REMOTE-SENSING SCALES RELATED TO THE FREQUENCY OF NATURAL VARIATION - AN EXAMPLE FROM PALEO-ICE-FLOW IN CANADA

Determining the appropriate or optimum remote sensing scale for use in specific applications in geoscience is a recurrent problem of compromise. Discussed is the importance of matching remote sensing scales to the frequency of variation in nature. This is illustrated by reference to the author’s work on the search for pattern amongst noise within the geomorphic imprint produced by the last North American Ice Sheet. Utilization of six scales of remote sensing in the mapping of ice-flow landforms has led to a radical new interpretation of the dynamics and behavior of the Laurentide Ice Sheet. It is shown that incomplete sampling of the frequencies of natural variation can lead to misleading or incorrect interpretations. This clearly has implications for remote sensing investigations of many other natural phenomena. © 1990 IEEE