|
<br />differences between the data sets are in the range of 23-87 em. In terms of monitoring ground topography within
<br />vegetation, the precision found in the RAMS data are unacceptable.
<br />Brock and Sallenger (200 I) detennined that aircraft location and pointing need to be known to within 5 em to obtain
<br />height accuracies of 10 em. When we pressed selected commercial LIDAR companies to obtain absolute vertical
<br />accuracies of 15 em, we were told that current IMU technology does not provide such accuracies and that significant
<br />ground control would be needed during the data collection to ensure this level of accuracy. This of course will
<br />significantly increase LIDAR collection costs.
<br />
<br />Sediment volume accuracy
<br />One reason topography is measured in the Colorado River ecosystem is to monitor the volumetric changes of the
<br />sediment deposits as a function of darn flow regime. To detennine the effect that the observed airborne elevation
<br />accuracies have on estimating the volume of sediment, we selected sites with the densest ground surveys within bare-
<br />sand and vegetated-sand deposits. Our analyses addressed three questions related to estimating terrestrial sediment
<br />volume from airborne elevation data. (I) How do the volume errors differ between bare ground and vegetated ground?
<br />(2) What is the effect on volume error in ignoring elevation points within vegetation, where vertical accuracies are
<br />lowest in airborne data? (3) Do any ofthe data sets meet or even closely approach the 3% accuracy in volume estimates
<br />that is currently obtained by ground surveys?
<br />For these assessments, two DEMs were generated from each LIDAR point coverage. The first LlDAR DEM
<br />excluded all first-return elevations in vegetated areas, but, in the case of the RAMS data, second-return data points were
<br />included in the DEM generation. The second LlDAR OEM did not exclude any first return data points, unless there
<br />was a second return. The average elevation along the lower..devation shoreline of each DEM was used as the base
<br />datum for calculating sediment volume of each 25-cm OEM cell. The sediment volume from a particular DEM was
<br />detennined by swnming the cell volumes within that DEM. For each test site, the ground-surveyed volume was
<br />subtracted from a particular airbome-derived volume; this volume difference divided by the ground-surveyed volume
<br />produced the percent volume difference or error.
<br />The results show that almost all LlDAR and photogrammetry data produce larger volumetric errors on vegetated
<br />ground than on bare ground. Volumetric errors for bare ground, vegetated ground, and total ground are mostly lower
<br />when both bare- and vegetated-ground data are used to estimate volume, which suggests that the topographic
<br />modulations in the vegetated terrain exceed the vertical elevation errors in the airborne data. Most of the LIDAR
<br />volumetric errors that are less than 3% were obtained on the bare-ground surfaces. On the other hand, the vofumetric
<br />errors from the photogrammetry data in our bare-ground and total-ground aSsessments were less than 3% and the
<br />volumetric errors in vegetated terrain were in the range of3.5% to 9.6%.
<br />
<br />CONCLUSIONS
<br />
<br />Considering the vertical elevation and volurretric accuracies provided by LlDAR and photogrammetry for both bare
<br />and vegetated ground, we have to conclude that photogrammetry provides better data for monitoring, but also requires
<br />more carefuJ processing to satisfy the stringent GCMRC accuracy requirements for monitoring the sediment deposits
<br />within the Colorado River ecosystem.
<br />
<br />REFERENCES
<br />
<br />Beus, S. 5., Avery, C. C., Stevens, L. E., Kaplinksi, M. A., and Cluer, B. L., 1992. The influence of variable discharge
<br />regimes on Colorado River sand bars below Glen Canyon dam. Chapter 3. In Beus, S. S. and Avery, C. C.
<br />(Eds.), The influence of variable discharge regimes on Colorado River sand bars below Glen Canyon dam:
<br />1991 Annual report. Northern Arizona University, Geology Department, Flagstaff, Arizona, 26p.
<br />Blank, B. L., 2000. Application of digital photogrammetry to monitoring sand bar change in Marble Canyon, Arizona.
<br />Report to the Grand Canyon Monitoring and Research Center Physical Resource Program, Utah State
<br />University. Department of Geology. Logan. Utah. 47 p.
<br />Bowen, Z. H., and Waltennire. R. G., 2002. Evaluation of light detection and ranging (LIDAR) for measuring river
<br />corridor topography. Journal of the American Water Resources Association, 38(1). pp. 3141.
<br />Brock, J., and Sallenger, A., 2001. Airborne topographic tidar mapping for coastal and resource management. U.S.
<br />Geological SUlVey Open File Report 01-46. 4 p.
<br />Hazel. J. E.; Kaplinski. M. Parnell. R.. Manone. M" and Dale, A., 1999. Topographic and bathymetric changes at thirt-
<br />three long-term study sites. In Webb. R. H,. Schmidt. J, C,. Marzolf. G. R., and Valdez. R. A, (Eds.). The
<br />Controlled Flood in Grand Canyon. A GU Geophysical Monograph Series 110. pp.161-184,
<br />
<br />EV ALUAll0N OF LIDAR AND PHOTOGRAMMETRY FOR MONl1ORING VOLUME CHANGFS IN
<br />RIPARIAN RESOURCES WITHIN 11IE GRAND CANYON, ARlWNA
<br />
<br />Pecora ISlLand Satellite Information IVIISPRS Commission IfFIEOS 2002 Conference Proceeding!l
<br />
|