coordinates in
MCPD
. Control points described as network
accuracy were chosen for this study as network accuracy
represents uncertainty of control points with respect to the
geodetic datum. Depending on the method used for data col-
lection such as static
GPS
or Real-Time Kinematic (
RTK
)
GPS
,
the accuracy (1
σ
) varies from
±
0.002 m to
±
0.05 m in both
horizontal and vertical data.
Studies relating to regional velocities provides good
evidence of whether the position of a control point should
remain constant or show a pattern of movement coincident
with overall regional velocity changes. For example, the
average difference between
NAD 83
(2011) epoch 2010.0 and
NAD
83(
CORS96
) epoch 2002.0 for easting was 0.05 cm
±
5.25
cm, for northing 2.12 cm
±
6.08 cm with vertical differences
of -0.66 cm
±
2.24 cm, whereas the average difference of
those two adjustments at the same epoch 2002.0 for easting
was −0.14 cm
±
1.04 cm, for northing 0.19 cm
±
0.94 cm with
vertical differences of 0.80 cm
±
1.89 cm (
CORS
Myear FAQ 6,
NGS
). It is apparent the difference between two adjustments
at the same epoch is very small, but the difference is larger
between epochs of 2010.0 and 2002.0 because of the impact of
velocity over time. Analyzing velocity of both the horizontal
and vertical components of existing
CORS
sites allows for the
subsequent analysis of
MCPD
control points in close proximity
to a
CORS
site to determine if these control points exhibit the
same positional behavior over time.
To accomplish this analysis, data in the
MCPD
were parsed
by surveyor. Next, data overlaps (spatially coincident control
points) between surveyors were identified using the intersect
tool in ArcMap
®
. These data overlaps were further analyzed
based on the township, range, direction, and section to verify
they were indeed representing the same control point. The
datasets used for all further analyses are described in Table 1.
Analysis involved the exploration of data derived from
differences in northing, easting, and orthometric height. The
objective of data exploration was to statistically analyze dif-
ferences between two datasets describing the same control
point. Data distribution patterns were analyzed using a his-
togram and box-plot. Normal probability plots were drawn to
check the normality of the data and
t
-tests performed between
the datasets. Both parametric and non-parametric
t
-tests were
used depending on resulting normality. Parametric
t
-tests
were performed under the null hypothesis, that
H
0
:
μ
Difference
= 0, against the alternate hypothesis,
H
1
:
μ
Difference
≠
0
Thus, the test of observed
t
-test statistic was
t
S n
Difference
=
−
µ
0
/
; where,
μ
Difference
is the mean value of the
difference,
S
denotes the standard deviation of the differenc-
es, and
n
indicates the number of observed control points in
the analysis. The observed
t
-statistic was compared with the
Table 1. Summary of control points statistical analyses completed in this study.
Sample Size
Survey Periods
Mean (m)
Standard Deviation (m)
t-statistic
Differences between horizontal coordinates (2013 and 2015)
Easting
82
Mar, 2013 vs Aug, 2015
0.000
0.023
0.18
1
Northing
0.000
0.042
-
Horizontal
0.008
0.039
0.08
1
Differences between vertical coordinates (2013 minus 2015)
Using GEOID03
52
Mar, 2013 vs Aug, 2015
0.072
0.052
-
Differences between NAD 83(1986) and NAD 83(2011)
Easting
85
Jun, 1997 vs Oct, 2015
-0.980
0.021
428.65
1
Northing
0.680
0.011
551.55
1
Horizontal
1.200
0.018
598.26
1
Differences between NAD 83(1986) and NAD 83(NSRS2007)
Easting
21
Aug, 2015 vs Dec, 2015
0.038
0.027
6.43
2
Northing
-0.037
0.030
5.74
2
Horizontal
0.062
0.025
10.94
2
Differences between NAD 83(CORS96) and NAD 83(2011)
Easting
16
Dec, 2012 vs Oct, 2015
0.009
0.011
1.90
3
Northing
0.014
0.021
4.16
3
Horizontal
0.022
0.017
4.84
3
Vertical coordinate differences using geoid models
GEOID99 vs GEOID12A
12
Dec, 2012 vs Sep, 2016
0.616
0.029
54.73
4
GEOID03 vs GEOID12A
8
Oct, 2012 vs Jul, 2014
-0.304
0.237
3.63
5
Differences in position due to velocity
NAD 83(2011) 2010.0 vs
NAD 83(CORS96) 2002.0
Easting
12
Sep, 2016 vs Dec, 2012
0.972
0.017
-
Northing
-0.693
0.019
Easting
9
Oct, 2015 vs Dec, 2012
0.977
0.020
-
Northing
-0.703
0.057
Easting
16
Oct, 2015 vs Dec, 2012
0.008
0.018
-
Northing
0.014
0.013
1.
t
-critical (0.05) value = 1.99
2.
t
-critical (0.05) value = 2.09
3.
t
-critical (0.05) value = 2.13
4.
t
-critical (0.05) value = 2.20
5.
t
-critical (0.05) value = 2.37
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
April 2018
217
