Geoinformatics 2007 Conference (17–18 May 2007)
Paper No. 3-11
Presentation Time: 2:30 PM-4:30 PM

GIS INTERPOLATION OF STRATIGRAPHIC CONTACT DATA TO RECONSTRUCT PALEOGEOGRAPHY: DERIVING A PALEOLANDSCAPE MODEL OF THE BASE ZUNI SEQUENCE AND SUBSEQUENT CRETACEOUS ISLANDS IN CENTRAL TEXAS

PROCHNOW, Shane J.1, MULLINS, Melissa2, CAPELLO, Stephanie V.3, PERRY, Anna F.3, AHR, Steven W.3, WESTFIELD, Isaac T.3, FALLERT, Kari L.4, GAO, Song4, HARTZELL, Kirstin T.4, and LANG, Kenna R.4, (1) Center for Applied Geographic and Spatial Research, Baylor University, Waco, TX 76798, Shane_J_Prochnow@Baylor.edu, (2) Center for Reservoir and Aquatic Systems Research, Baylor University, Waco, TX 76798, (3) Department of Geology, Baylor University, Waco, TX 76798-7354, (4) Department of Environmental Studies, Baylor University, Waco, TX 76798

Geographic Information Systems (GIS) raster modeling technology might constitute a quantum leap in visualization and data interpretation for paleogeographic studies.  Our research uses raster modeling to interpolate paleotopography and stratigraphic thickness maps within a roughly 4,200 km2 study area in south central Texas (Fig. 1).  Only readily available geospatial data was used.  High-resolution geological maps (1:24,000 scale) published by the Bureau of Economic Geology (BEG) at the University of Texas at Austin were digitized and georeferenced using ArcGIS software products.  Paleosurface modeling focused on four stratigraphic features, the: 1.) basal Zuni Sequence boundary set on Precambrian granite and Cambrian-Ordovician marine sedimentary rocks; 2.) fluvial to deltaic Hensel Sand Formation (Lower Cretaceous); 3.) marine to tidal Glen Rose Formation (Lower Cretaceous); and 4.) marine Walnut Clay Formation (Lower Cretaceous).  The basal Zuni sequence boundary is a continental-scale, angular unconformity, while the younger surfaces involved in this study are relatively conformable.  Stratigraphic contact traces between these surfaces were converted to three dimensional (3D) data by extracting elevation values from 28 m resolution digital elevation models (DEM) developed by the United States Geological Survey (USGS) National Elevation Dataset (NED) using ArcGIS 3D Analyst.  The NED has a published vertical accuracy of 7 m to 14 m.  ArcGIS Geostatistical Analyst was then used to interpolate paleosurface rasters using the simple Kriging method based upon 12 nearest points along contact traces and their respective elevation values.  The elevation of the interpolated raster was corrected for structural deformation since the Cretaceous using the ArcGIS Spatial Analyst Raster Calculator by setting the base of the Walnut Clay (mostly coincident with the upper surface of the Glen Rose) as a Z-axis datum and adjusting the upper surface of the Hensel Sand and the base of the Zuni Sequence elevation, respectfully.  The interpolated paleosurface rasters have estimated vertical root mean square and standard error within 5 m, roughly the same as the published NED accuracy used for the derivation of elevation.  The thickness of the Hensel Sand and Glen Rose was also estimated by using the Raster Calculator to subtract their corrected bounding surface interpolations.  The thickness maps were verified by using published measured section data that accompanied the original geologic maps.  Three dimensional modeling of the paleosurfaces was visualized in ArcGIS ArcScene.  The paleorelief for the base of the Zuni Sequence in the study area was about 351 m.  The interpolated paleosurface grid for the base of the Zuni Sequence has an estimated vertical root mean square error of 1.8 m and an average vertical standard error of 4.9 m.  The total paleorelief for the upper boundary of the Hensel Sand is estimated at 356.5 m, but is locally lower in relative elevation than the base of the Zuni Sequence unconformity.  The interpolated upper surface grid for the Hensel Sand has an estimated vertical root mean square error of 3.7 m and an average vertical standard error of 4.9 m.  The total paleorelief for the lower boundary of the Walnut Clay (upper boundary of the Glen Rose) is estimated to have been 197.0 m, but is locally lower in relative elevation than both the base of the Zuni Sequence unconformity and the upper surface of the Hensel Sand.  The interpolated lower boundary surface grid for the Walnut Clay has an estimated vertical root mean square error of 4.4 m and 1.4 m of average vertical standard error.  The total thickness of the Hensel Sand within the study area ranged from 0 to 96.6 m thick.  The Hensel Sand is locally absent on paleohighlands (interfluves) of Paleozoic rock that constitute about 7.5% of the study area (Fig. 2A).  The Glen Rose ranges from 0 to 183.8 m thick in the study area.  The area of zero Z-axis values (Glen Rose thickness) indicate terrestrial exposure above where the marine Glen Rose was deposited, probably reflecting islands that persisted throughout the deposition of this first marine unit of the Zuni transgression (Figs 2B-2C).  These islands shifted up dip (northwest) relative the Paleozoic interfluves, but still account for about 7.5% of the total interpolated grid cells (Fig. 2).  The paleotopography at the base of the Zuni Sequence may have been largely controlled by faulting, but was comparable in total relief and slope as the modern topography of the study area.  The interfluves during the Hensel deposition may preserve well developed paleosols because these features were never scoured by Cretaceous fluvial systems, exposed to the atmosphere longer, and eventually buried by subsequent low-energy marine sediments.  The Glen Rose eventually buried most of the Paleozoic highlands (interfluves) in the center of the study area (Figs. 2A-2B), but laterally terminates on islands of Hensel Sand in the northwestern portion of the study area (Fig. 2B-2C).  This suggests that island systems may have existed in the study area throughout the deposition of the Glen Rose.  The paleo-islands in the study area were subsequently buried by the locally thin (< 2 m thick) Walnut Clay.  Again, these paleo-islands may better preserve well-developed paleosols since they were buried by low energy deposits and exposed to the atmosphere for a longer interval than other areas where the upper boundary of the Hensel Sand is preserved.  Raster GIS modeling also allows for the rapid and consistent calculation of unit thickness, surface slopes, and boundary relationships across large areas to better analyze geologic data and serve as a predictive tool.  This study demonstrates how raster GIS modeling can be used in particular with stratigraphic and paleogeographic studies by spatially interpolating known data into buried, obscured or missing areas.  The ability for GIS technology to characterize paleogeographic features, including paleo-islands, has potentially profound implications for geological study.  These implications include, but are not limited to, identifying paleotopographic features for paleontological studies of terrestrial habitats, refining depositional models and interpolating geologic surfaces for petroleum exploration, and studying the connection between paleolandscape position and paleosol formation and preservation. 

