Full NSF Report
Authors and Credits
By Peter Pearson
This document describes the columns of the Siple Dome 1996 dataset and how each was obtained:
Columns 1 & 2 are local coordinates using a Lambert projection centered at the summit of Siple Dome (-81.65434,-148.8081,623.23). The 1610 xy points corresponding to the location of radar waveforms points were derived by interpolation at intervals of ~ 100m from a set of 69 static GPS surveys of poles located along the traverse.
Columns 3 & 4 are latitude and longitude coordinates respectively, taken from the continuous kinematic GPS dataset along the traverse with spacing of ~ 70m. Since the location of these points does not correspond exactly with that of the radar waveforms, a program was written to select the closest set of lat/long coordinates for each radar waveform. Thus, these are all measured (not interpolated) data points. However, there were 2 sections for which there were no kinematic GPS data. For these intervals to which no lat/long coordinates could be assigned, the latitude and longitude were both set to 0 (entries 1024-1043 and 1149-1317).
Columns 5 & 7 are identical, and were repeated for graphing purposes. These columns contain the cumulative distance away from the summit along the travel path (which wasn't entirely straight). These data were obtained by calculating the distances between successive sets of Lambert coordinates, and then summing them and adjusting the running distance to be referenced to the summit.
Column 6 is the elevation of the surface of Siple Dome in HAE (height above ellipsoid). These data were also obtained from the continuous kinematic GPS. The same procedure used in columns 3 & 4 for selection of the closest kinematic GPS value toa radar waveform was used here. However, in the intervals where there were no GPS data (see above), the elevations were linearly interpolated between the 2 endpoints of each gap.
Column 8 is the elevation of the bed of Siple Dome in HAE. At each of the 1610 points, the 2-way travel time of the radar signal was determined by picking the midpoint time between the max and min phase of the bed-echo wavelet. These times were converted into depth using a measured average speed of 1.74040e8 m/s. These depths were then subtracted from the surface elevations.