New Delhi, India
|  | Delhi region and the Himalayas | |
Delhi area city structure
This is an image acquired by the MERIS instrument onboard the Envisat satellite on 3 February 2004. Out of the 15 bands covering visible and near-infrared wavelengths, we have chosen the infrared band 14 (885 nm). In infrared bands, vegetation appears bright or light grey, and water, urban areas, and other infrastructure appear in dark grey to black.
This is why the star-shaped structure of the settlements and their interconnections are so clearly visible. Between the capital Delhi (centre lower right) and thousands of villages, depicted as grey spots and equally spread over the country, there are a number of sub-centres of unequal importance.
Try to set up different levels of such centres by their size and the distance between them. How many types of sub-centres exist in order to administrate and manage the country? To find the answer you will need to zoom into the image to recognise the villages, the smallest entities of the Indian community.
What is the mean distance between two villages within a certain region, and how much is it between two sub-centres of the same level? Pixel size of the image is 300 m. Use the Measurement Tool in LEOWorks. Now compare your result with the equivalent distances in your country!
Consult an atlas, and with respect to Delhi, find other important cities, such as Chandigarh and Dehradun.
Find out all about the linear features connecting the settlements. What do they represent? Having identified and classified them in one region (perhaps with the help of an atlas), set up the principal rules for identification (e.g. what characterizes a canal, etc.) and apply these rules to a neighbouring region.
If you need more information about urban development to answer these questions, you can see the following website: http://en.wikipedia.org/wiki/Central_place_theory.
| | | MERIS image of the Himalayas |
Snow in the Himalayas
These Meris images provide a view of a very small part of the Himalayas, just north of Delhi, with the city of Chandigarh (central left) and Dehranun (central right, north of the first hill chain).
There is a clear difference between the two images. On 25 November 2003, there was some snow only on the highest region, but on 3 February 2004, snow was also abundant on the lower parts. The snow of the Himalayas is the water reservoir of Northern India.
Any knowledge about the quantity of water held here can help to manage these resources better. It is therefore important to know the area of the snow-covered surface.
How can we measure it?
Simply by counting the very bright pixels over the mountainous area and multiplying the number by the surface of one pixel, which is 300 m × 300 m.
| | MERIS image of the Himalayas | |
To do this, let's look at the histogram.
First, display the image acquired on 3 February 2004.
Then open the histogram by clicking on the image and, in the main menu, click 'View' and 'Histogram'.
The histogram of the data values of the image sums up all the pixels with respect to their values.
In the 'Blue' channel, we find 39,017 pixels with a value of 255.
But we have to be careful - clouds also have such high values. In order to exclude them, we must make a subset of the image.
Copy the mountainous part in a separate image, assuming there are no clouds there.
Use the Crop Button (button with a square symbol) and 'Edit'.
In the new image, open the histogram of the blue channel and find some 20,000 pixels with value 255.
Let's assume that it has snowed an average of 20 cm. This corresponds roughly to 2 cm of rain. That corresponds to 20 litres per square metre. What is the total quantity of water stored in the snow in this area at this time, knowing that each pixel is 300 m × 300 m?
Important notes:
The numbers used in this exercise are fictitious. The objective of this exercise is to demonstrate the principles of satellite monitoring.
In order to have 'real' information on the water volume in the Himalayas, we would also need map information and some limited 'ground truth' (snow height, temperature, etc.). We would also need to compute the water run-off for each drainage basin. In this way we could accurately predict the quantity of water available downstream.
Depending on the season, this information could also be used for flood warnings. In fact, the snow cover measured by satellite data is the crucial input to water run-off models.
| 
1
