Field data collection utilized a Trimble NetRS GPS receiver with a Zephyr Geodetic antenna as a fixed Geodetic GPS base station located at the Barrow Arctic Science Consortium (BASC). This base station provides differential correction of data files at 1 second intervals as well as real time correction to field rover GPS units within approximately five kilometers and in line of sight of site of the Trimble HPB450 radio transmitter that is connected to the base station . The base station was upgraded in 2005 from a Trimble GPS 5800, which was used during the 2003 coastal survey described below. The rover system used for 2007 surveys included a Trimble R7 with internal radio that communicated with the base station radio transmitter for real time DGPS correction and a TSC2 survey controller that logged field data. Field surveys in 2003 and 2006 utilized a Trimble 5700 receiver, Zephyr antenna and a Trimble TSC1 controller. All DGPS equipment was provided by UNAVCO (www.unavco.org) to the local science logistics provider (BASC).
A total of eight surveys of the 10.7 kilometer ACD BEO observatory were conducted. These included August 2003, 2006, June 2007, August 2007 June 2008, August 2008, June 2009 and August 2009. During each survey, field personnel walked along the top of the coastal bluff edge. The initial surveys for the 2003 season were conducted by Serbin et al (2003). In subsequent years surveys were conducted by the lead author of this paper following previously documented survey methodology documented in Serbin et al. (2003). For field surveys, the rover receiver was installed in a standard Trimble backpack configuration with the GPS antenna mounted on the bluff side of the backpack. After antenna height was corrected for the height of the field observer, this configuration allowed for the location of the horizontal and vertical position of the coastal bluff edge to be documented.
Two survey methods were used. Real Time Kinematic (RTK) surveys were conducted whenever possible. This required a radio link to the base station. RTK surveys allow for instant DGPS processing and the acquisition of centimeter-accuracy horizontal and vertical location data. The second method utilized a Post-Processed Kinematic (PPK) survey style, which did not require a radio link, but instead relied on an unobstructed On-The-Fly (OTF) initialization period and post-survey differential correction with GPS data logged at the base station to provide field survey data with centimeter accuracy. For both survey types, the rover receiver was set to log data at a high resolution of one-second intervals to more accurately delineate the generally irregular coastal bluff edge. Repeat surveys using these vertically and horizontally centimeter-accurate field methods provide a means to accurately monitor the geospatial dynamics of the coastal bluff edge as it changes with coastal erosion.
DGPS survey data was downloaded from the controller and imported into Trimble Geomatic Office Software version 1.63 (TGO). Data was exported in text files suitable for ingestion by ESRI's ArcGIS (Workstation 9.2) Geographic Information System (GIS) software. Data from PPK surveys were processed by downloading base station files that were utilized for differentially correcting data collected by the rover GPS unit using standard DGPS correction procedures. Processed data were exported as for that of the RTK surveys.
Processed DGPS survey data were imported to ArcGIS as X, Y point data and saved to a point shapefile. Erroneous points that resulted from poor GPS signal when field personnel traversed small coastal gullies were deleted. Using the polyline conversion tool in the ArcGIS extension AlaskaPak Version 2.0 for ArcGIS 9.2, point shapefiles were converted to polyline shapefiles. For each survey, the length of the coastline was computed for each of the four monitoring segments using the attribute data field calculator tool associated with the ArcGIS polyline tool. Using the polyline shapefiles as a determinant location of the coastal bluff at each survey time, polygon shapefiles were created to establish the area of coastline lost to erosion for each monitoring segment for the following survey periods: August 2003 to August 2007, August 2003 to August 2006, August 2003 to June 2007, June to August 2007, August 2003 to June 2008, August 2003 to August 2008, August 2003 to June 2009, and August 2003 to August 2009. The area of each polygon shapefile was computed using the calculate area command from ArcGIS Toolbox. Rates of erosion were calculated for the following periods mentioned above. To enhance inter-comparison of erosion rates between segments of different lengths and different sampling periods, results were normalized by the length of the coastal segment at the earliest survey associated with a given survey period and reduced to a year-long time frame for survey periods spanning the 2003 to 2006, 2007, 2008 and 2009 periods.