0043 15.7 1.0768 0.00 0.00 0.3.0093 13.2 1.93 7.13 .35 13.2.00 0..85 TVALUE FOR A 5. 4 .0000 0.2262 0.0007 0.33 0..7 0.82 16.8 61.3 3.05 22.00 0.5313 0.0 0.16 0.2215 0.32 24.2 3.2545 0.0000 0. 8 FIVE TEMPERATURE SENSORS.47 1.2590 0.8 0.2 3.03 0.5 1.83 1.00 .19 0. 1. 3..84 24.0005 24.0820 0.20 0.0010 3..79 0.0709 0.73 0.7 3.5846 0.0043 14.2 3.03 0.49 1. 0.0000 0.7 129.4 386.44 58.5 2.00 0. OF BEING INSIG.6 0.0000 0.68 0.2 3.3 12.00 0.29 0.35 0.2 3.8 3.0 METRES.0012 1.2 3.0000 0.3 53.0008 17.7 3.6 2.0012 20.4 4.0 1.14 0.2 3.6889 0. 10 ..0352 0.2 3.74 20.09 17.0022 20.2 9.9 182.1 524.0004 23..2 3. 1.2 3.2 5.1 787..3846 0. 12 .88 14.2057 0.0004 0.00 0.2 3.0 AND 3.92 25.08 18.0004 12.1 37.2 3.2 3.1 2.0967 0. 2.0021 0.02 c 0.57 0.3.70 3.09 0.oooo 0.00 0.0183 11 .2 3.17 0.G 1.90 16..7 1.060 • FIVE TEMPERATURE SENSORS.49 0. #1 in Table 5.69 0.00 0.69 19.00 0..2.6 5.2 3.0795 0. 5 .4103 0. 2 .6 2.9 42.8 1.13 0.8 0.2 3.2720 0.2 3.9 532.0000 0.9 711..9 215.9 39. 1 95% TEST TABLE VALUE 3.23 0.0000 0.2 3.0 METRES.0069 0.2716 0.0006 0.68 ••••• 32. SENSORS ARE AT HEIGHTS 0.0010 16.0433 0.01 0.2257 0.2 3.6.5 136.00 0.9 1.52 0.2 THE MEAN STANDARD DEVIATION IS : F F F F F F F p p F F p F F F F F F F F F F F F F F F F p PRO.2 3.4940 0.2424 0..0283 0.0982 0.7 22.4903 0.18 18.1435 0.0000 0.6793 0.5800 0.86 31.35 0.00 0.0023 19.37 B 2.0009 0.7 2.0 METRES • .4 0.0048 16.0225 0.0 AND 4.16 24.1922 0.94 0.0967 0.15 2.6.0.9540 0. SIX II 3 4 10 3 8 7 7 11 5 5 16 10 20 13 7 5 2 3 4 11 25 30 13 10 4 5 12 9 21 (X) 0 .4 16. 3.3734 0.1 1.0000 0.0000 0. 1. SENSORS ARE AT HEIGHTS 0.00 0.0002 0.0196 0.8 1.1 334.38 1.0124 10.3369 0.38 15.4 2. 8 .8 B c 0. 2.0 10.40 3.60 0.0 2.0 3.6 0.00 0.1931 0.00 0.5 2.96 5. 7 .TABLE 5.2 3.3 96.0857 0.54 16.6979 0.4249 0.66 0.0076 15.3 1. 11 .2296 0.0137 18.4 22.00 0.55 2.91 0. TEMPERATURE SENSORS.1761 0.5827 0.0024 9. c B A 0.12 0.51 1.2.05 0. 1 .16 •17 •18 •19 t20 @21 622 823 ~24 825 •26 •27 t28 t29 TIME VARNC FACTR 10.2 0.50 12.0302 12.7 0.5 2..2 3.3 1.1 1.3198 0. 0.0006 8.0 AND 4.5923 0.2 0.00 0.0000 0.3 1.09 0.5 0.7 1.0945 0.2 3. 0.2313 0.00 0.0.00 0.1027 0.0907 0.00 o..0 3.3 1..2784 o..3 1.07 19.0878 0.3182 0.0876 0.4 1. 6 .2 3.14 .0044 13.2 3.5 4.3 50.1104 0.0 3.01 0.46 0..0119 0.00 0.2 2.0028 22.2 3.33 1.76 18.5054 0.0023 0.28 0. 1 1.0 4. 9 .54 0.59 6.00 0.0023 A 17.60 24.6 0.42 0..2 1.00 0.84 16.43 1.00 0.0002 0.0086 11. 00 0.09 0.7 Curve fitting with test of the significance of coefficient (Kukkamaki's model.0018 0.62 0.2 3.3.0108 0.1 3.00 0.0020 18.2994 0.4 l.14 19. • .0031 17.00 0.4945 0.00 0.oooa 7.22 4.9 33.1 0.2 3. 3 . SENSORS ARE AT HEIGHTS 0.8878 0.8975 0.03 0.33 21.0010 19.7 1. 2.8 0.15 0.65 15.0035 0.54 25.3444 0.4).49 3.5 0.7 10.0000 0.38 1.15 .0000 0.0753 0.
14 4..88 20.2 3.3761 0.3 126.5 3.0028 0.2 3.0000 0.59 22.3.22 2.04 0. .2 3.3.0147 0.0 1. 3.75 0.0003 0.4 67. TEMPERATURE SENSORS.5 147 . 2.5 5. SENSORS ARE AT HEIGHTS 0.65 14.6.0114 0.73 1.0015 0.2 521. 14 .7 65.0002 0.0216 0.2 3.0002 0. .18 0.02 1.58 24.3 82.0000 0.9 10.0043 0.0079 0.49 15.40 20.8 31.91 1...49 1.0 AND 4.2 THE MEAN STANDARD DEVIATION IS : F F F F F F F F F p p p p p F F F F F F F F F F F F F F p p p p p p p p p p PRO.01 0.51 12.0647 0.2 3.a 0. • FIVE TEMPERATURE SENSORS.0451 0.0018 0. SENSORS ARE AT HEIGHTS 0. .0016 0.0019 0.4 279.00 11.03 24.30 0.0000 o.09 0.0002 0.8 1.6 161.054 1.0163 0.3 48.08 2.17 0.20 1.9 1.0007 0.08 2.0917 0.86 2.2 9.2 3.04 0.1 1.0000 0.00 11.10 1.2 3.0023 0.0679 0.3704 0.00 3.7 3. c B A o.oooo o.0010 0.0004 0.19 1.0054 0.6 6.00 5.2.81 24.0012 0.00 6.0000 0.0000 0.00 4.0049 0.0000 0.46 19.00 19.63 19. TIME t .4 9.0006 0.2 3.7368 0. 2.4 4.37 3.2 3.37 3.4 0.2 3.0253 0.0001 0.00 23.0024 0.00 16. 12 .2671 0.1074 0.2 61.0010 0.OOOO 0.00 21.0003 0.00 12.00 24.0000 0.0057 0.01 0.00 22.0014 0.1 9..87 0.4 4..2 3.1 90.00 21..1 1. .2 3.2 9.0016 0.3 4.0 94.2 3.9 7.9 4.18 18.00 17.5 48.0000 0.0230 0.3 0.3 2.0000 0.15 0.0456 0.0.00 15.4 4.2319 0.0000 0.1 1.02 0..2 3.0000 0.00 1.0046 0.06 0.10 0.0714 0..5 4.0 10.29 20.3 0.0143 0.00 23.7 2.24 21.2 3.0000 0.04 0.33 0.4 30.8 17..25 0.7 1.3 266.0000 0. 0.16 25.0556 0.0 3.TABLE 5.7 4.9 146.6 235.2 3.2 3.1 0.16 "'17 "'18 .0122 0.35 c 0.4 2.9 1.0126 0.0031 0.2 3.oooo 0.86 14.21 13.07 0.50 0.6 5.7 5.2 1.4 190.8 of coefficient test Curve fitting with the significance (model #3 in Table 5.0061 0.50 1.0422 0.1047 0.84 ]3.0024 0...2.5 2.00 16.8 191.6 2.2 3.09 0.9 0.8 7.1 128.0000 0.00 10.0 AND 3.0383 0.1 95!1.2 3.0019 0.4).23 0..0000 0. 1.08 0.as •36 •31 •38 10.05 0.16 0.00 18.0001 0.2 3.0128 0.40 0.9815 0.5514 0.4 432.44 0.0208 0.1 53.0005 0.26 0.1 5.oooo o.97 16.1479 0.2 3.0436 0.8 6. 19 .6 154. SIX It 3 3 3 3 3 3 3 3 3 3 3 3 3 2 3 3 3 2 3 3 3 3 3 2 3 3 3 3 3 2 3 3 2 3 2 3 3 3 00 .50 1..00 19.00 15.36 12.2 3.2 3.4298 0.0102 0.5 2.2281 0.0014 0.0107 0.2 3.0434 0.05 0.7 2.4 155.0480 0.1 3.3 9.07 0.6.0000 0.0156 0.0000 0.4 65.0011 o..09 0..00 22.53 0.2528 0.03 0.0226 0.20 0.2 3.0000 0. 20 "'21 .3 0.05 TVALUE FOR c B A  2.00 12.2 3.89 0.8 ••••• 150.4 350.1028 0.9 20.2 3.0.0000 0.16 0.35 0.45 20.71 5.0000 o.85 20.00 13.88 21.7 31.0544 0.07 0.. .64 0.0007 0..00 20.1596 0.7 18.0165 0.2 3. 19 0.48 18.0091 0.07 0.3 45. 0.8 10. 1.2 3.09 0.oooo 0.10 0. e FIVE TEMPERATURE SENSORS.2 3. 1 3..0439 0. 37 1.11 0.to "'11 .0110 0.3783 0.00 VARNC FACTR 0.2 3.76 20.97 3..6943 0. 