During the snow-free period, we find that reflector height
does not remain exactly zero. Variations occur mainly when
the scattering medium is transitioning, from snow to slush/mud
and eventually grass-covered soil. This issue is both a challenge
and an opportunity. On the one hand, it poses the risk of being
mistaken for snow depth events. In fact, the identification of
the site-overall (i.e., noncluster specific) zero-level or bareground
reflector height is perhaps the weakest link in the whole
GPS processing chain of snow depth retrieval, as it relies on
only a few data points—we compute it as the fifth percentile
of site-average reflector heights over the snow-free period.
On the other hand, such observations attest to the prospects
of using GPS reflector heights for monitoring environmental
targets other than snow, such as vegetation biomass [10]. If
successful, the estimation of nonsnow targets would contribute
to guaranteeing that snow depth remains nonnegative.