. For. Res.
Symposium on Remote Sensing for Resources Development and Environmental Management / Enschede / August 1986
t Interpreta-
g von Fichte
JLlgem. Forst
Dwarfmistle-
tains of New
Aspects of
g of actual
ng und Luft-
ihre Erfas-
n, E. and G.
irben wissen.
pp. 97-111.
tigations to
onditions of
r irginia Agr.
he spruce in
)gy Bull. 28,
)ocumentation
1. J. Forest
Acid deposi-
nd Tech. 17:
ruce budworms
USDA Forest
stry Service
tatus of the
and Tennes-
v. State and
-1-1, 12 pp.
tates trees,
. USDA Misc.
sel zur Aus-
mr die Wald-
Lts 39:1093-
en, and W.M.
if red spruce
, Forest Pest
i, Fort Col-
worth. 1985.
• forests of
—Mt. Rogers
>untains and
l Appalachian
:t, 67 pp.
[ahrer. 1981.
iterpretation
L.) auf mit-
len. Memoirs-
Forsliche
.and 57:433-
lmann. 1982.
Mountains of
.09:162-168.
1966. Aerial
lg damage by
est Service,
Exp. Sta.,
, M. Miller-
ative survey
ind mortality
lire - 1984.
a, NA-TP-11,
italitat des
s Bundesins-
80 pp. plus
Field experience with different types of remote-sensing data
in a small-scale soil and land resource survey in southern Tanzania
T.Christiansen
Agrar- und Hydrotechnik GmbH, Essen, FR Germany
ABSTRACT: During a small-scale agro-ecological land resource and soil survey for the Iringa Region in south
ern Tanzania the advantages and imperfections of black & white panchromatic aerial photographs and two types
of Landsat MSS imagery (standard false-colour composites, colour-enhanced imagery) were examined. Both types
of satellite imagery turned out to be of little use in areas with high rainfall (900 - 1600 mm) and a dense
vegetation cover, but they gave very good information in drier parts of the survey area (550 - 900 mm). The
aerial photographs contributed little information on the flat parts of the drier areas. They were most useful
in an intermediate zone with complex relief. Colour interpretation keys for the satellite imagery of selected
areas were compiled. The results indicate that the validity of such keys is generally limited to relatively
small, agro-climatically homogeneous areas.
RESUME: Au cours d'une étude à petite échelle portant sur les sols et les ressources agro-écologiques de la
Région d'Iringa dans le Sud de la Tanzanie, ont été comparés les avantages et inconvénients respectifs de
photographies aériennes panchromatiques noir et blanc et de deux types d'images-satellite Landsat MSS (compo
sitions colorées, l'une avec fausses couleurs standard et l'autre avec accentuation des couleurs). Les deux
types d'images-satellite se sont révélés être peu utiles dans les zones à pluviométrie élevée (900 - 1600 mm)
et à dense couverture végétale, mais par contre ont apporté beaucoup d'informations dans les zones plus
sèches (550 - 900 mm). Les photographies aériennes n'ont livré que peu d'informations dans les zones sèches à
relief très aplani. Par contre, elles ont été d'un grand intérêt dans une zone modérément humide à relief
complexe. Des clés d'interprétation des couleurs des deux types d'images-satellite ont été élaborées pour
deux zones sélectionnées. Il en ressort que la validité de telles clés d'interprétation est généralement res
treinte à des zones peu étendues et agro-écologiquement homogènes.
1. INTRODUCTION
For a long time soil surveyors have successfully
been utilizing remote sensing data for small-scale
soil and land resource surveys. A considerable num
ber of publications already exist about the appli
cation of aerial photographs and satellite imagery
for this type of mapping (Allan 1978, Carrol 1984,
Goosen 1967, Hilwig 1982, Mathews et al. 1973,
Nieuwenhuis 1978, Siegal and Goetz 1977, Thompson
et al. 1984, van Sleen 1983, Westin and Frazee
1976).
Relatively little, however, is published about
the question of how different types of remote sens
ing data stand the test in daily routine work and
what their respective advantages and shortcomings
are in different climatic and physiographic areas.
During a small-scale survey of a large area in
southern Tanzania some interesting field experience
was gained which might help to reduce this informa
tion gap.
2. THE SURVEY AREA
The survey area is located in the southwest of
Tanzania. It comprises the Iringa Region except
from the Ruaha National Park in the northwest and
the Eastern Forest Reserve in the northeast of the
region. The area totals about 46,000 sq km (by com
parison: Switzerland 41,288 sq km).
The area includes strongly different altitude le
vels, landscapes and climates. It ranges from hot
semi-arid savanna areas with Acacia-thornbush vege
tation at 500 metres altitude up to warm -températe
tropical highlands with tea plantations and alti
tudes of almost 3,000 metres. For African standards
the area is comparatively densely populated.
The soils of the survey area vary from saline
dark and red loams and dark cracking clays at the
lower elevations up to very leached red and yellow
clays at the higher altitudes. In some areas soils
have developed from volcanic ash deposits.
3. THE SURVEY: BACKGROUND, TASKS AND WORKING CONDI
TIONS
A land resource map (scale 1:250,000) and a soil
map (scale 1:100,000) had to be produced within the
frame of a regional agricultural development plan
for the Iringa Region. The survey had to be carried
out under typical consulting conditions, i.e. very
tight schedule, almost no preparation time and li
mited base information.
The most important survey details are summarized
in the following table.
Table 1. Survey Details
Land
Resource
Survey
Soil
Survey
Total time available
(man-months)
6
19
Fieldwork time
(man-months)
3
12
Number of surveyors
2
3
Observation density
no specific.
1/1000 ha
Survey output p>er
mapping day
900 sqkm
300 sqkm
Time for fieldwork
preparation
1 day /
1 day /
4200 sq km
600 sq km
The survey followed a two-step procedure: in the
first step a map was compiled of agro-ecological
zones and land units. This map was mainly based on
the evaluation of climatic data, interpretation of
satellite imagery, and a reconnaissance field sur
vey. Beside the climatic information this map also
includes some major soil boundaries. In a second
397