 

Figure 1. A.) The study area is located in Central Texas where Paleozoic and Cretaceous bedrock crops out.  B.) The coverage of published 1:24,000-scale geologic maps and the respective contact traces used for this study.

 

 

Figure 2. ArcGIS™ ArcScene 3D model of interpolated paleosurfaces.  A.) The upper boundary of the Hensel Sand (Cretaceous fluvial) over the base Zuni Sequence unconformity showing Paleozoic highland interfluves.  View is from the northeast  B.) Same as Fig. 2A, but including the upper boundary of the Glen Rose and showing Cretaceous paleo-islands.  C.) Same as Fig. 2B, but shown looking from the south.

Geoinformatics 2007 Conference (17–18 May 2007)
Session No. 3--Booth# 11
Geoinformatics Poster Session
University of California: Main Hallway
2:30 PM-4:30 PM, Thursday, 17 May 2007


© Copyright 2007 The Geological Society of America (GSA), all rights reserved. Permission is hereby granted to the author(s) of this abstract to reproduce and distribute it freely, for noncommercial purposes. Permission is hereby granted to any individual scientist to download a single copy of this electronic file and reproduce up to 20 paper copies for noncommercial purposes advancing science and education, including classroom use, providing all reproductions include the complete content shown here, including the author information. All other forms of reproduction and/or transmittal are prohibited without written permission from GSA Copyright Permissions.