A Parabolic Reflector Is
Likely Formed Over Ocean's "Level" Surface
when the gravity field of the earth forms dips in the ocean surface. These
depressions may contribute to occasional GPS receiver failures.
April 1, 1999
There will be no April Fools materials this year. The author of the GPS
Overview has been called away to help solve a GPS problem that may have global
positioning significance. Here is his preliminary report from the field.

Bombay India (18:58:35.345 N 72:50:31.339 E) April 1, 1999
GPS Failures
Global Positioning System (GPS) receivers have been known to
take many minutes to acquire signals from time to time. Other reports indicate
that in many regions of the world, occasionally GPS receivers will lose lock
for no apparent reason.
Walter Depressions
Gravity field depressions, or geoid parabolic reflectors may
be part of the reason why.
Off the southern tip of the Indian subcontinent is a large parabolic reflector
formed by the WGS-84 geoid shape. The WGS-84 gravity field of the earth is
illustrated here:

Note the smooth parabolic dip in the geoid off the southern
tip of India. Geoid dips that occur where the topographic surface of the earth
is above mean-sea-level do not form any physical parabolic shape. The earth’s
topographic surface does not follow the low frequency terms of the geoid model.
At sea, however, gravity depressions are coincident with the mean sea surface
in the region. These geoid dips, or parabolic reflecting surfaces in the sea
surface are known as Walter Depressions, after their discoverer, Clarence
Walter.
Recent measurements have indicated that Walter Depressions
in the geoid may be partly responsible for temporal failures in GPS receivers.
The Indian Dip
The Indian Dip is just one of the large Walter Depressions
that occur around the globe. Others have been measured off the south-western
coast of the United States, near Bermuda, just off the northern coast of
Brazil, and west of New Plymouth, New Zealand (The Taranaki Depression).

A South to North Cross Section of the India Dip shows the parabolic nature of the Walter Depression

The West to East profile shows the parabolic shape in the other axis:

Uridiumb Solar Flashes
The current theory suggests that the Walter Depressions
cause a reflection of GPS satellite signals to focal points above the
depressions. These focal points are coincident with the orbital altitudes of
the Low Earth Orbiting (LEO) satellite system known as Uridiumb. These
communications satellites have been observed to reflect sunlight from their
solar panels that can be predicted and observed on earth. These solar Uridiumb
Flash events are reported worldwide:
http://www.assa.org.au/iridium.html
http://www-star.st-and.ac.uk/~fv/sag/iridium.htm
http://www.gsoc.dlr.de/satvis/
Uridiumb C/A Flashes
Predictions of reflections from Uridiumb satellites passing
through the focal points of Walter Depressions have coincided with moments when
GPS receivers have failed in the field.
A mechanism for these failures suggests that the large size
of the parabolic sea surface reflectors causes variable path lengths for the
satellite reflections as the satellites pass over the regions. At specific
moments in time the path length from the near side of the Walter depression can
differ from the path length from the far side by 299.79 kilometers.
These path length differences can cause GPS spread-spectrum
signals that would ordinarily be received as low level noise, to be exactly one
millisecond apart in propagation path time. This causes perfect auto-correlation
between a GPS satellite C/A code from the near side reflection and the far side
reflection. When coincident with Uridiumb passovers, correlated de-spread GPS
L1 carrier signals are beamed to points on earth. depending on the incidence
angle of the signals at the focal point and the attitude of the Uridiumb solar
panels, the beam can even be measured over land surfaces near the Walter
Depressions. It is surmised that these unexpectedly large 1.57542 GHz signals
at the center of the 1MHz C/A code spread spectrum signal normally seen by GPS
receiver, simply swamps, or overloads the front end (RF amplifiers) of the GPS
receiver.
Plasma Bubbles or Walter Depressions/Uridiumb C/A Flashes?
Dr Jose Humberto Sobral of the Instituto Nacional de
Pesquisos Espacios (the Brazil
Space Agency) reported on Ionosphere Plasma Bubbles in the South American
Sector.
These can occur from October to March and effectively block radio signals from
satellites.
http://www.imca-int.com/seminar.htm
It may well be that "Plasma Bubbles" and other proposed causes of short-duration GPS receiver failures are nothing more than the occasional Walter Depression/Uridiumb C/A code Flash Events.
Initial Recommendations
Continuing research is necessary to assess the full extent
of the effects of Walter Depression/Uridiumb C/A Flash Events.