2. 1 4.9 190.0002 A 18.5040 0.00 2.30 18.00 9. OF BEING INSIO.0257 0.92 1.0003 0.2 10.5 3.07 15. TEST TABLE VALUE 3.3. SENSORS ARE AT HEIGHTS 0.0023 0.0009 0.2 3.1 0.2 3.0 143.12 19.00 14.00 14.oooo 0.2 3.17 0...oooo 0..0 56.. . 22 •23 •24 •25 •26 827 828 629 830 831 •32 •33 •34 .0010 0.0075 0.21 0.22 19.0263 0.3189 0.2 3.0935 0.49 19.89 20. 1 2 3 4 5 6 7 8 9 . 15 .75 0.6 44.0008 0.09 19.0 METRES.9288 0.9 6.4 4..0007 0.49 18.7 5.0006 0.8 26.0 METRES.5 3.5 3. .2 77.00 17. 3.00 13.08 0.02 0.7 0.09 B 0.0002 0.0020 0.4236 0.00 20.28 0.00 18.2.0001 0. .90 2.0179 0.0 METRES • .0084 0.82 0.02 0.2 3.0096 0.0 AND 4.4407 0..43 17.21 0.0467 0.20 16.7 1.9 80.0186 0.0510 0. 13 .0000 G.7 95.07 0. 0071 0.2 3.0147 0.4 163.9 2.0642 0.8 222.5 1.9 2.91 15.0230 0.5 4.0007 0.0707 0.0000 0.0012 0. .07 23.00 8.0000 0.2 219.6537 0.0 39.0000 0.0000 0.00 7.
.JP c c r!i' c: ~c c e 18:00 G:e 18:00 a . c c c e c: 8 c:.4: c . c== e 7 18:00 c c 08:00 e c " ac <. ..........:.82 ..... c: c: c: e c: CB . c "'a "'9% c "' a c: e c:: c a 0.20 0. c c 0.10 0.. c 08:00 i c CA c 18:00 .... . . a c: E c: c: c 8 ... c "' " c:: ea Ce . • c: c: • e"' c: 18:00 c:  c a c: "' c: a '!!I . Test of the significance of coefficient for models in Table 5.. 4 c: .. . a:: c .. S' cC ·:: c..c:: Cl) I!! e c c 08:00 c c: a"' ....30 .. a c: c: F" 1!!1 c e 08:00 a II a c c:EI . c c B c "' c e E . . ... 6 .e . a B 18:00 c:: . •c c .e c c c !:! c c c:B c c c c c c: c: c c 0 5 N A 08:00 c .. c: c:: ..25 Probability of Being Insignificant Figure 5. c .4 0.. model ca C:& a 08:00 eeoc 1 c: ce e . ....... c c:: c ....15 0.. a a a <c .00 c c c E c:: .. c j=_cc .. c: c c a c: Cl) ~ fill a a 3 .05 0. ad!' c e c:ac c.
5. and iS produce not it very close can be results in spite of their different appearance. For results (all 35 developed least in order intervals for to compute squares estimation c of For model converge.3) to UNB trigonometric measured method on between each pair of the three bench marks at the SouthGym area. then c software was the based on either the coefficients 0.7 show the spline fitted plots and the models i4 the a and b were estimated by least i7.3 m.6 m and 1.10 and 5.6 and From these seen that 5. = 1. corresponding tables.discrepancies between effect versus the geodetic levelling and the UNBmethod).9.2 m). the results results were used to compute the section using the ~ Tables 5. full range using equation (3. whenever the was assummed solution did and the other coefficients were estimated through least squares.5.4. squares. model i1. Models i1 and i3 also give i6 gives very similar results. Model some discrepancies with the measured refraction effect. or (using the 0. a fixed value The other two coefficients.11 show computed refraction <.5 of comparison).4 field verification 2f The seven models in Table 5. 1985.83 ~ Comparison 5. large The same . 1/3 for exponent c.3) observed temperatures at heights assumption of have the the correction computation of exponent or the to two Figures of these results.2. 5.4 refraction correction (see height differences July 23 and 24. of the measured values 4. 5.
12 c) seen from the for that correlation does not exist corresponding Figure will be discussed the correlation coefticient factor) the alternative hypothesis later. The t statistic with In (n2) correlation The value 0.12 shows the The three cases. coefficient among correlation coefficient matrices high values of the results of expected (except for model 17.44 that can degrees of pass the freedom is given by e.g. the correlation are as the seven models asphaltgravel line).44. the correlation coefficients between the measured refraction effect and the individual computed values using the seven models are generally significant. the third case (Table as can also be The reason 5. first two matrices.82 which is greater than t value from corresponding table of percentage points: . The In the highest correlation exist between models 14 and 15.44. Hamilton [1964] as t = I 2 (n  2) r 2 1 . using = 2. is the smallest correlation coefficient test. the above equation: t For r = 0. for refraction effect similar to other models.84 Model i2 also results in values can be seen for model i7.7 • At the 1% hypothesis that r = 0 level of significance the null that r (r is can be rejected against ~ 0 if the coefficient is equal to or larger than 0. Table for the 5. 5.r where n is the number of observations.
85
t = 2.82
For
r = 0.43,
value.
t
>
t
33,0.005
= 2.75
= 2.74 which is smaller
than the
table
In total, the r values for models fl, 43, f4 and f5
are larger than
0.44 in the two first cases,
Table 5.12 a
and b, and for models f2, f6 and f7 are smaller than 0.44 in
the first case, Table 5.12 a.
The above discussion shows that the new proposed models
can safely replace
the Kukkamaki model when
temperature sensors is equal to or
number
choice.
of sensors
is 2,
then the
the number of
greater than 3.
model f2
is the
If the
only
To draw a firm conclusion about which model is the
best reperesentative of the temperature profile in the lower
atmosphere up to
4 m height and to confirm
the validity of
the new models, more investigations are needed.
86
TABLE 5.9
Refraction effect [mm] computed using the seven models
versus the measured value (BM1BM2).
