GLOBAL POSmONlNG SYSTEM AND TOTAL STATION FIELD METHODS UTILIZED TO DERIVE THE ACCURACY OF USGS lO-METER DIGITAL ELEVATION MODELS ON OAHU A THESIS SUBMITTED TO THE GRADUATE DMSION OF THE UNIVERSITY OF HAW AI'I IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS IN GEOGRAPHY MAY 2008 By Joseph R. Silver Thesis Committee: Everett Wingert, Chairperson Matthew McGranaghan Kim Bridges We certifY that we have read this thesis and that, in our opinion, it is satisfactory in scope and quality as a thesis for the degree of Master of Arts in Geography. THESi2 ~~son ii ACKNOWLEDGMENTS Thanks to the staff at the United States Geological Survey offices in Menlo Park, California and Reston, Virginia for their willingness to demonstrate the 10-Meter DEM production process. Additionally, I would like to thank my committee; Everett Wingert, Matthew McGranaghan, and Kim Bridges for their guidance and support. iii TABLE OF CONfENTS Acknowledgements ................................................................................. .iii List of Tables ..........................................................................................v List of Figures ........................................................................................vi List of Abbreviations ...............................................................................vii Chapter 1. Introduction ............................................................................... 1 Chapter 2. Literature Review ........................................................................5 Chapter 3. USGS DEM Production ............................................................... 13 Chapter 4. Methodology ........................................................................... 18 4.1 Problem Statement..................................................................... 18 4.2 Field Site Selection.................................................................... 20 4.3 GPS Data Collection .................................................................. 22 4.4 Total Station Survey .................................................................. 25 4.5 Post Processing ........................................................................ 29 4.6 Analysis ................................................................................. 30 Chapter 5. Analysis ..................................................................................33 Chapter 6. Conclusions .............................................................................74 Appendix: DEM Scanning Procedures ...........................................................81 Bibliography .........................................................................................92 iv LIST OF TABLES Table ~ 4.1 Field Site Areas ............................................................................ 27 5.1 Statistical Analysis ofField Site 1.. .....................................................37 5.2 Statistical Analysis ofField Site 2 .......................................................44 5.3 Statistical Analysis offield Site 3 .......................................................51 5.4 Statistical Analysis ofField Site 4 .......................................................58 5.5 Statistical Analysis of Field Site 5 .......................................................65 5.6 Statistical Analysis ofField Site 6 .......................................................72 v LIST OF FIGURES Figure ~ 2.1 Banding and Striping Effect ............................................................... 7 3.1 The Equation to Calculate Root Mean Square Error .................................. 16 4.1 lO-Meter Pixel Diagram of Resource Grade GPS Error.............................. 19 4.2 Generalized Method ........................................................................20 4.3 The USGS Quadrangle Boundaries for Oahu ...........................................20 4.4 Field Site Locations .........................................................................21 4.5 Ashtech Survey Grade GPS Receiver ...................................................22 4.6 Base Station Location on the Physical Science Building ............................23 4.7 The Base Station PSB 3 Monument .....................................................23 4.8 Leica TC 305 Total Station ...............................................................26 4.9 Leveling the Total Station ................................................................27 4.10 Examples of Radiant Total Station Shots ..............................................28 5.1 USGS DRG and Hillshade Context Image ofField Site 1.. .........................33 5.2 Total Station Points Overlaid on USGS DEM 10-Meter Grid-Field Site 1... ........ 