ETL 1110-1-183
1 Apr 98
ellipsoidal height differences, and the quantity )N is the
relative geoidal heights from a station with a known
NAVD88 elevation, than is possible from the direct
geoidal height difference computed from the geoid model
application of absolute geoid heights to GPS networks.
(see Figure 1).
b. This analysis was based on various methods used
b. The expected precision of the orthometric height
for determining NAVD88 elevations from GPS
from using GPS relative positioning and modeled geoid
ellipsoidal height data. These methods were tested on a
heights and the above relation can be calculated by the
network of points having known first-order leveled
summation of variance components corresponding to the
orthometric heights that were tied to first-order vertical
accuracy of the published orthometric height, the GPS
control. Results of the testing indicated that GPS-based
relative height determination, and the computed geoid
surveys could determine NAVD88 elevations to an
height differences.
accuracy of 30 mm when relative heights and differences
in geoid heights are applied.
c. Positional accuracy for orthometric heights on
benchmarks must be obtained from published sources
c. Note that the accuracy of NAVD88 elevations
based on the results of a vertical network adjustment.
determined from DGPS-derived heights and geoid
Without this information it is presumed that a fixed
modeling is dependent on the accuracy of the GPS
vertical control point contributes no additional error to the
coordinate solution and the geoid model.
height of the unknown stations. The uncertainties in GPS
relative heights are estimated from the vertical component
A-7. Additional Guidelines and
error estimate produced from the GPS data processing
Recommendations
and adjustment software. An error estimate of 10 mm is
commonly seen as the minimum baseline error produced
from static type surveys. Relative geoidal height (F)N)
In addition to the guidelines discussed in section A-5, the
following procedures and methods are recommended and
precision from geoid modeling can have an expected
should be implemented when using GPS for elevation
standard deviation of 10-20 mm.
determination:
d. The aforementioned error values lead to an
a. Keep project areas within a 20-kilometer radius
expected uncertainty in final orthometric height at the
of control points. GPS relative positioning accuracy
unknown station of approximately 30 mm (at the 95%
depends in part on the length of the measured baseline.
confidence level) relative to the published elevation at the
Positioning errors grow in direct proportion to baseline
benchmark reference station. A repeatable accuracy of 30
length at a rate of approximately 1 part per million. For
mm meets or exceeds most feature elevation tolerances
networks with an area less than 20 km, the distance
specified for many USACE surveying and mapping
dependent error in the GPS vertical component (relative
projects, excepting certain high precision surveys such as
ellipsoid height) will be limited. Occupation times of less
for structural deformation monitoring. In areas with
than 1 hour (i.e., 20-30 min) should produce good results
obstructions, dense vegetation, or high relief between
for these shorter baselines. For project areas greater than
monuments or projects site, GPS can exceed leveling
20 km, the occupation times should be increased to a
accuracy when time is critical to the project.
minimum of 2 hours for primary and secondary control
A-6.
Results of Field Testing
points. Control points should be spaced throughout
(surrounding and within) the project area.
a. Based on an evaluation of DGPS data and geoid
b. Observe when VDOP is less than 5.0. Vertical
modeling software capabilities by CETEC, it was
Dilution of Precision (VDOP) is a measure of vertical
determined that higher accuracy elevations are obtained
positioning accuracy (due mainly to satellite geometry)
by the transfer of ellipsoidal height differences and
A-4
Previous Page
