3 
In order to develop a consistent picture of trends in the earth system over time, the 
measurements made by satellite borne instruments must be reduced to calibrated physical 
variables which have meaning in terms of the processes and cycles being monitored. This 
process is shown schematically in Figure 1. In the initial processing step, the raw sensor 
measurements are calibrated and precisely located 
on the earth's surface. These processed 
measurements (say radiance for example) are 
stored in a geocoded data base along with data from 
other sources. Physical variables (reflectance, for 
example), which are themselves also stored in the 
database, are then derived from the processed 
measurements. The physical variables provide 
inputs to modelling processes which lead to an 
understanding of the ecosystem being modelled. In 
designing systems to make the measurements and 
perform the required processing tasks, we are led to 
ask three important questions: 
1. What do we want to measure? 
2. Why do we want to measure it? 
3. How are we going to measure it? 
The answers to these questions determine the type 
of system with which we end up, and they serve to 
remind us that the real objective of the system is to 
deliver information. They also emphasize the fact 
that the critical part of the system is the data 
handling subsystem, for it is this component of the 
overall system, together with the suite of sensors 
that are used, which determines the type and 
quality of information we ultimately receive. Thus 
the answers to these three questions have a fundamental impact on all aspects of the system 
design beginning with the choice of sensors and orbit characteristics, and reaching all the way 
through the entire processing system. The role of the spacecraft and launch vehicle is simply to 
deliver the instrument(s) to a certain point in space and time in order to make a set of 
measurements and send the resulting data to earth. Viewing the problem in this way changes 
one’s perspective on the relative importance of various system components compared to that 
which has traditionally been held. It leads directly to the view that a remote sensing system is in 
reality an information delivery system, the raison d'etre of which is to deliver information about the 
state of the planet to those who require this information for the purposes of making decisions 
which affect management of the environment and resource development. 
An Information Delivery System 
This section deals with the structure of an information delivery system suitable for the tasks 
outlined above, and the technological aspects and challenges of realizing such a system. Figure 
2 illustrates in schematic form, the basic structure of an information delivery system which uses 
remote sensing of the earth system as its major set of inputs. One can think of the information 
delivery system as consisting of a series of steps beginning with the acquisition of a measurement 
by a spaceborne sensor. This measurement is then transmitted to the ground where it is archived 
in its raw form. Processing of a measurement takes place in several steps: A preprocessing step 
converts the raw digital number (DN) into a calibrated quantity which is proportional to the 
electromagnetic radiance entering the entrance aperture of the instrument at the time of 
acquisition. At this stage is the process, the positional location of each measurement is described 
in a coordinate system which is dependent on the particular ground track and dynamic behavior of 
the spacecraft platform and the instrument itself, (i.e. the spatial location of the measurements is 
Figure 1 Data Flow