Symposium on Remote Sensing for Resources Development and Environmental Management/Enschede/August 1986 
© 1987Balkema, Rotterdam. ISBN 90 6191 674 7 
Approaches to monitoring renewable resources using remote sensing 
and geographical information system 
Lennart Olsson 
University of Lund, Sweden 
ABSTRACT: The report is aimed at introducing various remote sensing methodologies for the study of renewable 
resources. Different aspects of vegetation monitoring, including topics like supply/demand of natural 
resources and mapping of erosion hazards. The emphasis is put on applications of available satellite sensing 
systems on vegetation monitoring in developing countries. 
1. INTRODUCTION 
"...we are nowhere near to achieving the ultimate 
capability of the space vantage point for delivering 
useful information to the user. The information which 
such systems can produce is of immense economic, 
sociological and humanitarian value" (Landgrebe 
1983). This is very much true for the present sub 
ject, remote sensing of renewable resources. 
The term renewable resources is difficult to define 
explicitly. In this paper I have restricted the defi 
nition to cover different aspects of vegetation, but 
included some related fields, like land degradation 
(e.g. soil erosion and desertification). The emphasis 
will be put on applications of the most commonly used 
satellite systems, Landsat, NOAA and SPOT, on res 
ource monitoring in the tropics. A very important 
development of remote sensing is the geographical 
information system (GIS) and related technigues, 
which will be treated as well. 
2. REMOTE SENSING OP VEGETATION, AN OVERVIEW 
Remote sensing uses radiation in different parts of 
the electromagnetic (EM) spectrum as carrier of 
information regarding various phenomena on the earth 
and in the atmosphere. The EM spectrum ranges from X- 
rays to radiowaves, but only fractions of the 
spectrum have so far been used for remote sensing 
applications. Visible and reflected IR radiation is 
the most freguently used part of the EM spectrum, see 
Figure 1. However, research aiming at widening the 
range of wavelengths used for remote sensing is 
important. 
2.1 Visible and reflected infrared radiation 
Remote sensing by means of photographic technigues in 
the visible part of the electromagnetic spectrum (0.4 
-0.7 microns) have been used for a long time, though 
primarily for other applications than vegetation 
studies. The development of IR-sensitive film was 
initiated in 1931 (Reeves 1975). Although it was 
originally developed for military purposes, the main 
use has been within the field of vegetation studies. 
Among the earliest civilian remote sensing applica 
tions, were studies of vegetation type and stress 
(Colwell 1956). The introduction of electro-optical 
imaging sensors opened up a wider part of the EM 
spectrum. The research and applications have, 
however, mainly been restricted to the visible and 
the near infrared part of spectrum, see Table 1. 
In the early days of multispectral analysis of 
vegetation, much research was aimed at determining 
spectral signatures of different vegetation species. 
It soon became obvious that the concept of unique 
spectral signatures did not hold. Hence, it is very 
unlikely that automatic classification on board the 
satellite suggested by Ewalt (1979), will be reali 
zed. But the assumption of spectrally unique signa 
tures is still considered acceptable by many authors, 
when applying multispectral classifications, see 
section 5.1. 
An important feature of the visible and near 
infrared radiation is the relationship between 
reflectance and green vegetation, see Table 2. The 
combination of strong negative correlation with 
vegetation amount in visible red and the strong 
positive correlation in near infrared can be used for 
quantifications of vegetation amounts. 
The introduction of new wavelength bands with the 
Thematic Mapper sensor of Landsat 4 & 5, made some 
interesting parts of the spectrum, in the near infra 
red part, available to satellite remote sensing, the 
bands 5 and 7 (Figure 1 and Table 1)\. The reflectance 
from vegetation in this part is mainly controlled by 
the water content./ In combination With the shorter 
NIR and visible bands, we cah expect new 
possibilities for assessing vegetation'' conditions. 
2.2 Thermal infrared radiation 
All objects warmer than -273 C emit electromagnetic 
radiation, where the wavelength is dependent of the 
temperature of the object. The radiation emitted from 
the surface of the earth is typically in the range of 
3-50 microns. Remote sensing devices recording 
this radiation can be used to measure the radiating 
temperature of objects. However, to determine the 
actual temperature of an object, we must know the 
emissivity of the surface we are sensing, and this is 
a major constraint to the use of thermal sensors in 
remote sensing. Natural surfaces have typical 
emissivity values between 0.90 and 0.99. To determine 
actual temperature to an accuracy of 1C, an accuracy 
of 0.02 of the emissivity is required (Slater 1980). 
Most important remote sensing platforms for thermal 
IR sensors have been, and still is, meteorological 
satellites e.g. NOAA, SMS/GOES, and Meteosat. Other 
platforms are HCMM (Heat Capacity Mapping Mission 
1978-80), Seasat (1978) and the Landsat 4 & 5 band 6 
of the Thematic Mapper sensing system. For a review 
of thermal infrared remote sensing from satellite see 
Lynn (1986). 
The literature on the use of thermal IR imagery for 
vegetation studies is, compared to visible and NIR, 
scanty. There are several reasons to this: 
1041