| |
|
Description: |
In several papers originating within the Geophysics and Signal Processing community, the remote sensing of planet surface properties by antennae mounted on satellites is discussed. The mapping methods proposed there are based on a Backus-Gilbert theory. This is an interpolation theory based on least squares principles, and is incompatible with point data.
I shall try to illuminate some connections between the Backus-Gilbert theory and interpolation methods of moving least squares type, and will show how the use of improved antennae in conjunction with the Backus-Gilbert approach can lead to unsatisfactory maps of surface phenomena.
My talk may serve to alert the remote sensing community to the need for a more stringent mathematical analysis of their methods, and suggest fruitful areas for research in numerical analysis for those in studying interpolation methods.
Area(s):
|
|
Date: |
|
| Start Time: |
14:3 |
|
Speaker: |
K. Salkauskas (University of Calgary, Canada)
|
|
Place: |
Room 5.5
|
| Research Groups: |
-Numerical Analysis and Optimization
|
|
See more:
|
<Main>
|
|
