Jökull - 01.12.1994, Side 25

RESULTS AND DISCUSSION
The L5 ambiguity has a half-wavelength of about
43 cm, and should therefore be easily resolvable. Re-
solving this ambiguity does not in itself improve the
result but it is an important test of quality and a neces-
sary prerequisite to calculating the L1 and L2 ambi-
guities. Except for the long lines connecting Núpafjall
and the regional points, we could resolve all the L5
ambiguities in our data. Resolution of the L1 and L2
ambiguities is not very important where observation
windows are several hours long, since the ambiguity-
free and ambiguity-fixed solutions then approach
one-another. Which solution is superior then must be
determined by examining repeatability where points
were measured multiple times and ambiguity-fixed
solutions are often marginally better. For our data, all
the L1 and L2 ambiguities were calculable for the
short lines and 60% for the long lines involving the
regional points.
Network adjustments were performed on the am-
biguity-fixed results for (i) the Hengill network only,
and (ii) the entire network. The normalised RMS’s
calculated were 5.9 and 7.4 respectively. The scaled
formal errors are fairly homogeneous and sub-cen-
timetre at lo for both the N and E components, and 2-
3 times as large in the vertical, i.e., about 1 cm (Table
2, Fig. 4). The horizontal scaled formal errors are no-
ticeably larger for points DU18 and DU26, since
these points were measured once only. Within the
Hengill network, 91% of the points had lo errors of
less than 5 mm in the north and east components com-
pared to 85% when the regional points were included.
The vertical components have larger errors in both so-
lutions but the errors are smaller in the Hengill-only
solution with 91% being less than 1 cm compared to
70% when the regional points are included. From this
it is clear that the ambiguity-fixed, Hengill-only re-
sults are the best for the local network. The larger er-
rors associated with the four long lines are a conse-
quence of the line lengths being up to 12 times those
of lines within the Hengill network. The geographic
Table 2
Station Name Hengill Area dn de (cm) du Entire Survey dn de du (cm)
OS-1991-DUOl 0.27 0.25 0.69 0.34 0.31 0.86
OS-1991-DU02 0.28 0.24 0.70 0.36 0.30 0.88
OS-1991-DU03 0.22 0.19 0.50 0.27 0.23 0.62
OS-1991-DU04 0.28 0.24 0.69 0.35 0.30 0.86
OS-1983-7134 (Nupafjalt) 0.00 0.00 0.00 0.00 0.00 0.00
OS-1991-DU06 0.26 0.22 0.65 0.32 0.27 0.82
OS-1991-DU07 0.34 0.27 0.72 0.42 0.33 0.90
OS-1991-DU08 0.33 0.30 0.88 0.42 0.38 1.10
OS-1991-DU16 0.28 0.27 0.70 0.35 0.34 0.87
OS-1991-DU17 0.22 0.19 0.57 0.28 0.24 0.71
OS-1991-DU18 0.66 0.64 1.71 0.82 0.80 2.14
NE-79-037 0.34 0.31 0.88 0.43 0.39 1.11
OS-1991-DU20 0.24 0.20 0.58 0.30 0.25 0.73
OS-1991-DU21 0.25 0.20 0.60 0.31 0.25 0.75
OS-1991-DU22 0.24 0.20 0.60 0.30 0.25 0.75
OS-1985-7331 0.28 0.27 0.70 0.35 0.34 0.87
OS-1991-DU25 0.32 0.29 0.86 0.40 0.37 1.08
OS-1991-DU26 0.69 0.65 1.82 0.87 0.81 2.27
OS-1991-DU27 0.32 0.25 0.74 0.40 0.31 0.93
OS-1991-DU30 0.31 0.24 0.72 0.39 0.30 0.90
OS-1991-DU15 0.31 0.24 0.73 0.38 0.30 0.91
OS-1991-DU33 0.40 0.33 0.94 0.50 0.41 1.18
OS-7478-S027 0.32 0.26 0.78
OS-7193-S001 0.31 0.26 0.75
OS-7481-S061 0.67 0.56 1.57
OS-7480-S056 0.69 0.59 1.63
Table 2 Results of network ad-
justments for points within the
Hengill area only, and for the
entire survey. dn, de, du (shown
for both sets of results) are the
scaled formal errors in the
north, east and up components
respectively.
2. tafla. Niðurstöður útreikn-
inga á óvissu í staðsetningum
mœlipunkta á Hengilssvœðinu
(fremri dálkar) og effjórirfjar-
lœgir mœlipunktar eru teknir
með (aftari dálkar). dn, de og
du eiga við ónákvœmni mœling-
anna í norður-suður, austur-
vestur, og lóðrétta stefnu.
JÖKULL, No. 44
23