Al-Zadjali, Ibrahim and D.J. Stierman, Interpretation of Gravity from the Comayagua Graben, Honduras; Abstracts with Programs, North-Central Section of the Geological Society of America, Ames, IA, 1996, p. 25.
The Comayagua Graben of central Honduras, approximately 30 km long NNW to SSE, 15 km wide and 0,5 to 1 km below the surrounding highlands, is the most dramatic topographic low of the "Honduras Depression." Mountains east and west of this graben are capped by Tertiary volcanic rocks of Miocene age or younger. These volcanics overlie early Tertiary volcanics, Mesozoic sediments and Paleozoic metasediments that crop out on the north, northwest and northeast sides of the valley. We reprocessed and interpreted some unpublished gravity data collected by a consultant for the Honduran government in and around this graben. Prior to reducing these data and applying the same International Gravity Formula and Bouguer reduction density used for the rest of the Honduras data available to us, we had to convert UTM station coordinates listed in the consultant's report to latitude and longitude. A simple Bouguer anomaly profile running ENE across the northern half of the valley shows a residual gravity low of about -32 milligals. Assuming a density contrast of -400 kg-m-3, a 2-D Talwani forward model is consistent with a maximum valley fill thickness of 2700 meters and an average valley fill of more than 2000 meters. A more accurate calculation based on the complete Bouguer anomaly will no doubt result in a slightly shallower basin. A second profile across the southern half of the valley shows a narrower and shallower basin. This change in basin depth and width is abrupt, suggesting a fault striking at right angles to the valley axis.
NOTE: this abstract (as published) contains as error: because terrain corrections are largest for stations in rugged "bedrock" areas, complete Bouguer anomaly values will increase more at these stations (increase with respect to the simple Bouguer values) than at mid-valley stations, increasing rather than decreasing the amplitude of the residual low (and, of course, increasing the inferred depth to normal bedrock).