We first needed to develop consensus on standards, specifica-
tions, priorities and deliverables. Dewberry provided overall
project management and rigorous QA/QC, using checkpoints
from JOA Surveys. Data were delivered in the Alaska Albers
projection, then converted to multiple file formats for USGS
and NGA. Elevation data was delivered in both orthometric
and ellipsoid heights, using Geoid09 throughout for consis-
tency. Intermap’s Ortho-rectified Radar Images (ORIs) had
62.5-cm resolution, whereas Fugro’s ORIs had 2.5-meter res-
olution – both better than USGS’ 5-meter resolution require-
ment. Fugro delivered hydro-enforced DSMs and DTMs that
met USGS’ hydro specifications whereas Intermap’s stan-
dard hydro masks exceeded USGS specifications for lakes
and double-line streams. USGS received data that exceeded
specifications, but this caused edge-join issues and incon-
sistencies along seam lines between the two firms. Mapping
the easier terrain, Intermap routinely exceeded accuracy
specifications whereas Fugro struggled to satisfy accuracy
requirements partly because the P-band didn’t just penetrate
vegetation, but sometimes also penetrated the ground, de-
pending on soil wetness. Fugro developed an X-band/P-band
hybrid model for DTMs, but some issues remained. Geo-
SAR appeared to lower Denali’s peak elevation by 60+ feet,
possibly from IfSAR foreshortening, shadow or layover. This
caused Dewberry to launch two expeditions to climb Denali
in 2015 to GPS survey the official elevation of the top of the
ice and snow (at 20,310 ft), returning in 2016 to install a
USGS survey monument at Windy Corner and to use ground
penetrating radar (GPR) to measure the 6.1-meter depth of
ice and snow on the peak. Both years, Blaine Horner of Com-
passData led 4-person teams that carried equipment loads of
over 700 pounds from their base camp at 7,000’, to the peak,
under extremely difficult conditions.
204
April 2020
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
Figure 3. Rather than being funded efficiently for large contiguous blocks in two
years, small non-contiguous blocks were mapped each year between 2010 and
2019 as funds became available from different stakeholders.This also impacted
Dewberry’s ability to efficiently acquire annual QA/QC checkpoints in advance of
accuracy testing.
In addition to these technical and logistical challenges, USGS
lacked the funding necessary to acquire the data efficiently
in large contiguous blocks over a 2-year acquisition period.
Instead, as funds became available from diverse stakeholders,
funding was received in piecemeal fashion for smaller blocks
each year between 2010 and 2019 (see Figure 3) for annual
project areas. By the time funds were available in 2019 for
mapping of Kodiak Island which we had planned to map with
GeoSAR (because of the rugged terrain and dense forests),
GeoSAR had already been retired from service by Fugro and
was no longer operating commercially. Intermap did the best
they could to map Kodiak Island with X-band IfSAR only.
Because funds were received so erratically in small, inefficient
acquisition blocks, members of the Dewberry team acquired
much data on speculation, taking financial risks and not
knowing whether they would ever be paid.
Figure 4. Intermap efficiently mapped all Alaska 1-degree cells shown in pink,
and Fugro efficiently mapped all cells shown in blue.
Figure 5. JOA Surveys efficiently surveyed 1,044 QA/QC checkpoints in
remote terrain within each of the cells mapped by Intermap and Fugro.
“
“
What challenges did you face once
you received funded task orders?
