784 
3 METHODS. 
Table 1. Band width of the NOAA-7 AVHRR system 5 CONCLUS 
3.1 Outlines 
The contribution in waterflow of 90 smaller and 
larger Laga's is an unknown factor, which is 
difficult to analyse given the vastness and 
inacessibility of the lower catchment area (55,000 
km 2) and the ephemeral character of the outflow. 
To measure and monitor the flow in these Laga's in 
the field or even from the air in a systematic way 
is an operation which involves huge means in terms 
of logistics and manpower, and is therefore 
outside the scope of this study. 
A study methodology principally based on the use 
of satellite remote sensing techniques offers in 
fact the only way to obtain data of the area in a 
systematic way. 
Because of the dynamic nature of the Laga system a 
satellite monitoring program during the two rainy 
seasons in may-june and october-december in 1984 
seems to meet best the objectives. 
3.2 Selection of satellite imagery 
The satellite systems available for civilian use 
and capable of imaging the area whith regular 
intervals are the Landsat system and the systems 
based on meteorological satellites. The current 
Landsat satellite passes over the area whith an 
interval of 16 days. Only on few occasions Landsat 
data of the area are being collected, while there 
also is a great chance that the area is hidden for 
the satellite behind the clouds during the time of 
observations. Since Landsat data do not offer a 
reasonable possibility for the collection of a 
satisfactory amount of data during the cloudy 
seasons, the study relies on imagery provided by 
the meteorological satellites. 
Among the remote sensing instruments on board of 
the various meteorological satellites the Advanced 
Very High Radiometer (AVHRR) on board of the 
American satellite NOAA-7 offers the best possibi 
lities given its characteristics in spatial and 
temporal resolution (described in detail in 
section 3.3.). Use is made of of the Local Area 
Coverage programme (LAC) for the acquisition of 
the NOAA satellite data. In this programme user 
request NOAA-7 AVHRR data is recorded onboard the 
satellite for subsequent playback as Local Areal 
Coverage (LAC) data. 
3.3 The AVHRR system on board NOAA-7 
The American NOAA-series of polar orbiting envi 
ronmental satellites became a more interesting 
source for studying vegetation since the launch of 
the NOAA-6 satellite in june 1979. An improved 
sensor, the Advanced High Resolution Radiometer 
(AVHRR) was since then added to the basic confi 
guration of the NOAA-satellites. The main features 
of interest for studying vegetation with the AVHRR 
system are the high temporal, near daily revisit 
of any given place on earth, associated with an 
increased spatial resolution of 1.1 km at nadir, a 
+/- 56 csan angle and a swath width of 2700 km. 
The AVHRR system on board of NOAA-7, from which 
the imagery for this study was obtained, records 
in 5 channels. The band width of these channels 
are presented in table 1. 
3.4 Ordering of N0AA-AVHRR imagery 
DHV requested NOAA/NESDIS to collect NOAA-7 AVHRR 
imagery through the Local Area Coverage(LAC) 
program during two periods of two months (May-June) 
and three months (October-December) in 1984. 
Quick-looks of the best two images of each week 
were produced by NESDIS. From these hard copies 
Channel Bandwidth (urn) 
1 
.58 - 
.68 
2 
.725 - 
1.10 
3 
3.55 - 
3.93 
4 
10.5 
11.3 
5 
11.5 
12.5 
DHV selected the images suitable for digital 
processing. Unfortunately clouds were always 
present over the study area during the first 
period of observation. Only two images of May 3 
and June 20, with a relatively low cloud cover 
were worth digital processing. The digital tapes 
of these two dates were ordered for further inves 
tigation. During the second period of observation, 
data of only four heavily clouded days were 
archieved despite NOAA's adequate mission planning. 
So no suitable images for digital processing were 
made available during this second period of obser 
vation. 
3.5 Image processing 
The image processing facility at the Research 
Institute for Nature Management (RIN) was used for 
the processing of the NOAA tapes. The digital 
image processing routines that were used consist 
merely of image enhancements in order to obtain an 
optimal visual representation. Also the normalized 
difference images : (c2-cl)/(c2+cl) of the near- 
infrared and red channel resp. channel 2 and 1, 
(according to C.J.Tucker,1984) were calculated for 
both days. According to this author this so-called 
biomass index was presented as a succesful variable 
which correlate well with the green leaf biomass 
of vegetation. 
Consequently no atmospheric corrections were deemed 
necessary. 
4 RESULTS. 
Due to the presence of severe cloud cover 
especially above the catchment area only parts of 
this area could be observed satisfactorily on the 
imagery of both days. Therefore no systematic 
analysis on the hydrologic influence of the Laga's 
on the Tana River could be made. 
However, despite this severe obstruction the image 
ry shows that a considerable vegetation growth took 
place along the upper reaches, especially in the 
swamp areas, of the larger Laga's. No significant 
vegetation growth along the lower reaches of the 
larger Laga's and the other Laga's could be 
observed. Given the morphology of the Laga's as 
visible on Landsat-imagery, this indicates that no 
significant outflow from the Laga's took place 
during the period of observation. 
The biomass image of May 3 shows consistently 
higher values of the biomass index in comparison 
with the biomass image of June 20. These observa 
tions are in accordance with the field observa 
tions and are caused by the fact that the only 
rains of this first rainy season of May-June 
fell in the period prior to May 3. 
Although the N0AA-AVHRR system is especially 
suited for inventory of large areas, it can be 
concluded from the imagery that it offers suffi 
cient detail in spatial and radiometric resolution 
to observe and monitor vegetation growth of small 
regions, like the area along the Laga's. 
During a 
sible to 
the N0AA- 
revisit c 
cover, wh 
River reg 
optical s 
practices 
like rada 
as the ti 
world. 
Despite t 
indicate 
monitorir 
regions. 
sufficier 
no atmosj 
imagery t 
the fielc 
REFERENCE 
Blyth, K. 
Report 
Walling 
Colwell,E 
vol. 1 
grammet 
Deutsch, 
Satelll 
Associa 
DHV Const 
microli 
photogi 
DHV Const 
measure 
aerial 
DHV Const 
Tana Ri 
morpho] 
DHV Const 
riments 
ephemei 
DHV Const 
Study 1 
Nether] 
Donker, I 
process 
ITC, tl 
Lillesanc 
and im< 
NOAA p< 
Depart« 
mental 
Tucker, ( 
Engine« 
tation 
imager]