2004 
1993, 
cale: 
lates. 
34. 
'ctral 
tance 
mote 
uber. 
area 
ng of 
n of 
rtable 
ts of 
s on 
Zions. 
Tlggs. 
ndsat 
erate 
DETECTING AND QUANTIFYING LAND COVER AND LAND USE CHANGE 
IN EASTERN MAU BY REMOTE SENSING 
Kundu P.M', China S.S, Chemelil M.C and Onyando J.O 
Department of Agricultural Engineering, Egerton University, Box 536 Njoro, Kenya 
Abstract 
Eastern Mau is an important water catchment area for lake Nakuru drainage basin. It has great potential for agricultural 
development as manifested by the high yields of wheat, barley and dairy products from the few remaining large scale 
farms. This area was once covered by rich vegetation of hi ghland evergreen forests which extended from the Mau hills 
and turned into woodland dominated by acacia trees in the Rongai-Njoro plains. Remote sensing techniques and ground 
survey methods were used to evaluate the land cover /land use change that occurred in the area over a period of about 
40 years since 1964. A significant change of land use from forest to agriculture was evident from the interpreted results 
of the photos taken in 1964 and 1969, the SPOT image of 1987, the Landsat image of 1989 and the ground survey of 
2003. Deforestation, land fragmentation, cultivation of wetlands and rapid increase in human settlements have had 
negative impacts on water resources resulting in reduced stream flows and ground water. Land use has changed from 
planted forests and large scale farms of wheat, barley, and dairy to homesteads, kitchen gardens and small farms of 
maize, beans, potatoes and vegetables with some people keeping one or two dairy animals. Currently, studies show 
increasing land degradation in the area. Sheet and rill erosion is observed when it rains and gullies are beginning to 
develop along some of the farm tracks. To contain the situation. satellite remote sensing was used to monitor and 
provide data that can be used to evaluate and manage land use and land use change for sustainable development in Mau 
East. 
Keywords: Remote Sensing, Air photo-interpretation, Landsat, SPOT, Catchment 
Introduction Methodology 
Eastern Mau is in the Rift Valley Province of Kenya A systematic sequence of API was done on air 
and lies between Rongai-Njoro plains and the upper photographs of 1969 to ensure that definite, reliable 
slopes of the Mau hills. It is mainly covered by the and significant information was extracted. The central 
Quaternary and the Tertiary volcanic deposits. The photograph was analysed using mirror stereoscopes. A 
area is drained mainly by rivers Njoro and Lamriak full scene SPOT image (fig.5) covering Lake Nakuru 
which originate from Eastern Mau escarpment and and Lake Naivasha drainage basins was manually 
empty into Lake Nakuru and rivers kaumara, Nessuit interpreted and analysed. The general character as well 
and Rongai which drain northwards to lake Baringo as details of the geology, drainage, plant cover and 
and Bogoria. land use were drawn from surrogate indicators and 
interpretation elements in the images. Vegetation 
Remotely sensed data and ground survey methods cover was determined by the colour tone, the 
were used to evaluate the land cover /land use change distribution, form and texture of plant-covered areas. 
in the area for a period of about 40 years. A Land cover change was determined by comparing 
combination of fine, medium and coarse resolution multi-temporal images. A subscene Landsat image of 
images from different platforms were used to detect 1989 (fig.8) covering the recharge areas and the 
changes. The study showed a significant change of mature sections of the rivers originating from Mau 
land use from forest and woodlands to agriculture, and hills were interpreted. The large area covered in one 
rural built-up lands. This was evident from the image was important for various kinds of regional 
analysed results for the air photos of 1969, the SPOT studies with details of size 30-80 m being observed. 
image of 1987, the Landsat images of 1984 and 1989 TM bands 4, 5, 3 were selected as they showed 
and the ground survey of 2003 which gave a vegetation and land use features more clearly. 
chronological landscape record. Deforestation, land 
fragmentation, cultivation of wetlands and rapid Ground survey was carried out in the selected sample 
increase in human settlements have had negative sites in 2003. In order to achieve unbiased samples, 
impacts on water sources resulting in reduced stream surveys were based on an unaligned systematic 
flows. Land use has changed from planted forests, random sample of fixed-size segments of | km x 1 km. 
large farms of wheat, barley, and dairy to homesteads, Individual land parcels and ground cover classes were 
kitchen gardens and small farms of maize, beans, identified in each sample segment. By the method of 
potatoes and vegetables with one or two dairy animals. direct expansion, the area for each cover class was 
determined for the entire study area. A high 
densification of sample sites was preferred in order to 
validate the consistency of land use cover database. 
405