Kalemie is a small city on the shore of Lake
Democratic Republic of Congo (DRC) in central Africa. It
is located approximately 300 kilometers south of
Rwanda. In the mid-1990’s, several large conflicts broke
out in the DRC, started mostly by the Rwandan genocide
in 1994. Due to its geographic location and resources
available, Kalemie became heavily involved in the
conflict.
This study aims to analyze how the forest and
urbanization has changed in Kalemie and the
surrounding areas throughout the conflicts between 1994
and 2003 using historical Landsat imagery.
Introduction
Datasets and software platforms used
Outputs Methodology
Based on literature reviewed, data trends found in change detection and class statistics reflect the
changing patterns of land cover one would expect to observe in a conflict area. There is a clear increase in
urbanization shown at the time of the Rwandan Genocide, during which thousands of refugees flooded into
the Democratic Republic of the Congo. This is to be expected, as the city of Kalemie is near the border with
Rwanda. This increase in urbanization – and population – is accompanied by a decrease in forest cover, likely
due to the increased agricultural needs of the growing population.[1]
As the conflict in the Democratic Republic of the Congo began in 1996, urbanization slowed, and began to
fall, as Kalemie became a source of refugees. The forest began some regrowth from the decreased
agricultural requirements, although its peak would come after the eventual end of the conflict. As the conflict
wound down, the urban cover of the city began to slowly increase, a trend that continued until the end of the
study period.[2]
While the city of Kalemie was largely free of clouds, surrounding areas were not. Masking techniques using
the FMASK band provided by USGS allowed for some improvement, however this resulted in some
misclassification and representation of non-urban land cover being artificially lowered in some years.
References[1] Baumann, M., Radeloff, V. C., Avedian, V., & Kuemmerle, T. (2015).
Land-use change in the Caucasus during and after the Nagorno-Karabakh
conflict. Regional Environmental Change, 15(8), 1703–1716.
https://doi.org/10.1007/s10113-014-0728-3
[2] Butsic, V., Baumann, M., Shortland, A., Walker, S., & Kuemmerle, T.
(2015). Conservation and conflict in the Democratic Republic of Congo: The
impacts of warfare, mining, and protected areas on deforestation. Biological
Conservation, 191, 266–273. https://doi.org/10.1016/j.biocon.2015.06.037
Satellite data used for this project was derived mostly
from the Landsat 5 sensor, but some gaps from 1998 to
2001 were filled using Landsat 7 prior to the scan line
corrector (SLC) device failing. Data was acquired from
Google Earth Engine, and the specific dataset used was
the USGS Landsat 5 TOA Reflectance
(Orthorectified) with Fmask dataset. The
corresponding dataset for Landsat 7 was used as well.
Two images from each year between 1990 and 2009
were selected for a 40 image timeseries analysis.
ArcGIS was used for masking of images, and ENVI
was used for the classification methods as well as the
change detection.
Input Landsat tiles
Mask cloud and cloud shadow
Classify scenes using MLC
Derive training classes using false colour
composite
Accuracy assessment using confusion matrix
Accuracy assessment report
Thematic change detection analysis
Final maps, graphs and results
Results and Discussion
Change in land classification from 1990 to 2009
1990 1994 1998 2002 2006
Year of image
Perc
enta
ge o
f p
ixel
s cl
assi
fied
as
urb
an
Perc
enta
ge o
f p
ixel
s cl
assi
fied
as
hea
vy v
eget
atio
n
0.00%
5.00%
10.00%
15.00%
20.00%
25.00%
30.00%
35.00%
40.00%
45.00%
50.00%
0.00%
2.00%
4.00%
6.00%
8.00%
10.00%
12.00%
1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010
Change in urbanization and heavy vegetation cover over time
Classified imagery from every four years over the study area.
Study area
Location of Kalemie in relation to Rwanda. Basemap courtesy of National Geographic.
