From: hogrefek@geo.oregonstate.edu
Date: August 7, 2009 4:49:36 PM PDT

Even though I pulled out every trick in my book, the statistical results are not as nice as [hoped for]. This is the effect of the clouds/haze and or lots of sea surface glint in the images, limited multibeam coverage in shallow areas (to glean baseline data) and the very steep slopes around the island. However, the results are definitely better than what I managed during my master's work.

...The grid does provide for good terrain representation and should [serve well for tsunami run-up modeling]. I would suggest some sort of "high pass filter" algorithm to smooth out the "noise" before using it in such a model.... If you want to see a best case scenario for bathymetric derivation accuracy, I suggest you check out the results for Johnston, Kingman, Palmyra and/or Saipan....