__ ZONE
TIME
NO (ADT)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
 29
30
31
32
33
34
35
36
37
38
M0 D E L
i1
i2
i3
i4
i5
i6
i7
meas
10:00 2.0 1.0 1.7 1.7 1.8 1.9 2.6 o.o
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00
20:00
21:00
22:00
23:00
24:00
01:00
02:00
03:00
04:00
05:00
06:00
07:00
08:00
09:00
10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00
20:00
21:00
22:00
23:00
5.3
1.9
1.8
0.3
0.1
0.6
2.2
1.3
2.3
2.7
0.5
1.7
3.7
2.4
2.3
2.3
3.9
3.6
1.7
2.0
0.5
0.6
0.9
2.0
1.4
2.1
3.3
0.1
4.0
5.8
3.9
4.3
2.4
6.0
5.6
2.5
1.1
1.4
0.7
0.8
2.2
1.2
2.1
2.8
1.3
1.8
3.7
2.4
2.1
2.2
4.0
4.3
1.5
1.2
0.2
0.2
0.7
2.1
1.5
2.4
2.9
0.3
3.1
4.9
3.1
2.9
2.3
5.2
4.9
1.8
1.0
0.3
0.0
0.6
2.2
1.7
2.4
2.9
0.5
3.1
4.9
3.0
2.8
2.4
5.2
5.9
2.2
2.9
1.1
0.2
1.6
3.6
2.2
3.5
5.1
0.0
6.0
8.8
5.9
6.2
3.7
9.2
5.9
0.6
0.5
3.0
5.2
0.4
2.5
2.2
3.0
2.7
0.1
0.8
2.8
1.8
2.7
2.5
3.0
0.7
1.7
1.4
1.5
1.9
2.1
2.2
2.4
1.8
3.4
2.6
2.0
1.8
1.9
2.0
1.7
1.8
2.6
0.8
2.6
1.5
1.1
1.4
2.3
3.4 2.8 3.0 2.9 4.1 2.0 1.2
0.3 0.5 1.1 1.2 1.5 1.4 0.5
0.4 2.2 1.6 2.1 1.6 1.4 0.3
1.5 2.2 1.0 1.3 1.4 1.2 0.1
0.1 0.1 1.4 1.2 0.3 0.3 0.5
0.0 2.3 0.6 1.1 0.4 0.7 0.5
1.1 3.4 1.5 2.0 1.5 6.4 0.7
0.5
0.1
0.2
0.7
2.7
0.5
2.9
3.2
2.1
1.3
0.7
0.7
1.2
3.1
3.4
4.0
2.5
2.6
Mean
0.8 1.5 0.7 0.9 1.0 1.9 0.6 1.7
3.4 3.9 3.1 3.5 3.6 5.9 4.0 2.4
0.7
0.2
0.2
0.8
2.6
1.9
3.0
1.8
2.3
0.1
0.2
0.6
0.0
2.4
2.1
2.6
0.9
1.8
0.0
0.0
0.7
0.0
2.3
1.9
2.6
1.0
1.8
1.4
1.1
1.1
0.1
3.0
3.1
3.9
1.6
2.6
4.7
0.5
0.1
0.2
2.4
1.1
4.8
3.1
3.1
2.1
1.2
1.8
2.5
2.5
2.5
2.5
1.7
1.7

87
TABLE 5.10
Refraction effect [rnm] computed usino the seven models
versus the measured value (BM2BM3).
ZONE
TIME
NO (ADT)
M0 D E L
#1
#2
#3
#4
#5
#6
#7
meas
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
9:00
10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00
20:00
21:00
22:00
23:00
24:00
01:00
02:00
03:00
04:00
05:00
06:00
07:00
08:00
09:00
10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00
20:00
21:00
22:00
Mean
1.6
3.2
3.1
1.5
3.0
6.1
0.3
2.7
1.1
1.2
3.4
0.1
2.1
3.5
2.2
2.1
0.3
2.9
1.9
0.5
0.8
4.3
0.5
1.4
1.7
1.9
3.4
0.7
3.7
5.3
3.7
3.3
1.5
3.9
3.1
0.4
3.5
6.2
0.2
2.3
1.0
1.2
3.4
2.0
2.1
2.9
2.1
2.1
1.0
2.4
2.2
0.9
2.2
4.8
0.8
2.6
1.8
1.3
3.5
0.6
3.7
3.9
3.0
2.9
1.1
3.0
2.7
0.7
3.0
5.6
0.5
2.6
1.9
1.2
3.6
0.8
3.8
3.8
2.9
3.0
0.6
4.2
2.8
1.1
1.8
6.8
1.1
3.1
3.0
2.4
5.4
0.9
6.3
7.5
5.6
5.2
1.5
6.4
1.0
0.4
0.7
6.8
0.0
1.1
1.8
1.1
3.4
0.1
1.7
2.7
1.8
2.0
1.4
1.4
0.2
0.5
0.6
0.6
0.3
1.8
1.2
0.0
1.1
1.7
1.7
1.9
1.9
1.1
3.2
2.6
2.8
5.3
4.4
3.0
3.2
2.7
2.8
4.2
3.6
3.1
4.3
3.6
3.3
7.9
6.7
4.9
2.5
2.6
3.4
2.0
1.6
1.0
1.3
2.7
1.4
2.0
1.8
3.8
0.5
0.4
0.2
1.3
2.1
2.3
3.2
5.2
0.9
0.7
0.8
0.7
2.6
0.4
3.8
0.4
3.0
0.0
1.3
1.6
2.1
2.2
3.1
2.9
0.5
0.8
1.1
2.8
3.4
3.7
1.6
2.7
0.2
1.6
2.5
4.9
4.0
6.5
1.1
0.4
0.8
0.8
2.5
1.7
3.7
1.3
2.6
0.5
0.7
1.4
0.1
2.2
3.8
1.2
0.8
0.4
0.0
2.6
1.5
3.8
0.1
1.5
0.2
1.2
1.9
1.8
2.8
4.7
1.3
0.6
0.6
0.1
2.4
1.4
3.9
0.5
1.8
0.9
2.4
2.5
3.1
3.5
5.3
3.9
0.1
0.2
0.5
2.9
2.2
5.5
1.0
2.2
0.4
1.9
0.4
0.2
4.0
5.9
6.5
3.8
1.9
0.3
2.3
1.3
4.2
0.4
3.2
1.0
2.0
1.6
1.5
0.5
2.0
0.5
1.7
0.2
0.4
0.3
1.5
2.3
2.6
1.7
0.3
0.8
0.2
0.7
0.5
1.0 0.1
      
0.3
   
8 1.0 0.9 1.7 0. 1 1.1 0.9 0. 9 0.5 0.0 0.2 .2 0.6 0.9 0.8 1.5 1.8 1.1 0. 2.2 1.2 0.9 2.8 0.2 0.5 0.3 .2 1.2 0.3 0.4 0.2 0.6 0.5 2.5 1.0 0.0 0.1 0.9 0.9 1.3 1.5 1.0 0.5 1.8 1.6 0.8 0.1 0.2 1.7 1.1 0.2 0.1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 9:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00 19:00 20:00 21:00 22:00 23:00 24:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00 19:00 20:00 21:00 22:00 Mean 0.0 0.1 4.0 1.6 0.9 0.0 0.1 2. 2 1.11 Refraction effect [mm] computed using the seven models versus the measured value (BM3BM1).9 0.0 0. 7 0.1 6.3 2.2 2.0 1.0 1.5 2.8 o.5 0.3 1.1 0.0 0.2 1.4 0.2 0.9 0.2 0.9 0.4 0.3 1.2 1. 0 1. 5 1.2 me as 0.2 0.2 0..6 0.9 0.6 1.0 0.1 0.9 0.1 0. 3 0.1 0.1 0. 0.0 0.6 2.1 0.9 0.4 0.4 0.2 0.1 1.6 0.9 0.4 0.1 0..1 0.7 0.2 1. 1 0.0 1. 6 0.9 0.2 1. 9 0.0 0.3 0.2 1.1 0.5 2.5 0.2 0.8 1.2 0.3 1. 3 2.2 0.0 0.1 0.3 0.6 1.0 2.3 0.1 0. ZONE TIME NO (ADT) M 0 DEL i1 i2 i3 +4 45 +6 i7 ..8 1. 0 0. 7 2.8 0.4 2.8 2.3 0. 2 1.7 0.0 0.5 0.2 0.3 1. 6 0.5 0.9 1.9 0.6 0.3 0.2 1.6 0.1 0.2 1.2 0..4 1.5 2.7 0. 9 2.7 0. 3 2.6 0. 8 2. 1 0.0 0.2 1.3 0.4 0.4 2. .9 2.6 0.0 2.1 1.2 0.0 2.0 2.0 0.0 0.1 1.1 1.4 0.1 0.1 2.0 3.2 0.0 0.7 0.2 0. 1 1.7 0.1 0.6 1.3 0.9 1.8 0.0 0.1 1.1 0.4 0.1 0.9 2.0 0.1 0.2 1.88 TABLE 5. 3 2. 3 0.3 0.9 3.3 0.1 0.5 0.3 0.2 2.2 2.6 1.5 0.2 0.1 0.3 0.9 2.3 0.8 1.5 0. 1 1.3 5.5 0.3 0.4 1.1 2.3 2.0 1.9 0.9 0.3 1.2 1.7 0.9 0. 8 1.2 0.4 0.3 5.8 2.4 1.1 1.2 0.0 0.3 2.3 0.8 1.3 0..4 0.8 2. 8 0.0 2.0 0.9 0.5 0.2 0.5 0.6 0.1 0.7 0.9 0.5.2 1..1 0.5 0.7 1.3 0.1 0.4 0.2 0.0 0.0 0.5 0.8 2.5 0. 4 1.4 0..8 1.7 5.0 0.2 0.
