Lake Chad 1973 - Public domain US army map

Just 40 years ago, Lake Chad encompassed 25,000 square kilometers, and fishermen collected over 230,000 tons of fish from the lake every year. The lake has now shrunk to an area of typically around 500 square kilometers, with seasonal fluctuations, and the average yearly catch has dwindled to about 50,000 tons. People are now farming in areas that just decades ago were submerged.

Why is Lake Chad shrinking?

In order to explain the significant loss of water in the course of the last decades, a whole series of different theories have been proposed.

All of these theories attach a lot of importance to climate change. A rise in temperature results in an increase in the evaporation rate which is then accompanied by a decrease in the volume of tributary waters. Obviously, the decrease in vegetation cover in the savannah regions inside the watershed of the lake plays a decisive role due to its contribution to a change of the regional climate, which gets dryer. This decrease in vegetation is mainly attributed to overgrazing. Another factor is the increased use of the lake and feeder rivers for drinking water and field irrigation.

Chad’s population has tripled in the last four decades. From 1983 to 1994, there was a 400% increase in farming areas. Agriculture is therefore responsible for roughly 50 per cent of the aggradation of the lake.  
 
Measure the surface area of Lake Chad
 
In this exercise you will take measurements of the surface area of Lake Chad in order to quantify the extent of the retreating water. You will then perform a multitemporal change detection analysis with the use of GIS layers. Because the images are geocoded to the same map system, these GIS layers can be overlain on georeferenced images of different sensors, with differing spatial resolutions, and varying image extents. Geocoding is especially important if you want to detect changes over very long time spans, such as in the case of Lake Chad.

For the sake of simplicity, we define the lake in this exercise as the largest remaining water body to the southeast, without including the smaller isolated water bodies nearby.

For this exercise you will use a satellite image acquired by the MERIS sensor in 2011 (the same image used in Exercise 1) together with two other images acquired by Landsat satellites: one of the Landsat images was acquired by Landsat 1’s MultiSpectral Scanner (MSS) in 1973. This will enable you to perform change detection over a longer time span. The name of the image is “Landsat_MSS_LakeChad_January1973” and it can be found in the “Eduspace - Download” section in the right menu. The other Landsat image originates from the Thematic Mapper (TM)sensor of Landsat 5 in 1986 and is named “Landsat_TM_LakeChad_December1986”.

Now you must analyse, qualitatively as well as quantitatively, Lake Chad’s dramatic decline over the past 40 years.

1. Open the .tif file of the 1973 Landsat MSS image using LEOWorks. The bands of Landsat 1’s MSS are detailed in the following table:
 
 

Channel   Wavelength Range (μm) Landsat 1,2,3 Landsat 4,5   MSS 4 LEOWorks: band1 MSS 1 0.5 – 0.6 (green) MSS 5 LEOWorks: band2 MSS 2 0.6 – 0.7 (red) MSS 6 LEOWorks: band3 MSS 3 0.7 – 0.8 (near infrared) MSS 7 LEOWorks: band4 MSS 4 0.8 – 1.1 (near infrared)

 
 
Open an RGB view which is suitable for clearly displaying water areas. Try different combinations, including an RGB_3-2-1.

In order to roughly measure the former surface area and perimeter of Lake Chad, use the measurement tool and trace a line around its coastline. Make a note of your results.

2. Now create an RGB view of the 1986 Landsat 5 TM image.

Repeat the steps you followed for the MSS image above.
 
 

Channel Landsat 4,5 Wavelength Range (μm) TM 1 0.45 – 0.52 (blue-green) TM 2 0.52 – 0.60 (green) TM 3 0.63 – 0.69 (red) TM 4 0.76 – 0.90 (near infrared) TM 5 1.55 – 1.75 (medium infrared) TM 7 2.08 – 2.35 (medium infrared) TM 6 10.4 – 12.5 (thermal infrared)

 
3. Now open the Envisat MERIS image, which is the most recently acquired image of Lake Chad. Create an appropriate RGB view for water detection, just like you did in the previous exercise.

In order to calculate the perimeter and surface area of the lake, repeat, for this MERIS image, the steps you previously performed with the Landsat images. Don’t forget to write down your results.

4. Compare your results.

Calculate absolute as well as relative values, which should be expressed as percentages (differences divided by initial values X 100%) . Use the following tables to record your findings:
 
 

Perimeter Area January 1973     December 1986     Absolute difference     Relative difference

 
 

Perimeter Area December 1986     January 2011     Absolute difference     Relative difference

 
 

Perimeter Area January 1973     January 2011     Absolute difference     Relative difference

 
 
5. Using these three data sets, draw two graphs in a Cartesian coordinate system to visualise the development of Lake Chad’s perimeter as well as its area (ordinate = Y-axis) over time (abscissa = X-Axis). Visualising it this way will facilitate your interpretation of the decline.

6. Now use the GIS tool to trace the lake’s year-dependent coastlines in a GIS layer using polygons. Add as new attribute area and length attribute.

Save each of the layers in a .shp file with a distinct name such as “Lake_Chad_coastline_1973.shp”.

Finally, overlay these three GIS layers onto the up-to-date MERIS image and select different colours using the Layer Style Editor.

Write a small essay based on your analysis of the shrinking of Lake Chad based on your measurements, graphs, and images.

Compare your results with the other students in your class.