Ocean Remote Sensing Using GNSS-R - GNSS Remote Sensing: Theory, Methods and Applications - page 226

multi-satellite altimetry; 7.5 cm precision for static PIT in 1-s and 1-GNSS satellite;
and 0.6 m for 1-s 1-GNSS satellite airborne PIT. Note that the processing of
air-borne altimetric campaigns have in general the drawback of trajectories with
highly and non-predictable dynamics. Those are sometimes difficult to solve at a
few centimeter precision level.
The theoretical expected performance of the PIT technique from space-based,
LEO, platforms has been detailed in Martín-Neira et al. ( 2011 ). This study assumes
much higher antenna gains, and it predicts 13-17 cm precision in 100-km along-
track averaging for an in-orbit demonstration mission (23 dBi gain antennas, 800 km
orbital height), and 5-7.5 cm from an operational one (30 dBi antennas, 1,500 km
orbital height).

9.2 Ocean Surface Roughness

The GNSS, L-band signals, have electromagnetic carrier wavelengths longer than
the fine surface ripples generated by instantaneous winds. In principle, only surface
features of typical length longer than the electromagnetic carrier wavelength can
be sensed, meaning that L-band signals are not in an optimal frequency for
wind monitoring. However, as the wind blows, it transfers energy to the Ocean,
increasing the waves' height and length. One of the discussions among the GNSS-
R community is the strength of the link between GNSS-R observations and wind
speed. Some studies adjusted or calibrated the apparent Ocean surface slopes at L-
band, in the form of a modified relationship between the variance of the slopes
and the wind ( Katzberg et al. 2006 , Eq. 3), and valid for a wide range of wind
speeds. Some others present L-band roughness parameters as a product by itself
(e.g. Cardellach et al. 2003 ; Germain et al. 2004 ). Based on Elfouhaily et al. ( 1997 )
spectrum, Fig. 9.7 displays different relationships between the variance of the slopes
and wind at different stages of development of the sea. It shows that the variance for
a given wind speed depends on the stage of development of the sea. On the other
hand, the drag coefficient is a relevant parameter to model momentum exchanges
between the sea waves and the atmosphere. It can be a function of both the wind
speed and the wave age (e.g. Nordeng 1991 ; Makin et al. 1995 ). This opens potential
inversion schemes, closer to data assimilation approaches, in which independent
wind information could be combined with GNSS-R observations of the L-band
roughness to infer information about wave age or dragging-related parameters.
The L-band roughness parameters might also be suitable sea surface descriptors
to provide roughness corrections to L-band radiometric missions for improving their
sea surface salinity measurements. The L-band radiometric measurements of the
surface salinity have a major systematic effect induced by the surface roughness,
in particular, to the portion of the spectrum to which L-band signals are sensitive.
There are a set of promising studies along these lines, such as Marchan-Hernandez
et al. ( 2008 ), Valencia et al. ( 2009 ), and Camps et al. ( 2011 ).