86 0.59 1 0.97 0.48 1 0.58 1 1 .97 0.94 0.34 .56 1 0. 1 1 a: gravelgrass (BM1BM2) line model #1 model #2 model #3 model #4 model #5 model #6 model #7 measured 1 0.94 0.82 0.94 0.25 .99 0.95 0.75 0.12 Correlation Coefficients Matrices model #1 model #2 model #3 model #4 model #5 model #6 ·model #7 measured #1 #2 1 0.92 0.79 0.10 1 0.95 0.49 #6 #7 0.92 0.82 0.96 0.92 0.94 0.98 0.92 0.93 0.97 0.36 1 1 0.74 0.81 0.77 0.03 1 measur.50 1 0.94 0.91 0.02 c: asphaltgravel (BM3BM1) line .68 0.89 0·.89 0.90 0.93 0.42 #3 MODEL #4 #5 0.84 0.78 0.67 b: grassasphalt (BM2BM3) line model #1 model #2 model #3 model #4 model #5 model #6 model #7 measured 1 0.96 0.97 0. 79 0.10 1 0.97 0.18 .56 1 1.38 .97 0.87 0.95 0.89 TABLE 5.68 0.52 1 0.91 0.89 0.05 1 0.75 0.79 0.99 0.89 0.94 0.00 0.96 0.82 0.89 0.77 0.10 .74 0.95 0.02 1 0.88 0.41 1 0.60 1 0.04 1 0.71 1 0.38 .
~ \: 2 0 ·n .' ..5: t1odel #4 ···· Model · . #3 #7 (h] ...._I /' I ../ 11 4 ~ 4l Q) 0:: . I .. .. /... .. ... I ' 1 2 ~ \ I ./ I '..0 0 \ '.. ' ~ o... \.Model #I #5 12:00 .Model Model Refraction effect computed using the seven models versus the measured value for BM1BM2. \ \ t!l 41 41 I J 1... ~ ~ ...'.~..jJ 0 () Q) ·.. .... \_/ I' 6 I I I .jJ C) ctl I ~~~\~ \ i ..6 I I' I 1 I I 1 I I I 4 . ..I I I \ I ·.. .1 II .·. \ ' ·./' o 0 12:00 LEGEND• 04:00 20:00 oEla Me115ured Figure 5.. .Model · Model #2 #o 2o:oo Zone Time ..
. ..1l 1.. I fr... \ ' ' \ 'l!l. cO ~ ~ (!) ~ 4 I I ·6:J I \/ 6~~~~~~~~~~~ 0 20:00 12:00 04:00 12:00 20:00 Zone Time [h) LEGEND• oEJ<3 Me~5ured Model Figure 5.... f.. '" 1 0 ' 1. __ .. \ \ "' h .6 ..1l J.. ' ('.: ''.6: #4 Model Model #l #5 Model 11odel #2 #6 Model Model #3 #7 Refraction effect computed using the seven models versus the measured value for BM2BM3. \• • 0 ..... \ ~' .."' ·. \_../ I• \ I/.'" //. A.. '' I I~J ·"·' I: ~.. a.\ I i X ~' ). / ~ .... ......
...i 10 •\ '\..' ~ / ''T:J .l ~ \ li'r:l ###BOT_TEXT### N \~~... +" I . .· .1 I ~ ril P I .4 ''\ G:l\~ /j':~' : I '.I 2 " t ~ Q) ex. .... #3 #7 ..Model ····· Model #2 #6 20:00 Zone Time [h) Hodel Model Refraction effect computed using the seven models versus the measured value for BM3BM1.Model #l #5 ... >! ~.Jill v. • •..:.\ 'I 'f/1•I ''.. •• m I 1 ' 1 1 I I \ I I
8 2.6 2. [mm]  NO mean of 2 tar LOCAL TIME 2.lm tar 3.6 1. the between Table 5. windy Sunny periods.5m tar 10:25 10:45 11:02 11:16 11:32 11:54 12:12 12:30 13:01 13:15 13:47 14:00 14:13 0. Only the height difference between repeatedly for four shows the BMl and BM2 was measured hours from 10:20 to discrepancies 14:30. .2 3. windy Sunny periods.5. 72 1.9 1.8 0.6 1.2 0.01 1. TABLE 5.57 1.93 5.9 0.9 2.5 4.6 0.2 1.25 1.2 0.9 0.4 1.31 REMARKS cloud cover \ condition 1 2 3 4 5 6 7 8 9 10 11 12 13 Mean s.5 0.9 50 50 50 70 90 100 100 75 50 50 75 75 75 Sunny Sunny Sunny.5 1.4 1.5 1.1 Tests on 20 June The first test ~ was performed on 20th of June as a preliminary observation preparation for the long term tests.0 0.13 Preliminary test measurements using UNB trigonometric method at SouthGym area from BMl to BM2 Geod.1 2. windy Windy Windy Windy .0 3.1 1.8 0.4 2.1 0.9 1.3 0.1 2.95 1.2 0.9 1.Trig. Windy Windy Windy Breeze Windy Sunny periods.4 0. D.0 2. 0.0 1.7 4.3 1.13 geodetic height differences and UNB method trigonometric height differences.1 0.5 Computed Versus Measured Refraction Effect 5.
The meteorological were: air temperature measured distances. (the middle region 3. inversion with light below this range.94 The day on which the reported as windy.g. and observations done and pressure on 20 June 1985 for correcting a record of cloud the cover and wind.s. Webb. (see one windy day [1984]). cover.g. e.3 for of of thermal more detail) extends usually to more than 30 m (one Obukhov length) and according to theories temperature. first three rows show positive usually longer The results of the values for the lower targets and negative values for the heigher targets. be due to inversion of but later it was found refraction effect can change sight regardless of On a the windy day.6).1. stability. on a cannot appear days. Obukhov length observations were carried out was In a moderatly windy day of summer.1.2 is Greening. However momentarily inversion first three metres of cloud on air is less of calm than 2 of temperature the atmospheric this does not necessarily mean that the lower targets will result in a sign of refraction effect different from the results of the higher targets (see also section 5. when m. . the defined in section than 30m (see e. of summer may expect But. the horizontal movement of gradients within the layer. thought to gradient. 3.3. see section the temperature that the magnitude of the sign with the sign of the region II This was first elevating line temperature gradients. [1985]).
It was supposed to continue for 24 weather conditions it hours. BM2 and BM3 were measured repeatedly over the whole 13hour period. For all other tests the wind velocity and direction.15 and using show the obtained the can be interpreted as effect by assuming levelling as being errorless.95 No temperature gradient was measured. and temperature of measured and cloud cover this information was useful the changes of gradient was used to compute the was recorded. the possible their only the temperature magnitude of refraction effect. for understanding atmospheric condition and if correlation with the refraction error.16 differences levelling discrepancies refraction 5. temperature gradient.14.5. Tables between 5. atmospheric pressure. but due to unfavourable was interrupted after 13 hours of continuous observations. 5. The height differences between BM1. the discrepancies using precise UNBmethod. ground surface were Although all humidity. These mostly due results of to the geodetic . therefore the magnitude of refraction effect using the meteorological data could not be obtained for the first test.2 Tests Qn l i July ~ The second test was carried out on 19 July 1985. the geodetic height and 5.
96
TABLE 5.14
Discrepancies between the results obtained using
trigonometric height traversing and geodetic levelling for
BMl to BM2.
Geod.
N0
1
2
3
4
5
6
7
8
9
10
11
12

Trig. [mm]
mean
of 2
tar
LOCAL
TIME
2.lm
tar
3.5m
tar
11:13
12:09
13:24
14:54
15:49
18:05
18:56
19:31
20:09
21:56
22:47
23:25
0.1
2.0
1.4
2.7
3.6
1.7
2.4
0.8
2.6
1.7
1.9
0.7
0.9
0.5
2.7
2.0
1.2
1.5
0.8
1.0
0.3
2.2
0.7
0.7
1.80
±0.97
1.13 1.45
±0.87 ±0.64
Mean
s. D.