34 5.3 3D Model ofField Collected Surface & USGS DEM Surface-Field Site 1... ......35 5.4 USGS DRG and Hillshade Context Image ofField Site 2 .......................... .40 5.5 Total Station Points Overlaid on USGS DEM 10-Meter Grid-Field Site 2 .........41 5.6 3D Model ofField Collected Surface & USGS DEM Surface-Field Site 2 ........42 5.7 USGS DRG and Hillshade Context Image ofField Site 3 .......................... .47 5.8 Total Station Points Overlaid on USGS DEM 10-Meter Grid-Field Site 3 ........ .48 5.9 3D Model ofField Collected Surface & USGS DEM Surface-Field Site 3 ........49 5.10 USGS DRG and Hillshade Context Image ofField Site 4 ........................... 54 5.11 Total Station Points Overlaid on USGS DEM 10-Meter Grid-Field Site 4 ......... 55 5.12 3D Model ofField Collected Surface & USGS DEM Surface-Field Site 4 ........ 56 5.13 USGS DRG and Hillshade Context Image ofField Site 5 ........................... 61 5.14 Total Station Points Overlaid on USGS DEM 10-Meter Grid-Field Site 5 .........62 5.15 3D Model ofField Collected Surface & USGS DEM Surface-Field Site 5 ........ 63 vi 5.16 USGS DRG and Hillshade Context Image offield Site 6 ...........................68 5.17 Total Station Points Overlaid on USGS DEM 10-Meter Grid-Field Site 6 .........69 5.18 3D Model ofField Collected Surface & USGS DEM Surface-Field Site 6 ........70 6.1 Slope Map of Honolulu Quadrangle Section ...........................................77 vii LIST OF ABBREVIATIONS D3D DELTA 3D USGS Proprietary Software Program DEM Digital Elevation Model DRG Digital Raster Graphic EDM Electronic Distance Measure ESRI Environmental Systems Research Institute EXTRON USGS Digital Scanner GIS Geographic Information System GPS Global Positioning System IDW Inverse Distance Weighting LT4X USGS Proprietary Software Program NASA National Aeronautics and Space Administration NGS National Geodetic Survey NSRS National Spatial Reference System PDOP Percent Dilution of Precision PSB Physical Science Building RMSE Root Mean Square Error TIN Triangular Irregular Network TID Tactical Terrain Data USGS United States Geological Survey UTM Universal Transverse Mercator WGS84 World Geodetic System of 1984 viii Chapter 1. Introduction Three-dimensional models are useful in terrain representation and mapping. Map quality, whether two or three dimensional, is limited by the accuracy of the data and resources used to create the finished product. Digital Elevation Models (DEMs) are one of the primary sources of data for three-dimensional maps. Although obtaining DEM data for the United States is easy because of the United States Geologic Survey (USGS) and the accessibility of their digital data sets, this does not assure the accuracy of the data (Childs, 2001). DEM data is only as accurate as the techniques used in the generation process. This project analyzed the accuracy of 10-Meter USGS DEMs utilizing a Global Positioning System (GPS) and Total Station surveying equipment. Reliable and accurate data for use in cartographic representation are difficult to acquire, and data provided by institutions such as the USGS are no exception. Inaccuracies have been found in 30-Meter DEMs of Oahu in past research and can be verified by viewing images made from those data sets. When viewed through standard DEM viewing software, errors can be seen on sloped and flat areas (Clouet, 1997). This process of visual inspection, or visualization, is crucial before using such data. Portions ofWaikiki and other coastal flatlands on Oahu are represented as being below sea level in both the 30-Meter and lO-Meter DEM datasets. Coastal areas with more relief, such as the Northern shores ofKauai, depict sea level as being eight meters high. This becomes problematic when all elevations within the DEMs of Hawaii are based on sea level as the absolute zero. This could point toward a problem not only in the data production method, but also in the data source used to generate the models. 1 Past research on DEM accuracy has examined photogrammetrically derived data sources as well as current array sampling methods. Depending on the accuracy thresholds applied in the studies, the results varied. An accuracy assessment of elevation data using a Global Positioning System found that USGS l-degree/3-arc-second DEMs in Texas were of high vertical, z dimension, quality and met stated vertical accuracy standards (Adkins and Merry, 1994). This accuracy standard for the I-degree 30-meter by 30-meter grid cell DEMs had a range ofplus or minus 30-meters in the vertical, z, dimension. This is not satisfactory for many cartographic and planning opemtions. An engineer does not want to know plus or minus 30-meters, or IO-meters for that matter, how much earth to move on a job site. He or she