GNSS Atmospheric and Multipath Delays - GNSS Remote Sensing: Theory, Methods and Applications - page 29

Multipath propagation degrades the GNSS positioning accuracy. Since GNSS
multipath effects depend on the environment near the receiver, e.g., path geometry,
signal characteristics, diffraction and reflection mediums as well as their changing
nature together with a complex antenna and receiver design, it is very difficult to
remove or mitigate the GNSS multipath effects. Furthermore, the site-dependent
characteristics of multipath decorrelate the errors caused by multipath propagation
at different antenna locations and thus even differential techniques, like double
differences (DD), also cannot remove or mitigate it.
GNSS multipath affects both code and phase measurements. The carrier phase
multipath is the fraction of the GNSS signal's wavelength, whereas code multipath
is limited by the chipping rate (Teunissen and Kleusberg 1998 ). The pseudorange
multipath can degrade the accuracy of GNSS navigation, positioning, timing,
ionospheric monitoring and RTK surveying. In particular, when the pseudorange is
employed for ambiguity resolution purposes, multipath effects on the pseudorange
can increase the time requirement for initialization. Pseudorange multipath on both
frequencies can be investigated and evaluated using the linear combinations for both
pseudorange P 1 and P 2 and carrier phase L 1 and L 2 data with eliminating the effects
of receiver and satellite clocks as well as atmospheric delay. In the following, MP 1
and MP 2 are linear combinations using both pseudorange and carrier phase data
to eliminate the effects of station clocks, satellite clocks, tropospheric delay, and
ionospheric delay (Teunissen and Kleusberg 1998 ):

1 C L 1 C 2

˛ 1

L 2 D M 1 C B 1 1 C m 1 C 2

˛ 1

m 2
(2.24)

2
˛ 1

2
˛ 1

MP1 D P 1

2˛
˛ 1

L 1 C 2˛

˛ 1 1 L 2

MP 2 D P 2

2˛
˛ 1

m 1 C 2˛

˛ 1 1 m 2

D M 2 C B 2 (2.25)

D 1 C

1 n 1 1

C 2
˛ 1 n 2 2 , ˛

D 2˛
˛ 1 n 1 1

f 1
f 2

2

where B 1 D

, B 2

C

˛

2˛
˛1

1 n 2 2, ˛

f 1

f 2 . P 1 and P 2 are the dual-frequency pseudorange obser-
vations, L 1 and L 2 are the dual-frequency carrier phase observations, and m1
and m2 are the dual-frequency carrier phase multipath. The report of TEQC (the
toolkit for GPS/GLONASS/Galileo/SBAS data, UNAVCO: http://facility.unavco.
org/software/teqc/teqc.html ) is the Root Mean Square (RMS) variation of MP1 and
MP2 for each satellite and a mean RMS for all satellites. Therefore the remaining in
MP1 and MP2 is the effects of pseudorange noise (<25 cm), carrier phase multipath
(<7 cm), and carrier phase noise (<2 mm), but these are much smaller when
compared to the pseudorange multipath that can be as large as 10-15 m at low
elevation angles (Hilla and Cline 2002 ).

D