Since the profile
line is
the same as the
expect that over
0.5
0.7
2.0
2.3
2.4
1.6
1.6
0.9
1.5
1.9
1.3
0.7
REMARKS
cloud
cover
condition
'
75
100
50
75
100
100
100
100
100
Fair
Windy
Windy, little rain
Little wind
After a period of rain
Cloudy
Cloudy
of surface on the
foresight of one
backsight of the other,
a long period the misclosure
height differences would be near
for the lower targets is
zero.
0.03 mm
i.e.
one can
of the three
The mean misclosure
as expected.
But
the mean misclosure for the higher targets is 0.85 mm which
is considered as too large.
Column 2 of Table 5.15 shows that the mean discrepancies for
target 3.5 m is a negative
number.
This value was expected
97
•
TABLE 5.15
Discrepancies between the results obtained using
trigonometric height traversing and geodetic levelling for
BM2 to BM3.
REMARKS
Geod.  Trig. [mm]
N0
1
2
3
4
5
6
7
8
9
10
11
12
mean
of 2
tar
LOCAL
TIME
2.1m
tar
3.5m
tar
11:31
12:28
14:00
15:16
16:04
18:29
19:07
19:41
20:21
22:18
22:59
23:35
1.5
1.6
2.2
0.3
0.0
0.3
0.8
2.0
2.3
1.1
0.7
0.6
1.8
2.7
0.8
0.5
0.1
0.1
0.4
0.7
1.7
0.5
0.6
0.5
0.60
±1.26
0.22 0.19
±1.18 ±1.13
.Mean
S. D.
cloud
cover
condition
%
shimmers on BM3
75
Shimmers on BM2.
25
Windy
50
Windy, little rain
100
Little windy
100
100
100
100
100
100 After a period of rain
100
Cool
Little windy
100
1.6
2.1
0.7
0.1
0.1
0.1
0.6
1.3
2.0
0.3
0.7
0.5
to be positive. The computed refraction using the profile of
the lines and
measured temperature gradient in
Table 5.17,
indicates positive mean values for both targets.
This
RBRF > 0
can
be
AHGAHT < 0
then
AHGAHT > 0.
The
lower targets
The temperature
explained
first case
and the second
gradient on
ground (0.3 m to 1.2 m)
by
using
and
Figure
RBRF < 0
when
had mostly
is larger
then
occurred for
case for the
the grass
When
5.8
the
higher targets.
field close
in magnitude
to the
than the
98
TABLE 5.16
Discrepancies between the results obtained using
trigonometric height traversing and geodetic levelling for
BM3 to BM1.
REMARKS
Geod.  Trig. [mm]
NO
LOCAL
TIME
2.1m
tar
3.5m
tar
mean
of 2
tar
cloud
cover
11:49
12:46
14:17
15:36
16:32
18:43
19:19
19:59
22:35
23:12
00:05
1.5
2.7
3.1
1.2
1.7
0.2
0.2
0.4
1.1
0.8
0.4
1.9
0.3
1.8
0.4
0.9
0.6
0.1
0.4
0.0
0.0
0.7
1.7
1.2
2.4
0.8
1.3
0.2
0.1
0.0
0.5
0.4
0.5
75
Breeze
Sunny
0
95
100
Little rain & wind
Little windy
100
100
100
100
100 After a period of rain
100
100
1.17
±1.03
0.5
0.83
±0.79 ±0.75
\
condition

1
2
3
4
5
6
7
8
9
10
11
Mean
S. D.
temperature gradient over asphalt close
it is smaller when going higher
4 rn).
cannot
This may explain the
explain the
higher targets.
was not
repeated in
nearly zero.
and
above the surface (1.2 rn to
different signa of values,
large miaclosure
However,
to the ground,
of
0.85 mm for
but
the
this rather irregular misclosure
the three other
cases and
was always
1 1.4 0.6 0.5 0.5m Mean 2 .5 0.6 0.83 +2.5m Mean 2.2 0.65 +1.3 0..2 0.8 1.7 0.6 2.0 2.7 0.2 0.2 0.99 TABLE 5.7 4.0 1.5m Mean ..5 1..0 0.3 3.4 0.2 0.2 0.4 1.3 0.6 0.1 0..11 0.17 Computed refraction using measured temperature gradient Computed Refraction [mm) LOCAL BMl to BM2 BM2 to BM3 BM3 to BMl NO TIME 2.1m 3. D.7 1.2 0.33 :t0.85 .9 3.6 0.0 0..0 0.3 1.7 0.7 0.7 1.6 1.1 1.8 1.2 1. 2.7 0.1m 3.9 4.6 2.. 1 2 3 4 5 6 7 11:00 12:00 13:00 14:00 15:00 16:00 17:00 8 18:00 9 19:00 10 20:00 Mean s.2 0.2 0.1 0.4 0.4 1.9 1.6 3.4 0.84 4..7 1.9 0.1m 3.2 1.8 0.
RF) < 0 THEN (6HG .6HT) < 0 (RB .100 6HF 6HG 6HT 6H 6HT I) If II) If height difference using geodetic levelling height difference using the leapfrog trigonometric method 6HG = 6HB + 6HF = 6HB + RB + 6HF .5. was measured during profile of the the lines and .8: Back.RF) Figure 5.6HT) > 0 (RB .and foresight magnitude of refraction difference 5.RF = 6HG + (RB . the SouthGym area started at of air Knowing the to the next day until continuous observations.RF) ~ >0 THEN (6HG .3 The third test survey in 9:43 on 23 July 1985 23:50 for The a total of and continued 38 hours of temperature gradient entire test period.
is defined by the regression equation (the cr. the refraction effect for all three lines was computed for one hour intervals. The relationship between the refraction.6HT).3 + 1.9 shows: 1.12) .101 using equations (3. and the correlation is more pronounced in the latter for BM2 to BM3.).10 has the expected hypothetical regression mr = cr).10): mr = 0.3.67). (the solid line) A fitted linear regression line in Figure 5.9 parts a and b.2 cr (5. The correlation between the two can be seen in Figure 5. The computed refraction at one hour intervals using the observed temperature gradients.8. The measured discrepancy between using geodetic the refraction the height levelling and UNBmethod of effect which difference the results trigonometric is the determined obtained with height traversing (6HG. Figure 5. and computed solid line in Figure 5. small deviations from line (the dashed line: This figure shows the measured refraction effect against the computed refraction effect for the line BM2BM3. measured. The error bars in this figure the observations and account other errors are based on the precision of they have been increased to take into (this will be briefly discussed in section 5. mr.66) and (3. 2.
Using these graphs and considering the corresponding time of neutral condition when the refraction fluctuation around this refraction and angle effect is expected can be observed at corresponding difference is zero point is computed simply neutral condition measured the to be almost the mostly because by subtracting time vertical angles angle of zero. 5.102 = 0. and from all the other assuming The of angle the of . r =0 of significance Ho : r At the • 1% level r is rejected and thus Ha ~ 0.11 found by detecting 3. Figure temperature 5. i. In order to investigate the computed refraction we instrument station assessment of may difference of measured and consider to the bench the refraction the lines from marks separately.2) from the temperature depicts the approximate time of neutral condition for different sufaces and elevations.4. computation over the A good each ground surface is also possible by examining them individualy. This is almost the The lowest measured and same for the correlation exists the computed refraction between the effects for the line BM3BM1 (see Table 5.60 with correlation coefficient. can be (see section Figure 5. i.1. refraction.12 gradient shows the crosses the fluctuations of zero observed vertical angles to the lower targets.e. The neutral refraction angle condition measurements.12). correlation exists.e. can be accepted (for details of the test see section line BM1BM2.4). when the periodic line.
12:00 20:00 04:00 12:00 20~00 Zone Time [h) Figure 5. . ! I ~ · II ~ I b I ~ 0 ~ "' 0 m ~ 4l <l) o:. b. BM2BM3 and c. BM1BM2.9: Measured refraction effect versus the computed value. BM3BM1. a.It m: j 103 computed measured 2 4 e ~ 1 2 1 I.
