2006 INEGI Sierra Cucupah Empirically Corrected Lidar Dataset

These airborne lidar data were gathered by the Instituto Nacional de Estadistica y Geografia of Mexico as part of a regional mapping activity in northwestern Mexico. They span the area that ruptured in the April 2010 M7.2 El Mayor Cucupah earthquake which was laser scanned and for which data are available in OpenTopography's holdings. Alejandro Hinojosa of CICESE is the contact person for these data. This version of the data has been empirically corrected by Craig Glennie and colleagues at the University of Houston. Details for the data corrections as follows: Rather than trying to correct the whole dataset, we just concentrated on the portion that overlaps with the post-event data. Here is a brief summary of what we did: (1) Pre-Event Data was given in ITRF 1992 (1988.0 epoch) and post-event NCALM data was processed in ITRF2000 (Epoch 2010.627). NGS software package HTDP was used to compute a coordinate shift between these two reference frames (-0.900 m East, 0.429 m North, 0.004 m Up). To correct the 2006 data to the same datum as the NCALM data, we added the vector (-0.900,0.429,0.004) to all of the pre-event data points. (2) Original dataset contained all scan data out to +/- 28 degree scan angle. There are significant problems at the outer edge of the scan, so all scan lines were cropped to +/-24 degrees. This results in minimal overlap between scan lines, but doesn't create any data gaps between flight lines. (3)Dataset was then re-boresighted. We determined a roll and pitch offset for each flight line individually, plus a global mirror scale factor. (4) Finally, we determined an individual delta "z" correction for each flightline. Note that for all of the above adjustments, none of the post-earthquake data was used. We purposely sequestered the two datasets so as not to inadvertently remove differences caused by the earthquake. To give an idea of the magnitude of the improvement, on the pre-event dataset (as delivered to me) in the overlap, we were seeing average elevation differences of 95 cm (1 sigma). After cropping the data to 24 degrees, the average elevation differences were 70 cm (1 sigma) After steps (3) and (4) above, the average elevation differences were reduced to 52 cm (1 sigma). So overall, it appears we were able to reduce the vertical errors by almost a factor of two.