extract velocity model of the last iteration form IterVel.out and save
it as MOD.true, change maxiter as 0 (line 12) and synthetic flag as 1
(line 19) in 'Surftomo.in', then rerun the code
add model uncertainty estimation with random models, also indent all the
code to make them a little bit easy to read
bug fix about shift when output the velocity model
segmentation fault appears in the case of very large dataset, there will be problem if the nonzeros of G matrix exceed the range of INTEGER, now change maxnar = spfra*dall*nx*ny*nz, remember order matters.
also show the lat and lon when station outside the inverted region