10: The measured refraction effect [mm].3. its accuracy due to the about 0. introduced this way is as precise i. This angle of refraction is converted to linear refraction and compared with the computed refraction for all three lines in Figure 5.104 4 :&• .e. refraction found in vertical angle.13 • In all three plots the strong . of the bias detecting the neutral conditon or the zero refraction angle time. The discrepancies of height difference determined by trigonometric height traversing and geodetic levelling for BMlBM2 line versus the computed refraction error for the same line. The dashed line is the hypothetical regression line. and the solid line represents the actual linear regression.6" can be (see as the measured section lower because uncertaintity for 5. however.8). ~ e e 2 • ~ • •• • 0 ~ ~ ~ ~ ro m ~ • • 2 • • ~ • • ·~ • • • • e 4 2 4 4 2 0 computed [mm] Figure 5.
refraction effects. the correlation coefficient. Figure 5. These plots are prepared for the mean corresponding values of the two targets. the regression equations show that the computed refraction effect. cr. According direction of the to Webb [1968] this can . is too large compared to the measured refraction effect.105 correlation between the two computed and measured refraction effects can be seen. 2.18. Specifically over asphalt. The wind was blowing almost in the line of sight to BM3. The site measurements chosen to was too make the temperature close to the gradient ground covered by gravel (about 2 m) and was not characteristic of the road on which reason for the line of sight to choosing that spot transportation to move BM3 was extended. In Table 5.18 gives and measured effect. mr. was the lack of the equipment from one The proper site to another. equations relating and at hypotheses the computed to the the that 1% there level is which exist of zero Table linear regression measured refraction significance the correlation null between the measured and the computed refraction effect (Ho : r = 0). the is measured refraction smaller than the effect computed one.14 gives the linear correlations between computed 5. explanations of why the results more than There are four two times possible of asphalt are so different from the other two surfaces: 1.
11: Fluctuations of point temperature oradient a. gravel....1 .5 2 2.8 E ' 0.106 a 0. grass and c. 0 . b.4 u 0.. · 0.5 3 Height Above the Ground (m] Figure 5. · c 0..~~r12:00 20:00 04:00 20:00 12:00 Zone Time [h] .l. ._: ·.6~~~r~~~~...· .4 ·. ... asphalt...
and c.. .. \ 0 ~ 0 '• 36 b I 12:00 20:00 04:00 • 12:00 20:00 Zone Time [h] Figure 5. BM3 (asphalt). BM2 (grass).. ~ a c:: «3 ... ~ \ \) ~~\ . ... ! \f!\A . I~A~~l 4i' ~ 39'J ~~~ \.107 Ul G) 17 .....12: Fluctuations of observed vertical angles a. b. BMl (gravel).
... '. ISBM2 (grass) and c.: . a. ~ _.. ' ... ·' : . . ....... .. . ... {a) Gravel . " ' ' ' .: .. . ...13: Computed refraction effect versus the measured value. . (c) Asphalt I ' r I '' 12:00 20:00 04:00 12:00 I ' 20:00 Zone Time [h] Figure 5.• ~~ ...... ' ... . ... ..'. ..' ~ ·. ISBM3 (asphalt}.. b.108 computed measured mm 4 ... ·... I\ '. .' .'· I .. e e .. 4 ......• f··'\ 4 .. ISBMl (gravel}.. .·.' 2 (b) Grass " ( .'I I .
for the as can be BM3BM1 weak correlation equations in Table 5.12 It was Figure previously mentioned that. The correlation coefficient improved to 0.109 result in lower (computed) value. the traffic on the during road could also result in a mixing of the atmosphere and a reducing of the effect of The lower refraction effect over asphalt can refraction. the day. 5.18 computed refraction two equations. The significant refraction effects refraction or trigonometric against the correlations (either derived from discrepancies methods of height computed detailed knowledge that of refraction the from the between differnce is measured angle geodetic of and determination) mostly we have about the two due to the surface profiles and to the long term temperature gradient measurements. refraction effect than the expected Other than the wind direction.9 levelling c. from the small be noticed fluctuations of the vertical angle in comparison to the other two in Figure 5. .12).68 . given cr computed then mr will be the corrected refraction effect. significant improvement in the correlation found between the two after was as A coefficient was the corrections were made. discrepancies and trigonometric versus the asphalt gravel. corrected according to these the seen in the geodetic traversing of have very Using the regression and between height computed refraction (Table 5.
ISBM3 (asphalt).::s 0. 2 '0 (I) . b... 2 ::l • p. e e 0 0 •• • '0 • Q) ..14: 2 0 2 measured [mm] Linear correlation between the computed and measured refraction error.. 2 e 0 e 0 0 • • •• 4 0 4 6 4 2 2 0 measured 4 2 measured [mm) 2 (c) Asphalt •• 2 e e 0 4 [mmJ •• •• 0 '0 Q) . ISBMl (gravel). ISBM2 (grass) and c. e 0 0 • 4 • 4 Figure 5.. ==' 0. ...110 (a) Gravel (b) Grass 4 • • • 4 • 2 • 2 g .... a.
31+0.005   ** gravel BMl 0.75 * t  and the table values for t at 0. of the height traversing over a temperature can be very realistic.2.13+0. results 33.3 > 2.7 R asphalt BM3 0.111 TABLE 5. 0 2.0 significantly high.18 ttest on the significance of the correlation coefficients Surface from Corr.05 level are: t = 2. IS Coef.005 work. to * Regression Equat.0. known is accepted. H >t t 33.75 R grass BM2 0.23 cr > 2.0.7 > 2.75 5.36+0. trusted in (see that than the practical and may longer cause too 100 m section simulations in 4. cannot be refraction effect in trigonometric known ~ not accurate enough trigonometric the r = t 33.87 mr= 0.51 cr 10.025 lines traversing proved profile with profile.01 = and 0.3). R gradient along the and useful information may .40 cr 10.87 mr= 0.66 mr= 0.0.75 and ** Ho: r =0 is rejected and Ha: Although the correlations are computed refraction error since large it is corrections for height However.
AH n is the simultaneously taken) height (5. details see Chapter 6. Estimation of the standard deviation of one ~H is given by Chrzanowski [1985] a= [ where ~i ] 1/2 2 I 2n d. synchronized in taken time.4 The last test at the SouthGym area was using two independent theodolites separated 29 July. one may expect higher precision than what has resulted from this experiment.5 Since the their not but hours starting measurements were differences were at not affected by shortterm fluctuation of temperature gradient. 11:45 ending were independently a duration for at 17:07. Thus.112 be extracted from For more such simulations. number and An d~H of is average observations (usually the difference of the two standard deviation of . the Two electronic theodolites were used.13) i traversings.5. The main purpose to get sufficient carried out by about 2 m on in making these measurements was data to estimate the actual precision of trigonometric height traversing without the influence of the refraction error. of 5. and observations simultaneously. Tests on 12 ~ ~ and estimation Qf standard deviatipn of vertical angle measurements 5.
18 • Considering the errors involved in both computing the refraction effect. measurements carried out in and 5. 3. five groups indication {only of values that the for these agree one can see that the above closely. equations in particular measuring and lines) Table This can be 5. computed presented in measured on 4.58" for a that corresponds to a zenith angle sight lengths of 200 m.79 mm was found of 0. and corrected Table 5. standard deviation measured in four sets with The bars plotted in Figure 5. extracted from measurements carried out in 23 July 1985 and corrected using equations given in Table 5. measured using electronic theodolite #1.18.15 shows the following refraction errors: 1.9 were computed considering the above estimated standard deviation as well as the contribution of other sources of errors such as a change of the height of targets due to the expansion or contraction of the rods. Figure 5.18 are useful and can improve an the . 2. are in the discussed in distance detail by Chrzanowski These errors [1984] and Greening [1985]. extracted from 24 July 1985 and corrected using equations given in Table 5.18 according and to equations temperature gradient 29 July 1985. measured using electronic theodolite #2. nonverticality of the rods. and errors measurements.113 0.
4). 0 0 0 ~ 08 8.J 0 ~ 0+~~==~~~~r~0 ~ G) ~ 0 6. f::::P 6 8 "8 Q Q !:. b.. [:] 6 •6 0 S 6 0 0 0 0 08'0 0 0 8. 0 0 8. 2 0 0G 0 ~ 0 G) ~ ~ tll s::: 2 0 0 0 "M +. a. • 0 0 • ~ 8. 1::... BM3BM1 (a key to this figure is given in section 5. BM2BM3 and c. . 6 0 0 0 t::.114 A: 1 0:2 0:4 o:s • 2 0 1::. 0 2 <b 0 • 0 0 El • 2 0 . BM1BM2...p 8.:. 0 0 08 0 0 2 12:00 14:00 16:00 Zone Time (h) Figure 5. 0 0 ro e e .... 6 ' 0 06 68 .5. 08..15: Measured refraction error versus the computed value.
