbology.

  1. Click on the Load Raster Layer button:
../../_images/6_0011.png
  1. The Load Raster Layer dialog will open. Find the file named popmap10_all.tif in QGSI for Disaster Management/Working with Raster folder. Open it.
  2. When it appears in the Layers Panel, right-click on it and click Rename. Give it the name Indonesian Population.

Note

High resolution, contemporary data on human population distributions are a prerequisite for the accurate measurement of the impacts of population growth, for monitoring changes and for planning Interventions. The AsiaPop project was initiated in July 2011 with an aim of producing detailed and Freely-available population distribution maps for the whole of Asia.*

When it loads, you’ll notice that the new raster image appears as whole Indonesia’s Islands in grey colour.

../../_images/6_0021.png

The layer appears grey (and doesn’t give us any information) because its symbology hasn’t been customised yet. In the colour aerial photograph we loaded via OpenLayer Plugin, everything is already defined. But if you load a raster image and it’s just a grey rectangle, then you know there’s no symbology for it yet. It still needs to be defined. That’s what we will do next.

6.2 Symbolise Raster Data

  1. Open the Layer Properties dialog for the Population layer, which is now named Indonesian Population.
  2. Switch to the Style tab. This shows the current symbology settings, and as we’ve seen, they don’t give us much information on the layer. Let’s make sure the layer has data in it.
  3. Change the Render type to Singleband pseudocolor. Set colour to YlOrBr and change mode to Equal Interval with 6 Classes. Click Classify to show 6 classes. As shown below:
../../_images/6_0031.png
  1. Click OK. The raster should look like this:
../../_images/6_0041.png

Good! This tell us that there is data in this layer. And by looking at it we can get an idea in Java Island, population number higher than other island in Indonesia.

Let’s stop for a minute and understand what is happening here. Remember that an image is made up of pixels, individual cells that contain a value, which is usually a colour value. For example, if you zoom in very closely on a photograph you can see those individual pixels, like this:

../../_images/6_0051.png

The value of each cell is saved in the file. Imagine the file being saved something like this, where each square is a pixel:

../../_images/6_0061.png

Of course the computer doesn’t understand words for colours. In fact the value of each cell would be a number, which the computer then associates with a certain colour. For our aerial image, this is already defined. Since it is a normal image, it knows to associate the numbers for each pixel in the file with the common colours that we see every day. But this new raster image is different, because the values of each pixel don’t represent colours, but rather altitude, and QGIS doesn’t know automatically how to display it. Hence it shows every pixel in the image as grey, even if the values in each pixel are different. When we change the symbology to Psuedocolor, we can see all the different pixel values shown with various colours.

It would be nice to represent our Indonesian Population layer as a greyscale spectrum, rather than a variety of bright colours. Next we will tell QGIS to symbolise the layer with colours in a spectrum, beginning at the lowest pixel value in the file and ending at the highest pixel value. In other words, if the pixel values looked like this:

../../_images/6_0071.png

QGIS would create a spectrum equating numbers to colours like this:

../../_images/6_0081.png

And render the image like this:

../../_images/6_0091.png

To do that, let’s start to symbolise Indonesia Population:

  1. Open the Layer Properties Again.
  2. Switch the render type back to Singleband gray (1).
  3. Check the box next to Min / max (2).
  4. Next to Contrast enhancement select Stretch to MinMax (3).
  5. Under Load min / max values, select Estimate (faster).
  6. Click the Load button:
../../_images/6_0101.png

Notice how the custom Min and Max values have changed. The lowest pixel value in this image file is 0 and the highest is about 3024.93. But what are the minimum and maximum values that should be used? The current values are those that just gave us a grey rectangle. Instead, we should be using the minimum and maximum pixel values that are actually in the image. You can determine those values easily by loading the minimum and maximum values of the raster.

  1. Click OK. You should see the values of the raster properly displayed, with the darker colour representing small number of population in each pixel and the lighter one, high number of population in each pixel:
../../_images/6_0111.png

We’ve learned to do this the tricky way, but can we do it faster? Of course! Now that you understand what needs to be done, you’ll be glad to know that there’s a tool for doing all of this more easily.

