Publication Date:
2016-06-21
Description:
With the increasing volume of satellite observations, dimension reduction techniques are more and more important for storage or transmission. Furthermore, they are essential for inversion schemes that, in practice, cannot handle the huge amount of information provided by modern instruments for near real time inversion. In this paper, we compare the theoretical advantages and limitations of the two general strategies: compression (i.e., feature extraction) and channel selection (i.e., feature selection). The statistical “Input Variable Selection" framework is adopted to revisit the basis of these remote sensing techniques. The flexibility of these methods to specific applications is demonstrated. Special emphasis is put on the optimization of the observation dimension reduction for the simultaneous retrieval of several variables (e.g., temperature and humidity). In addition to considering the signal-to-noise ratio for the variable to retrieve, we propose to account for the contamination by other unknown variables. We also introduce a new approach named “Bottleneck Channels" (BC) that combines compression and channel selection techniques and can therefore benefit from the advantages of both strategies. Various configurations of the BC approach can be considered: strict, linearly or non-linearly projected, each with advantages and drawbacks. In the companion the paper, experiments will be conducted on microwave data to illustrate the practical advantages of each approach. BCs are able to compress and suppress noise in a similar way as PCA and they can be interpreted as channels like in a channel selection.
Print ISSN:
0035-9009
Electronic ISSN:
1477-870X
Topics:
Geography
,
Physics
