Skip to content
- Per arc: from millimetres (calm water, long arcs, synthetic tests) to a
few centimetres (real arcs with waves, short arcs, low C/N0).
- Per session: the robust combination of many arcs reduces the spread; on
Lake Lugano a 32-minute window with 11 arcs gave a 27 mm spread
(see Field tests).
- Waves broaden the peak; wind-roughened water gives weaker oscillations
(diffuse reflection) and fewer accepted arcs.
- Published results with geodetic and low-cost equipment on lakes report
millimetre-to-centimetre precision for daily means with several
constellations (Sources).
- Elevation timing. The receiver reports elevation in whole degrees.
reflexa micro records the second at which each degree changes (1 Hz), so the
conversion to sin(elevation) is not biased by the rounding. In
simulation this halved the error of a nominal arc (RMSE from 8.2 to
4.5 mm on synthetic data).
- Sampling. The faster the oscillation (higher antenna, faster
satellite), the shorter the sampling interval must be. The app keeps the
interval Auto; a fixed interval may not exceed K ÷ RH max (K ≈ 192 m·s
for GPS, 160 m·s with GLONASS).
- Search range. The RH search range follows from the elevation window;
with Auto the unit searches the widest range it can resolve and moves
the band when reflections land on its edge (the change is reported).
- Edge blocks. Abrupt, persistent drops of C/N0 at the start or end of a
pass are trimmed automatically (see jump detection).
- Changing water. The basic method assumes the level is constant during
an arc. For slow changes (lakes, reservoirs) this is fine; fast tides
need corrections (rate of change, refraction) that are part of the BREVA
processing roadmap.
- How arcs are combined. Each arc that passes the quality rules gives one
reflector height with its own uncertainty; the session result is a robust
(Huber) weighted combination, and weak or band-edge arcs are
down-weighted rather than discarded (details in
Processing & quality).