  1. Remove Indonesian Population from the Layers panel, by right-clicking it and clicking Remove
  2. Load the raster image again, renaming it to Indonesian Population as before. It will be a grey rectangle again.
  3. Enable the tool you’ll need by enabling View → Toolbars → Raster These icons will appear in the interface:
../../_images/6_0121.png

The button on the right will stretch the minimum and maximum values to give you the best contrast in the local area that you’re zoomed into. It’s useful for large datasets. The button on the left will stretch the minimum and maximum values to constant values across the whole image.

  1. Click the right button labelled (Stretch Histogram to Full Dataset). You’ll see the data is now correctly represented as before!

6.3 Terrain Analysis

Certain types of rasters allow you to gain more insight into the terrain that they represent. Digital Elevation Models (DEMs) are particularly useful in this regard. In this section we’ll do a little bit more with our DEM raster, in order to try to extract more information from it.

6.3.1 Calculating a hillshade

The DEM you have on your map right now does show you the elevation of the terrain, but it can sometimes seem a little abstract. It contains all the 3D elevation information about the terrain that you need, but it doesn’t really look 3-Dimensional. To get a better look at the terrain, it is possible to calculate a hillshade, which is a raster that maps the terrain using light and shadow to create a 3D-looking image.

To work with DEMs, we will use the all-in-one DEM (Terrain models) analysis tool.

  1. Open the file named SRTM_Merapi.tif, which is located in QGIS for Disaster Management/Working with Raster folder.
  2. When it appears in the Layer panel, rename it to DEM.
  3. Click Stretch Histogram to Full Dataset button.
  4. Go to Raster → Analysis → DEM (Terrain Models)…
../../_images/6_0131.png
  1. In the dialog that appears, ensure that the input file is the DEM layer.
  2. Set the output file to hillshade.tif in the directory Merapi/SRTM/.
../../_images/6_014.png
  1. Check the box next to Load into canvas when finished.
../../_images/6_015.png
  1. Leave all the other option unchanged.
  2. Click OK to generate the hillshade.
  3. When the processing is complete. Click OK on the notification.
  4. Click Close in the dialog.

There should now be a new layer called hillshade that looks like this:

../../_images/6_016.png

This looks more Three-Dimensional, but can we improve on this? When you open DEM (Terrain Models) tools, you will notice in the Mode Options for hillshade there’re Azimuth of the light and Altitude of the light.

Azimuth of the light defines in which direction the sun is, whereas 0° represent north, east is 90°, south is 180° and west is 270°. While Altitude of the light is the angle between the horizon and the centre of the sun’s disc.

On its own, the hillshade looks like a plaster cast. It will look better if we can combine it with our more colourful DEM. We can do this by making the hillshade layer an overlay.

6.3.2 Using a hillshade as an overlay

Hillshade can provide very useful information about the sunlight at a given time of day. But it can also be used for aesthetic purposes, to make the map look better. The key to this is setting the hillshade to being mostly transparent.

  1. Change the symbology of the original DEM layer to use the Pseudocolor scheme.
  2. Click and drag the DEM layer beneath the hillshade layer in the Layers panel.
../../_images/6_017.png
  1. Ensure that Control rendering order is checked.
../../_images/6_018.png
  1. Now we will make the hillshade layer somewhat transparent. Open its Layer Properties and go to the Transparency tab.
  2. Set the Global transparency to 50%:
../../_images/6_019.png
  1. Click OK in the Layer Properties dialog. You should get a result similar to this:
../../_images/6_020.png

Using a hillshade in this way, it’s possible to enhance the topography of the landscape. If the effect doesn’t seem strong enough to you, you can change the transparency of the hillshade layer; but of course, the brighter the hillshade becomes, the dimmer the colours behind it will be. You will need to find a balance that works for you.

Go to next chapter –>