Concluding remarks

Abstract

In opening this Discussion Meeting, Professor Houghton said that it was the aim of the organizers to put most emphasis on the consideration of problems of the interpretation of data from remote sensing. That aim does indeed seem to have been achieved and we have had some very useful studies of the significance of data and of the principles and methods used in interpretation. Two points have in particular occurred to me. In the first place, as with other types of satellite observation, the techniques of observation and the capacity for obtaining very large amounts of data are outrunning the understanding of the results and the ability to extract significant conclusions; it should, however, be noted that whereas in research it is the new, ill-understood observation that leads to further progress, the many, well-understood, routine observations are the ones that are needed in applications. The second point, and it is related to the first, is the great complexity of some processes that produce detectable signals, so that the interpretation of the response of a remote sensor may only be possible in some limited, well-defined circumstances. A clear instance was given by Dr Curran in his discussion of the determination of leaf area index, while Dr Tucker showed how incompletely the imaging of sea waves is understood.