12)) .93 and 0.115 computed values. 5. computed refraction from mid effect for individual point to the three bench than these estimated (previously the knowledge along refraction effect of the variablity the line account. a short time after sunrise. effect from one day to another as long as the measurements are carried out within the same portion of and day over the same profile under similar weather conditions. from one bench mark to another) the over estimations mostly cancelled out (see equation (5. 0.5 Comments on SouthGym The refraction treating the the neutral The effect is surveys successfully measurements carried out condition time as computed correlated ~ values with the using measured correlation coefficients of durino (or free from this by near to) refraction error. BM2BM3 and BM3BM1 respectively. a short time before sunset. estimated are strongly effect. Also. can of sight which is larger called "measured" based on detection of neutral condition) values.83 with for lines BM1BM2. as expected. of However.e. realized that the refraction it is is almost the same. . of marks) lines be The over estimation due to insufficient the temperature could not gradient be taken into in the computation of a full line (i. and in the afternoon. The preferable times for observations durino clear days are early in the morning.e.78. procedure refraction 0. The (i.5.
2.5 m height) are less affected by refraction.5. which.15 and 5.6 ~ The final test survey test line. between zenith angle started at spot out in each setup and distance 12:15 and ended at Figure 5. effect height levelling method at a about 0. 5. but this does not necessarily always cause randomization of refraction error. Tests on QQ August 5. temperature measurements to BM4.1 m height) different from the higher targets.2. some of the cases the discrepancies listed 5. sian for lower targets (at 2. 1. on asphalt 39 as consisted of the two SouthGym sets of Observations measured discrepancy using geodetic trigonometric 2. .14. given in section and which difference results is the determined of UNB (~HG.~HT).3 m from the about 20:00 on 06 August 1985.16 show a although in in Tables 5. 5. sidewalk.13.16 shows: refraction the The concrete were carried observations. The computed refraction at 15 minute intervals using the observed temperature gradients. was carried out on the HeadHall A description of the line is The change in temperature at different heights from the ground was measured using site for close six temperature sensors.116 Measurements to the higher target (at 3. The was selected setups of measurements in measurements.
profile of the baseline was well known and the temperature gradient was determined every five minutes. between BM2 and BM4 at HeadHall test line. The average of the measured refraction effect (for the mean of the two targets) came out to be 2.117 o measured • computed .....7 mm. Assuming that the last observation refraction error (since it is the at 20:10 is free of closest possible . 0 • • 0 (. the computed refraction effect observed height differences generally It should be mentioned that the to correct the improves the results..2 mm and the corresponding computed value is 2.j 0 • 0 0 0 Q) ~ ~ • 2 Figure 5. ~ 6 4 0 t:z1 • 2 a 0 • 0 0 • •• ~G e 0 • • ••• 0 • 0 0 0 0 • • •• o c 0 • • 0 0 0 0 0 0 • • • • • • (:) 0 0 0 0 0 0 0 +~~~r.~ v • 14 ~ Q) 0:: 0 c ~ 6m G{:l.16: Applying 20:00 18:00 16:00 Zone Time [h) 14:00 The discrepancies of height difference determined by trigonometric height traversing and geodetic levelling...
12). The refraction effect on the or less backsight is oscillating more around the zero value and on foresight is always less than zero as expected. foresight.93 measurements carried can be assumed and then ground. total of the above two (measured.19 the refraction the was obtained following values of refraction effect: 1. This method of neutral condition detection can be utilized in precise measurements of monitoring the vertical movements of large structures. Table foresights shows 5. Excellent values agreement exists (11 and coefficient of 12 0.19) which the two with as a reference (free effect can The neutral temperatures at a condition different measured correlation substantiates out during the neutral the refraction measurements. backsight. AHG .AHT (measured. 2. 3.118 condition time). tl). computed. observation to the neutral effect on and backsights separatly. 5. measuring in Table between that the condition time from refraction error) be computed can for other be detected elevations above by the In this case the profile of the line is not needed. 4. . and 5. such as dams.
0 2.84 2.25 2.5.9 31.65 3.5 2.60 2.85 2.70 A key to this table is given in section 5.55 1.20 2.24 0.0 1.2 2.2 0.1 28.6.2 3.55 2.45 3.3 31.42 2.20 2.96 3.70 5.3 27.3 25.58 2.4 1.4 31.03 0.2 2.13 0.21 2.0 3. [mm] [mm] (mm] Temp.29 2.83 0.20 2.3 1. 2.06 1.5 27.0 26.85 4.09 0.4 1.4 32.17 0.66 0.0 2.55 2.89 2.40 0. Comp.1 30.1 30.6 30.7 2.21 2.71 4.3 1.5 2.0 0.35 3.09 1.0 2.75 2. Remarks calm and clear all day .0 31.8 1. i1 42 Refr.05 2.5 0.23 0.91 3.0 2. Refr.75 2.15 5.42 0.8 31.60 0. F.5 1.65 3.38 2.15 2.97 1.28 0.53 2.119 TABLE 5.74 0.5 31.19 Computed refraction effect versus the HeadHall test line measured value for 2 3 4 5 I I Meas.41 0.28 0.22 0. Refr.59 0.4 3.1 1.4 32.92 1.85 2.30 1.4 29.06 2.10 4.9 28.25 0.49 3.2 1.2 1.3 2.5 2.0 31.2 29.3 29.25 2.09 0.40 0.35 2.2 4.68 0.27 0.05 0.25 3.7 2. [mm] [mm] 0.0 3.24 0.08 0.45 2.85 0.23 0.06 0.45 0.7 2.03 3.30 2.3 30.16 2.75 2.3 0.25 2.50 3.6 2.74 0.51 0.51 3.25 0.55 3.15 1.01 2.1 2.35 0.61 0. 79 1.84 0.8 32.65 4.37 1. 1 NO LOCAL TIME 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 12:00 :15 :30 :45 13:00 :15 :30 :45 14:00 :15 :30 :45 15:00 :15 :30 :45 16:00 :15 :30 :45 17:00 :15 :30 :45 18:00 :15 :30 :45 19:00 :15 :30 :45 20:00 :10 Mean B.40 1.20 0. [oC] 29. Meas.6 3.s.09 0.20 2.43 3.16 1.13 1.69 0.4 2.85 2.s.06 3.53 2.7 29.9 3.8 29.3 31.50 0.20 2.3 3.0 3.5 1.12 0.22 0.26 1.7 1.60 2.41 0.8 31.20 2.9 3.2 2.2 1.87 0.7 2.7 2.0 0.3 31.75 0.1 2.6 3.6 31.7 30.0 31.
.
.Chapter 6 SIMULATIONS Qf REFRACTION ERROR IN TRIGONOMETRIC HEIGHT TRAVERSING 6.120  is less where K During . levelling and an terrain in 82 km long error in with temperature gradient. A number of new simulations has been done by the author with the same actual purpose as in previous levelling conducted data. 1978]. of the The simulations were cumulative influence height traversing. The results reported by Chrzanowski in details trigonometric of refraction in geodetic trigonometric heighting were simulated profile environmental helpful to understand the of refraction The influence extreme simulated along an assumed average of these simulations are [1984] and Greening [1985]. along a line The new of simulations but using simulations actual geodetic have been levelling of special order (the allowable discrepancy between independent forward and backward levelling between bench marks than +3 mm~K [Surveys and Mapping Branch.1 Simulation Alonq g Geodetic Levelling Line on Vancouver Island In 1984 a number of levelling simulations were carried out at UNB to assess the dependence of the refraction errors on the profile conditions. is the distance between bench marks in kilometres).
2 km km km km Mervi1le Kelsey Bay Merville Nanoose Bay Nanoose Bay Merville Kelsey Bay Merville Lines #1 and #4 are extended. parts for forward and The backward levelling traverses: line #1 #2 #3 #4 to from Forward Forward Backward Backward 108. levelling instrument to back and foresight levelling rods were measured using stadia cross hairs. project on Vancouver The from the above data . The operations.5 115. geodetic levelling was carried out the Geodetic Survey of Canada.5 109.5 m and distances from the temperatures were 2.totally over a parallel to Georgia. Surveys and Resources Canada.7 115. out over a line ~224 into two Mines The Mapping Branch. of the line was generated for 10 m (or less) The profile intervals along the line based on measured stadia distances and height . and started in late May 1984 The middle of October 1984.121 the geodetic observed at heights 0. and was finished in the simulations have been carried km line chosen is divided Island by Department of Energy. Kitchener and Hkusam mountains with maximum height differences of about 300 m. the southern Lines #2 and shoreline #3 pass of flat terrain. the partially Strait over a of hilly terrain south of the Menzies.5 m in every setups.
