Global Positioning System Reference
In-Depth Information
10°
0.04
6
2
TEC = 50
10
electrons/m
×
30°
f
= 1.575 GHz
0.03
60°
0.02
0.01
0
0
20
40
60
80
100
120
Separation (km)
Figure 8.8
Variation of ionospheric delay difference due to elevation angle differences.
in magnitude 95% of the time, even during the peak of the 11-year solar cycle. Vari-
ous physical phenomena, including traveling ionospheric disturbances (TIDs),
which are small-scale irregularities in the ionosphere, can cause steep spatial
gradients in TEC over distances as short as 10 km.
Ionospheric delays typically change very slowly with time, normally following a
daily cycle of very low values at local nighttime, followed by a ramping up to a max-
imal delay in the early local afternoon, and then a decline back to the steady night
value again. In midlatitude regions, the time rate of change of vertical ionospheric
delays rarely exceeds 8 cm/min [9]. In other regions of the world, rates of up to 65
cm/min have been observed [9]. Some recent studies have indicated that rates of over
3 m/min may occur on rare occasions. These observed rates include both the effect of
changing elevation angles and TEC.
8.2.5 Receiver Noise and Multipath
Unlike the other error sources considered thus far, receiver noise and multipath
result in pseudorange and carrier-phase errors that are uncorrelated between receiv-
ers separated by even very short baselines. Multipath, in particular, often dominates
error budgets for short-baseline code- and carrier-based DGPS systems for two rea-
sons. First, it causes pseudorange and carrier-phase errors that are generally statisti-
cally larger than those caused by receiver noise. Second, the fact that multipath
errors are uncorrelated from receiver to receiver means that the difference in mea-
surement error caused by multipath between two receivers has a variance described
as the sum of the multipath error variance attributable to each alone. As discussed in
Section 6.3, the magnitude of multipath errors varies significantly depending on the
type of receiver and environment. In this section, we will assume a 1-sigma
pseudorange multipath error of 0.2m for each receiver, which is consistent with the
performance of a high-end C/A code receiver in a relatively benign multipath
environment.
 
Search WWH ::




Custom Search