We investigated the
effects of GCP addition on consecutive model solutions by
examining the RMS error of adjusted GCP positions and the
total image unit weight (a good indicator of the quality of the
entire solution). Total image unit weight provides a measure
of the overall precision of control and tie-point measurements
and is calculated as the RMS (in pixels) of adjusted
control and tie-point residuals following the bundle adjustment
(Wolf and DeWitt, 2000). Lower image unit weights
indicate more tightly controlled model solutions. DEMs were
processed from each of the model solutions at 10m post
spacing using a triangular irregular network (TIN) data structure
and an adaptive terrain extraction matching algorithm
(using SocetSet software, BAE Systems), and were validated
by calculating elevation residuals between each GPS check
point and its spatially coincident DEM cell value.
We investigated the effects of different GCP configurations
on DEM error by differencing each lidar point-controlled
photogrammetric DEM (summer 2003) from a DEM derived
solely from the lidar data. This gave a straightforward qualitycontrol
test whereby differences closest to zero represented
the best-quality DEM surfaces. We also examined the effects
of GCP configurations on estimates of glacier volume change
by computing differences between photogrammetric DEMs
(summer 2003) and a DEM derived from repeat survey
lidar data from summer 2005, and