Kathmandu Valley with optical images - General overview - Continued
|  | Natural colour combination of Kathmandu using bands 3,2,1 | |
Natural-Colour Combination
The natural-colour combination requires the Red (band 3 - 0.63 - 0.69 µm), the Green (band 2 - 0.52 - 0.60 µm) and the Blue (band 1 - 0.45 - 0.52 µm) Landsat channels.
This combination of Landsat bands leads to an image closely resembling natural colours, because it uses the natural colour reflection of the Earth's surface. The wavelength between 0.63 and 0.69 µm (band 3) represents the red part of the electromagnetic irradiation of visible light, and is therefore used for the red channel in the RGB colour system.
Landsat band 2 is used for the green channel in the RGB system, because green light is visible between 0.49 and 0.575 µm. This part of the spectral colours is represented in Landsat band 2. The same physical characteristic is used for the blue channel, which is represented in Landsat band 1.
Why does this combination lead to an image closely resembling a colour photograph?
What happens if you change the order? For example, make Red = band 1, Blue = band 2, and Green = band 3.
Open the LEOWorks programme. If you have not downloaded the images of Kathmandu yet, do so now.
Choose File>Open. A dialogue box will pop up. Choose the folder
Kathmandu and select the first image
Kathmandu_Landsat_2001_Band_1.tif. Open
Kathmandu_Landsat_2001_Band_2.tif and
Kathmandu_Landsat_2001_Band_3.tif as well.
Choose Image>Combine from...>Red Green Blue. A pop-up menu will open. Select image Kathmandu_Landsat_2001_Band_3.tif for Red,
Kathmandu_Landsat_2001_Band_2.tif for Green, and
Kathmandu_Landsat_2001_Band_1.tif for Blue, and click OK.
The new image is the natural-colour image produced using the 3 visible light channels. But it is not in natural colours yet, because the data still has to be enhanced. Select (activate) the first image Kathmandu_Landsat_2001_Band_3.tif and choose Enhance>Interactive Stretching. A histogram will appear. Shift the left (blue) bar in the Input Histogram to the left start point of the Input Histogram. Then shift the right (red) bar to the right start point of the Input Histogram and click Apply. Note the changes in the combined image.
Convert the other two images the same way.
How far you have to shift the bars for each band depends on your perception of colour and the colour adjustment of your screen. To produce an almost 'normal' natural-colour image, it is recommended that you open the already combined image Kathmandu_Landsat_2001_Band_321.tif and use it as a control element.
Describe the image in general.
Which features of Kathmandu can you identify in the Landsat natural colour image?
Point out the airport.
Describe the location of Kathmandu within the city surroundings.
Think of mountains, rivers, agriculture, forests, city parks and large areas of land.
Are you able to identify the city limits in this natural-colour image?
Why is the city centre kind of fuzzy?
| | | False-colour image of Kathmandu using bands 4,3,2 |
False-Colour Combination
To increase the interpretability of satellite images, false-colour images are often used.
In most cases a false-colour image uses at least one infrared channel. The infrared range is very useful to interpret the Earth's surface, because it consists of reflected energey, and, in the thermal range, emitted energy as well.
Infrared light is not visible to the human eye but reveals a lot of information. In particular plants reflect much more energy in the near infrared than in the visible range of the electromagnetic spectrum. Even the health of a plant can be assessed from the intensity.
There are different band combinations possible to produce a false-colour image. Two of the most common combinations are Landsat channels 4,2,1 and 4,3,2. The use of the near infrared (band 4) and the red channel (band 3) is important. The near infrared is very sensitive to vegetation, whereas the red channel provides the best representation of surfaces that are not covered by vegetation. We will use the band combination 4,3,2, due to the atmospheric disturbance over Kathmandu.
The atmospheric diffusion is much higher in shorter wavelengths than in longer ones. The highest diffusion of visible light appears in the blue spectrum, due to its short wavelength of 0.42 - 0.49 µm.
Open the LEOWorks programme. If you have not downloaded the images of Kathmandu yet, do so now.
Choose File>Open. A dialogue box will pop up. Choose the folder Kathmandu and select the first image
Kathmandu_Landsat_2001_Band_2.tif. Open
Kathmandu_Landsat_2001_Band_3.tif and
Kathmandu_Landsat_2001_Band_4.tif, too.
Choose Image>Combine from...>Red Green Blue. A pop-up menu will open. Select image Kathmandu_Landsat_2001_Band_4.tif for Red,
Kathmandu_Landsat_2001_Band_3.tif for Green, and
Kathmandu_Landsat_2001_Band_2.tif for Blue, and click OK.
Improve the raw data as described in the Natural-Colour Combination exercise.
This new image is a false-colour combination of three greyscale pictures.
How have the colours changed in general? Pay close attention to the vegetation and the built-up areas.
Why does the vegetation appear in red?
What appears in blue, and what appears in green? Explain the colouring by the degree of reflection of the different surface features.
The false-colour image of Kathmandu using the band combination 4,3,2 is still pretty fuzzy in the city centre area, due to strong diffusion in the short wave green channel. But there is a way to reduce the atmospheric influences by combining a false-colour image using Landsat bands 4,5,3.
| | False-colour image of Kathmandu with a reduction of atmospheric noise using bands 4,5,3 | |
There are two reasons why this combination leads to an image with reduced atmospheric influences. The first is that the channels used have longer wavelengths, and therefore the diffusion is smaller. The second is that there is the greatest spectral difference between red light, near infrared and middle infrared. The image features are much more distinct than in a false-colour image using two channels out of the visible light and one near infrared.
Choose Image>Combine from...>Red Green Blue. A pop-up menu will open. Select image Kathmandu_Landsat_2001_Band_4.tif for Red,
Kathmandu_Landsat_2001_Band_5.tif for Green and
Kathmandu_Landsat_2001_Band_3.tif for Blue and click OK.
Improve the raw data as described in the Natural-Colour Combination exercise.
We will need this image in the following exercises. Therefore save the image as Kathmandu_Landsat_Band_453 (TIF) to your Kathmandu folder.
It is definitely nice to have a noise-free and pretty colourful image. But it is worthless if it is not used, for example, to interpret and classify data. The purpose of a satellite image is to provide information. It is up to the spectator, to you, to filter this information out. For details on how to do this, see the 'Multispectral Image Classification' exercise.
| 
