March 19, 2010

Hi David,

Our GJI study explored some of the intricacies of estimating strain-rate (and rotation-rate) fields from a set of spatially irregular velocities (with or without verticals).  It sounds like you want the strain-rate fields only.

Attached are these files (zipped):

onesub.csh
onesub_reason_3D.ps
sixsub_reason_3D.ps
socal_d01_q03_q07_b1_3D_s1_u1_masked_pts.dat
socal_d01_q03_q07_b1_3D_s1_u1_strain.dat

onesub.csh is a GMT script used to generate onesub_reason_3D.ps. (This is only a guide to show what columns should be plotted and will not work directly.)
sixsub_reason_3D.ps is Figure 11 of Tape2009.

I have also included four additional files, each with one line of headers indicating each columns.  D is for strain-rate, W is for rotation rate.  The 2D/3D tag in the file names indicates whether verticals were not used -- verticals have a major effect on W but not on D.

socal_d01_q03_q07_b1_2D_s1_u1_Dtensor_6entries.dat
socal_d01_q03_q07_b1_2D_s1_u1_Wtensor_3entries.dat
socal_d01_q03_q07_b1_3D_s1_u1_Dtensor_6entries.dat
socal_d01_q03_q07_b1_3D_s1_u1_Wtensor_3entries.dat

All files were estimated using spherical wavelets with scales 3-7, so relatively coarse.  (For example, in the Parkfield region would allow up to scale 9, as indicated in Figure 15 of our paper.

Detailed documentation is here:
http://www.people.fas.harvard.edu/~carltape/research/pubs/Tape2009gps.pdf

But it might be quicker to ask me anything in particular.  Let me know if you are looking for something else.

Thanks,
Carl