753 
6 CONCLUSIONS 
>here was app. 
in averaged 
l for the NOAA- 
l if the atmos- 
! of the order 
lospheric cor- 
i data set are 
! there is a 
¡n the grey lev- 
>f the satellite 
temperatures 
1 K. Having in 
le compared 
lommonly accep- 
fle infrared 
n area derived 
irature value of 
ire is within 
.sius respec- 
■e also been 
s often obser- 
i group at the 
of Bergen has 
of the currents, 
can be of dan- 
. oil-fields 
current gradi- 
iddies off the 
The image is 
:rom NOAA- 6 
. 13. 1981. 
irn of eddies is 
the pattern 
from approxi- 
! Celcius. For 
lospheric cor- 
lave not been 
i TM thermal 
r HRR, offers the 
>T-algorithm to 
igainst internal 
lata is located 
at present 
Figure 4.3. A NOAA-6 derived SST image covering the 
Sothern Norway and the North Sea. The upper part of 
the image is covered by clouds (Cfr. area C fig. 4.1). 
they are not available to the users. Therefore, it is 
impossible to apply the SST-algorithm to the present 
TM data. The SST image presented here is generated 
from a look-up table based from comparisons of digital 
values and in-situ temperatures at known locations in 
the image. 
The image data presented in figure 5.1 is a Landsat- 
5/TM channel 6 sub-scene from the Troms0 area (Path 
197/Row 11) (Area B in fig. 4.1). Troms0 (69.6 N/18.9 
E) is located in the upper right corner. The data were 
acquired at Kiruna on June 3. 1984, and have been pro 
cessed at Troms0 Telemetry Station's image processing 
laboratory. 
By combining the thermal and a near-infrared channel 
the land areas have been removed. Land is represented 
by the color black in the image. The different grey 
levels of the sea surface have been assigned a tempe 
rature as indicated by the scale at bottom left. 
When the data were acquired, the current around the 
island of Troms0 was moving north (up the image), 
transporting relatively warm surface water (app. 10 
deg. Celcius) northwards. The island is linked by two 
bridges, one to the east, and one to the west (not 
seen in the image). As the water flows through the 
bridges, the supports of the bridges cause turbulent 
mixing of the warm surface water with the colder sub 
surface water. This mixing is clearly identified in 
the image. 
In the centre of the image there is an area where 
the surface temperature is approximately 2-3 degrees 
below that of the surrounding areas. This is caused 
by the upwelling of colder subsurface water due to 
the interaction of the small island and the local cur 
rent pattern in the narrow sound. Also there is a cold 
snow-melt water outlet of a local river. 
In the bottom right part of the image, a fjord which 
is an outlet for cold water can be seen. The cold wat 
er results from the river Malselv transporting cold 
snow-melt water from the local mountaineous ares to 
wards the sea. 
Although the Arctic weather conditions very often lim 
its the applications of optical satellite remote sen 
sed data in Norway, the results carried out by diffe 
rent Norwegian remote sensing institutes show the app 
licability of these services in research and to some 
extent also in (semi-)operational processing. With 
the development of the next generation of all-weather 
sensors, specially the operational processing of sat 
ellite data seems very promising. 
The medium resolution AVHRR-data have been very use 
ful in the research towards better understanding of 
oceanic processes. The eddies observed off the west 
ern coast of Norway (Cfr. figure 4.3) was fully dis 
covered when satellite data became .available. SST 
studies from airborne remote sensed data have also 
been very important in the understanding of the 
generation mechanisms for the eddies. In this case, 
the research has resulted in a system for forecasting 
the eddies from local SST studies. 
The present algorithm operated at Troms0 Telemetry 
Station apply rediosondes data profiles for atmos 
pheric corrections. Since there is a lack of spatial 
coverage from the radiosondes, the applications of 
these profiles assume a stationary atmosphere, which 
is a very idealized, never-occuring assumption. How 
ever , a combined application of AVHRR- and TOVS (Tir 
os Operational Vertical Sounder) data will offer the 
opportunities for taking the spatial atmospheric in- 
homogenitites into account when deriving atmospheric 
corrected sea surface temperatures. Therefore, the fu 
ture activity in Troms0 regarding applications of 
NOAA-data will include registered AVHRR- and TOVS- 
data processing. 
Compared to the 16 day repeat cycle of the Landsat, 
the high frequency of repetivity for the NOAA-satel- 
lites makes this system very useful for operational 
processing in Norway. 
Figure 5,1. Landsat-5/TM derived SST image from the 
Troms0 area. The observed features are discussed in 
the text. 
The high resolution thermal infrared TM data seem 
very suitable for studying surface temperatures and 
currents within the Norwegian coastal-zone. For a 
coastal-zone planner specially the applications in 
the fjords seem very interesting for the development 
of the natural resources. 
At present there is a limitation in applying exist 
ing SST-algorithms to the TM, due to the lack of cali 
bration data. However, this-seems to be a temporary 
problem since the ESA/Earthnet plans to include the 
in-flight calibration data on the supplied data.