Evaluation of methods for erosion surveys — a case study from the Northern Province of Rwanda
Rwanda, also called “the land of a thousand hills” is densely populated and 90% of the population lives on subsistence farming. A combination of a hilly landscape, extensive land use, overgrazing and intensive rainy seasons makes the erosion problem very complex and extensive in Rwanda (Clay, 1990).
The aim of this project was to evaluate the use of two methods for erosion studies in Rwanda: the Caesium method (Walling and Quine, 1992, Kulander and Strömquist, 1989, Bengtsson Sjörs, 2001) and supervised classification (Lillesand, Kiefer, Chipman, 2004) of satellite images made from Landsat 5 and 7 data.
The Caesium content in the soil samples was analysed twice, once with a Germanium detector and once with an Iodine detector. There were big differences between the two results. The discussion and conclusion are based on the result from the Germanium detector.
The expected erosion pattern with heavy erosion on the steep hillsides and an accumulation in the bottom could not be seen. Instead, no significant Caesium content was found either in the profile or on the reference sites. Possible explanations could be that the soil has been moved and mixed to a greater extend than expected or that all soil has eroded from the hill and the reference site.
Supervised classification based on Landsat 6 and 7 images was also an unsuitable evaluation method. As a result of cultivation, there are great variations in land use even on a very small scale, which complicates the work to map erosion on a landscape scale. Due to local conditions in Rwanda, a simple inclination map is currently the best alternative to overview the erosion problem.
Evaluating the differences between the two methods to analyse Caesium content in soil would be an interesting topic for further studies. More excessive erosion studies using The Universal Soil Loss Equation or the Soil Loss Estimations in Southern Africa (SLEMSA) might be necessary to further understanderosion in Rwanda.
Norén, K. & Spörndly, N. (2009). Evaluation of methods for erosion surveys — a case study from the Northern Province of Rwanda. Bachelor’s Thesis. Committee of Tropical Ecology. Uppsala University.