Open the landsat.qgs project.

The image should now look like the following:

Colors representing a given pixel are defined using the RGB color space, which requires three different components. A normal image (such as the one you will get from a digital camera) has three bands containing the intensity for each one of these three components: red, green, and blue.

Multispectral bands, such as the one used in this recipe, have more than three bands and provide more detail in different regions of the electromagnetic spectrum. To visualize these, three bands from the total number of available bands have to be chosen and their intensities have to be used as intensities of the basic red, green, and blue components (although they might correspond to a different region of the spectrum, even outside the visible range). This is known as a false color image.

Depending on the combination of the bands that are used, the resulting image will convey a different type of information. The combination chosen is frequently used for vegetation studies, as it allows you to separate coniferous from hardwood vegetation as well as providing information about vegetation health.

The combination is applied, in this case, to a Landsat 7 image, which is taken with the ETM+ sensor. The wavelengths covered by each band are as follows (in micrometers):

Different combinations are frequently used for Landsat layers. One of them is the following:

This is a natural color combination, as the bands used for the R, G, and B components actually have the wavelengths corresponding to the colors red, green, and blue:

If you are using an image that is not a Landsat 7 one, each band will have a different meaning, and using the same combination of band numbers will yield different results. The meaning of each band must be checked in order to understand the information displayed by the rendered image.