Balancing good data with bad

Roughly, choosing the size of $\bold \epsilon$ chooses the stiffness of the curve that connects regions of good data. Our first test cases gave solutions that we interpreted to be too stiff at early times and too flexible at later times. This led to the idea of modifying the goal $\bold 0\approx\bold D\bold u$to include a depth-dependent scaling. The first try would be proportional to the travel-time-depth $\tau$.Thus we define a diagonal matrix $\bold \Lambda = {\bf diag}(\tau)$and revise our model-damping goal to be $\bold 0\approx\bold\Lambda\bold D\bold u$.