Environmental information extraction from satellite remote sensing data

Remote sensing is a particular instance of the more general measurement process, which is intrinsically an inverse problem. Some of the key issues of inverse theory in the context of the interpretation of remote sensing over terrestrial surfaces are outlined. Earth Observation from space provides a unique opportunity to acquire data on the state and evolution of the environment globally repetitively, and at spatial resolutions suitable for the simulation of dynamic processes in the relevant geophysical media. Various approaches can be used to address this inverse problem, including the formal inversion of physically-based or empirical surface reflectance models against remote sensing measurements, or more empirical approaches such as the exploitation of correlations between the variables of interest and spectral indices derived from the measurements. This field of study will evolve considerably in the next few years as a result of the introduction of a new generation of space sensors with greatly increased performances and capabilities. The higher quality data generated by these sensors will permit the evaluation of the theoretical bases of existing models, and the latter will foster a much better exploitation of the data. However, scientific and technological advances will need to be coordinated to take full advantage of the opportunities, as the availability of reliable models and high performance computers and networks will rival as bottlenecks to the full exploitation of Earth Observation data.