Small, reasonably cheap, hand-held GPS receivers have been available since
about 1990. Until May 2000, however, their accuracy was no better than
a few hundred metres in position and even less in elevation, because of
deliberate signal degradation for military reasons (‘selective availability’ or
SA). The instruments were thus useful only for the most regional of surveys.
For more accurate work, differential GPS (DGPS) was required, involving
a base station and recordings, both in the field and at the base, of the estimated
ranges to individual satellites. Transmitted corrections that could be
picked up by the field receiver allowed real-time kinetic positioning (RTKP).
Because of SA, differential methods were essential if GPS positioning was
to replace more traditional methods in most surveys, even though the accuracies
obtainable in differential mode were usually greater than needed for
geophysical purposes.
Accuracies in hand-held GPS receivers
The removal of SA dramatically reduced the positional error in non-differential
GPS, and signals also became easier to acquire. It is often now possible to
obtain fixes through forest canopy, although buildings or solid rock between
receiver and satellite still present insuperable obstacles. The precision of the
readouts on small hand-held instruments, for both elevations and co-ordinates,
is generally to the nearest metre, or its rough equivalent in latitude and longitude
(0.00001◦). Accuracies are considerably less, because of multi-path
errors (i.e. reflections from topography or buildings providing alternative
paths of different lengths) and because of variations in the properties of the
atmosphere. The main atmospheric effects occur in the ionosphere and depend
on the magnitude and variability of the ionization. They are thus most severe
during periods of high solar activity, and particularly during magnetic storms
(Section 3.2.4).
Because of atmospheric variations, all three co-ordinates displayed on a
hand-held GPS will usually vary over a range of several metres within a
period of a few minutes, and by several tens of metres over longer time
intervals. Despite this, it is now feasible to use a hand-held GPS for surveys
with inter-station separations of 100 m or even less because GPS errors,
even if significant fractions of station spacing, are not, as are so many other
errors, cumulative. Moreover, rapid movement from station to station is, in
effect, a primitive form of DGPS, and if fixes at adjacent stations are taken
within a few minutes of each other, the error in determining the intervening
distance will be of the order of 5 metres or less. (In theory, this will not
work, because corrections for transmission path variations should be made
individually for each individual satellite used, and this cannot be done with
the hand-held instruments currently available. However, if distances and time
intervals between readings are both small, it is likely that the same satellite
constellation will have been used for all estimates and that the atmospheric
changes will also be small.)
Elevations from hand-held GPS receivers
In some geophysical work, errors of the order of 10 metres may be acceptable
for horizontal co-ordinates but not for elevations, and DGPS is then still
needed. There is a further complication with ‘raw’ GPS elevations, since
these are referenced to an ellipsoid. A national elevation datum is, however,
almost always based on the local position of the geoid via the mean sea level
at some selected port. Differences of several tens of metres between geoid
and ellipsoid are common, and the source of frequent complaints from users
that their instruments never show zero at sea level! In extreme cases, the
difference may exceed 100 m.
Most hand-held instruments give reasonable positional fixes using three
satellites but need four to even attempt an elevation. This is because the
unknown quantities at each fix include the value of the offset between
the instrument’s internal clock and the synchronized clocks of the satellite
.bmp)
constellation. Four unknowns require four measurements. Unfortunately, in
some cases the information as to whether ‘3D navigation’ is being achieved
is not included on the display that shows the co-ordinates (e.g. Figure 1.15),
and the only indication that the fourth satellite has been ‘lost’ may be a
suspicious lack of variation in the elevation reading.
about 1990. Until May 2000, however, their accuracy was no better than
a few hundred metres in position and even less in elevation, because of
deliberate signal degradation for military reasons (‘selective availability’ or
SA). The instruments were thus useful only for the most regional of surveys.
For more accurate work, differential GPS (DGPS) was required, involving
a base station and recordings, both in the field and at the base, of the estimated
ranges to individual satellites. Transmitted corrections that could be
picked up by the field receiver allowed real-time kinetic positioning (RTKP).
Because of SA, differential methods were essential if GPS positioning was
to replace more traditional methods in most surveys, even though the accuracies
obtainable in differential mode were usually greater than needed for
geophysical purposes.
Accuracies in hand-held GPS receivers
The removal of SA dramatically reduced the positional error in non-differential
GPS, and signals also became easier to acquire. It is often now possible to
obtain fixes through forest canopy, although buildings or solid rock between
receiver and satellite still present insuperable obstacles. The precision of the
readouts on small hand-held instruments, for both elevations and co-ordinates,
is generally to the nearest metre, or its rough equivalent in latitude and longitude
(0.00001◦). Accuracies are considerably less, because of multi-path
errors (i.e. reflections from topography or buildings providing alternative
paths of different lengths) and because of variations in the properties of the
atmosphere. The main atmospheric effects occur in the ionosphere and depend
on the magnitude and variability of the ionization. They are thus most severe
during periods of high solar activity, and particularly during magnetic storms
(Section 3.2.4).
Because of atmospheric variations, all three co-ordinates displayed on a
hand-held GPS will usually vary over a range of several metres within a
period of a few minutes, and by several tens of metres over longer time
intervals. Despite this, it is now feasible to use a hand-held GPS for surveys
with inter-station separations of 100 m or even less because GPS errors,
even if significant fractions of station spacing, are not, as are so many other
errors, cumulative. Moreover, rapid movement from station to station is, in
effect, a primitive form of DGPS, and if fixes at adjacent stations are taken
within a few minutes of each other, the error in determining the intervening
distance will be of the order of 5 metres or less. (In theory, this will not
work, because corrections for transmission path variations should be made
individually for each individual satellite used, and this cannot be done with
the hand-held instruments currently available. However, if distances and time
intervals between readings are both small, it is likely that the same satellite
constellation will have been used for all estimates and that the atmospheric
changes will also be small.)
Elevations from hand-held GPS receivers
In some geophysical work, errors of the order of 10 metres may be acceptable
for horizontal co-ordinates but not for elevations, and DGPS is then still
needed. There is a further complication with ‘raw’ GPS elevations, since
these are referenced to an ellipsoid. A national elevation datum is, however,
almost always based on the local position of the geoid via the mean sea level
at some selected port. Differences of several tens of metres between geoid
and ellipsoid are common, and the source of frequent complaints from users
that their instruments never show zero at sea level! In extreme cases, the
difference may exceed 100 m.
Most hand-held instruments give reasonable positional fixes using three
satellites but need four to even attempt an elevation. This is because the
unknown quantities at each fix include the value of the offset between
the instrument’s internal clock and the synchronized clocks of the satellite
.bmp)
constellation. Four unknowns require four measurements. Unfortunately, in
some cases the information as to whether ‘3D navigation’ is being achieved
is not included on the display that shows the co-ordinates (e.g. Figure 1.15),
and the only indication that the fourth satellite has been ‘lost’ may be a
suspicious lack of variation in the elevation reading.