The actual accumulation geodetic levelling has refraction effect been computed and the refraction accumulation was done of the using the same in trigonometric in simulation of height traversing temperature gradient and the same profile.1. the slope height of points was interpolated 6.5 for the profile with along of the lines. 5. Computation 21 the refraction error in geodetic levellina 6. adjacent every two adjacent a straight Assuming line turning points of the as being turning points were connected and then the maximum 10 m horizontal separation See Figures 6.2) . by turning points.122 differences of ground two between constant.50 [ 1  c + 1 C+l ( 50 C+1 250 ) + 150 c . 2 (S/50) . Kukkamaki [1939a] to a form equation was simpilified by more convenient for computation using a hand calculator Cr = 10 5 G .1) in which.2 to the line. At (6.1 The temperature differences measured during levelling were used in Kukkamaki's formulae for refraction correction using equation This (3.94 G = c c 250 . 200 ] (6.19). Ah .
19) given by Kukkamaki [Remmer.1 in all circumstances However. 1980] 2 S Cr = 6 where: the dn d4 ()( d2 ~h dt dn/dt second and is + 3 ~h given by equation (3.1). c is the refraction difference Ah.5 m above the ground surface.3) ) 80 c2 d2 and d4 are of Kukkamaki's .1. correction varies Kukkamaki assumes that linearly with the height and the measured temperature difference In equation (6.08 mm can = 50 m and formula which is an adaptation of 1/3 G = 80.5 m and 0. Ah = the levelled height difference in scale division 0.5 mm.25 oc. ~h computed using Remmer's of equation (3. ~t.4 corresponding average value can be used of c = 0. for c = 0. suggests that an G = 69. the author equation assumed a (5.07 mm and assuming ~t constant value obtaining = 0. and ~t = the measured temperature difference between 2. Kukkamaki [1979].11). For Cr = 0. with be found = 2 m. without causing any significant loss of accuracy. Based on Hytonen's [1967] investigations. Cr = 0. S Geodetic refraction for c was also as in c = 1/3. the fourth derivatives temperature function respectively which are d2 = {C1) C b z (6.2).11).123 where. In equation (6. the exponent in Kukkamaki's model.5.
3) single setup.4) R in which 6 = = S 2 = = this can be written as 8 I n For a special case. I when = 6 1 (6.5). Figura 6.1).2 to 6.19) and (6. The refraction in geodetic levelling The mean the close agreement. (3.1. 2 The refraction effect in trigonometric height traversing was computed using equation (2.1. Refraction error in trigonometric height traversing 6 .1).13) Cr = 1 I k (S  dx X) R is given by equation (4. were c4 compute the refraction effect in geodetic levelling along the above lines and as can be seen in Figure the answers 6.g. came out to be in figure shows the effect of along the line comparison with 43. . of the three simulated refraction is used effect in for the trigonometric methods (see e.124 and d4 = (c3) (c2) (c1) c b z S is the sight length and used to ~h is the height difference for a All three equations (6.
(S . Results of simulations Figures simulations.5) 8 n where 502.7) z and S is the sight length of s (S 6 . 3 = divided into n equal subsections n s).1.. special order other hand..125 8 =k I 1 s 2 + k 8 .[ 0. but.0342 + (dt/dz) ] . one can would be expected. to 6. keep the maximum To length show this of sight . the refraction errors in trigonometric height traversing are not correlated with the profile of the route. the geodetic levelling refraction effect is highly correlated with the the levelling route. As can be seen.7 p k = .S)+k 8 2 k 2 (6. They are highly dependent on the clearance and the length of the line of sight as it dependency. fluctuates profile of within the limits of On the of Canadian accuracy specifications..2 6.6) i 2 i T = (1/3) b dt/dz 4/3 (6..2 B)+ 3 2 1 ••••• + (S . (6.5 The starting show the results heights of of the profiles four were chosen arbitrarily.
the author's in motorized A considerable improvement the results of simulations whenever a height of larger than 2. the of these it is cases and substantiated that if the lines leapfrog method are less clearance greater than 1 m. keeping unchanged while increasing the sight length. will result in a higher refraction error. usually more trigonometric height traversing. those presented case of reciprocal method. the limits of specifications ray for the first U ~ 2. is noticed in height of results shown according than 2 m for to 2. . that the assumed for of is all the instrument. where K is in The line of sight should be less than 250 m for the clearance of 1 m in the should also be mentioned instrument is experience with the given by Chrzanowski [1984] in terms of standard deviation as height of sight in The Canadian specification for first order and for oneway levelling is The than 150 m in this the refraction effect is within the Canadian order levelling.126 unchanged while changing the error the clearance.0 mm~K. of refraction will where clearance. Many more simulation were results obtained for different cases of the sight lengths and the ray clearances. kilometres. and decreasing the sight length will help to decrease the refraction effect.2 m for the instrument is assumed. By inspection thesis. increased as well. the clearance the In all be reduced by number of four lines increasing the observations will be Conversely in most of the cases.2 m It the in here.
8) Equating the right hand sides of equations (6.127 6.5 rn above the the ground).5 m height) and higher (at 2.7) and (6. mild. The H using To can look at the the route of the can be estimated gradient profile in the conditions.5 c = 1/3 c . . b is given in terms of ~t according to equation (3.5) b H =[ (6.2 Simulation of the Refraction Error Using other Values of Temperature Gradient Measurements According to the measured temperatures (at height 0.4) b = ~t where c I ( 2. dt 2/3 = 0.10) ~t is the temperature difference between lower (at 0.9) 0. middle region equation (3.5 m height) sensors.32) and neglecting the adiabatic lapse rate r = 0.8).1.23)). H condition during one along (#1 to #4). flux.0274 dz H 4/3 z (6.0. H can be written in terms of b (see section 3.5 m and 2. levelling in investigate the changes of Vancouver changes of sensible heat above levelling lines from the temperature under unstable the weather Island was quite weather.0822 in which.5 ) and (6.0098 (see equation (3.
67 23.5 .75 0.6 along the The line i2 . For other examples of (Fredericton.2 and 6.33) or be slightly figure.1 shows the averaged and its corresponding H values along the four levelling lines.69 28 17 25 24 MEAN 0.B. TABLE Average ~t.3 show other parts of Canada United States. these values in in the in this with the help of coefficient b and negative values may Table 6.128 shows the variations of sensible heat flux Figure 6.36 0.71 0. different from those depicted negative values reflect to be computed (3.53 0. Line 6. one can expect higher H and b values during the summer than those from Vancouver Island. example in N.) Fredericton. according to meteorological obsevations carried out.1 b and H along the levelling routes ~t b [oC] H [W 2 m ] il i2 i3 i4 0.35 0.39 0. the they have the corresponding equation 6. since the stable condition and Table (3.37 0.34).28 0.
geodetic levelling needs to In be corrected .7 0.7 0.57 1.0.8 0.7 and 6.9 0.7 0. of refraction Figures 6. 53 1. 34 l.7 0. 34 1. 72 1. b and H in Fredericton.7 l.53 1.129 TABLE 6.0 0.3 0. 91 1.2) in the one obtains a set of results which can represent the refraction error in the summer conditions of the Fredericton Figures 6. 72 0.6 0.5 .1 0.2 Average Date 1985 Ground Cover Jul19 Jul19 Jul19 Jul23 Jul23 Jul23 Jul24 Jul24 Jul24 Jul29 Jul29 Jul29 Aug06 Aug10 AugAug14 Aug15 Aug15 gravel grass asphalt gravel grass asphalt gravel grass asphalt gravel grass asphalt asphalt highway highway highway highway highway MEAN ~t.3 and 6..15 1. the area.7 1. 34 1.6 0.15 0.25 (see Table simulations along the Vancouver Island profile. N..7 0.19 1.8 show the simulation error on the same profiles as in these cases.B.7 0.. 53 52 18 66 66 52 66 80 66 66 66 95 80 66 112 66 95 4 80 10:2519:59 10:4519:35 10:1020:32 10:1018:20 09:3018:40 09:5019:00 07:0017:20 07:2018:48 07:4519:05 12:1017:10 11:3016:30 11:5016:50 11:0020:15 07:2815:23 13:3215:04 11:1316:~5 07:2812:34 13:0615:36 partially cloudy partially cloudy partially cloudy mostly sunny mostly sunny mostly sunny mostly sunny mostly sunny mostly sunny partially cloudy partially cloudy partially cloudy mostly sunny mostly sunny mostly sunny mostly sunny mostly cloudy partial