Geoscience Reference
In-Depth Information
oceanographic, cryospheric and geodetic satellites, progress in understanding some of the
world's major river, floodplain and wetland systems can be made. In Sect.
3
, we describe
the proposed NASA/CNES Surface Water Ocean Topography (SWOT) satellite mission,
which would provide the first dedicated observing system for surface water by measuring
water height (h), water slope (oh
=
ox) and water height change over time (oh
=
ot)at*100-
m spatial resolution between 78N and 78S every 11 days. SWOT would not measure
river discharge directly, and instead, this would need to be estimated using a hydraulic
model driven by the SWOT water elevation observations. Constructing such a model
requires knowledge of the channel bathymetry, which may be poorly known for many
rivers, and estimation of the unknown channel friction. Section
4
therefore describes recent
studies that have explored data assimilation techniques that could use the anticipated
SWOT and existing observations to infer the unknown bathymetry and friction and hence
estimate discharge. Section
5
summarizes progress to date and future prospects for
research in this area.
2 Observing Global Flood Dynamics
An ideal set of measurements for observing global flood dynamics would comprise data
describing channel bathymetry, floodplain topography, river discharge, inundation extent,
water level and water storage at appropriate spatial and temporal resolutions as determined
by our understanding of flood wave physics outlined in Sect.
1
. Determining these reso-
lutions in some situations, however, may not be straightforward. For example, we know
that floodplains consist of features such as former channels, levees, pans and crevasse
splays which give a complex microtopography that can affect both local-scale patterns and
larger-scale flow routing (see, for example, Neal et al.
2012
). Similarly, on the basis of a
small number of unique and opportunistically acquired data sets, we know that water levels
in inundated floodplains and wetlands show significant variability in time over periods of
24 h and in space over length scales down to 10-100 m (see, for example, Nicholas and
Mitchell
2003
; Bates et al.
2006
; Alsdorf et al.
2007b
), yet currently available observations
are incapable of capturing this. Appropriate sampling density will therefore vary with event
dynamics, which will be controlled to first order by basin size and climatology (Bian-
camaria et al.
2010
) and complicated by such factors as basin shape, geology and land use.
Measurements can be taken either through ground observations or using remote sensing
platforms, and these are discussed in more detail below.
Ground observations of surface waters are made through discharge gauging stations;
however, these are located on main rivers only where flow is confined to a single channel
and can be fully sampled by a single measurement. Developed countries may have
extensive ground gauging networks with long records but, worldwide, the number of
gauges is declining (V¨r¨smarty
2002
) and there can be significant barriers to data access.
Moreover, floodplains and wetlands, which may convey a significant quantity of the total
flow (e.g., Richey et al.
1989
), are almost entirely ungauged. We therefore do not currently
possess a comprehensive and globally consistent observing system for surface water.
Nevertheless, at ground gauging sites, frequent water depth measurements can be taken to
centimetric precision and made available in near real time with appropriate telemetry
systems. If the gauge site is geodetically levelled, then absolute water elevation mea-
surements referenced to a local ellipsoid or global geoid are possible. Flow rating curves
constructed by fitting a relationship between repeated simultaneous measurements of flow
cross-sectional area, velocity and depth can then be used to determine discharge through
