IEEE Geoscience and Remote Sensing Magazine - September 2022 - 22

* Direct Comparison
* Indirect Evaluation
* Cross Validation
Integrated
Validation
Strategy
SWS
[6]. Currently, however, derivation approaches for SWS parameters
are not mature enough, specifically those tuned
for accurate estimates of SWS variations of rivers and
floodplain wetlands [6], [24], [207]. Due to the smaller
target size, in these areas, satellite signals are likely to be
degraded and blocked by adjacent land features, including
river banks, wetland vegetation, and mudflats. Some satellite
data processing methods, such as altimetry waveform
retracking, the decomposition of mixed pixels, and downscaling
techniques, continue to be challenging. For these
reasons, monitoring SWS and its dynamics in areas with
complex channel networks is much more difficult through
present satellite sensors.
To overcome this issue, the SWOT mission will help
FIGURE 6. The framework for future research on the derivation
of SWS.
area and water level from remote sensing data [20], [202].
The accuracy of satellite-derived SWS is jointly impacted
by surface variables of water level, water area, and bathymetric
measurements [35]. However, the utility of existing
techniques for satellite-derived surface variables is often
limited by the spatiotemporal resolution of the used sensors
[14]. For water level, the intertrack spatial resolution of
current radar altimeters is generally low and easily misses
many water bodies, such as rivers, floodplain wetlands,
and small reservoirs.
Even though measurements are available, altimetry accuracy
can be attributed to a number of factors, including
target size, surface complexity, and instrument-tracking
logic, ranging from 3-5 cm (for larger water bodies) to
tens of centimeters (for small water bodies). Furthermore,
various satellite observations provide retrieval estimates of
surface water area and could achieve high accuracy of 90%,
but they suffer from limitations, such as spatiotemporal inconsistency
with altimetry sensors. Considered a new generation
of satellite mission for SWS, the upcoming wideswath
SWOT mission is expected to provide simultaneous
water area and water level estimates both temporally and
spatially, which will be an important breakthrough in the
remote sensing of SWS [30], [203]. Based on these observations,
accurate and precise SWS products are envisioned
for surface water bodies with areas as small as 250 × 250 m
under swath coverage with a height accuracy of around
10 cm [204]. These will certainly contribute to continuous
measurements of SWS in terrestrial surface water bodies
at global scales with an unprecedented spatial resolution
[35], [205], [206].
Second, determining how to generate SWS data sets at
a global scale with high accuracy remains challenging and
still needs to be explored [16], [204]. SWS has been considered
an essential water variable for land surface hydrology
22
characterize changes in SWS for all rivers as narrow as 50 m
under the swath coverage [30], [204], [205]. Moreover, the
increasing abilities of high-spatiotemporal-resolution imagery,
such as CubeSat, small satellite, and unmanned aerial
vehicle measurements, will undoubtedly address this issue
[208], [209]. Therefore, innovative and robust algorithms
are direly needed to obtain reliable and high-resolution information
on SWS parameters for these new sensors [210],
which will no doubt inspire and stimulate further excellent
works. From a longer perspective, data assimilation is an alternative
methodology [211]. Further work is also required
to develop advanced data assimilation algorithms to integrate
ground-based and remotely sensed observations into
hydrological models, which can achieve superior and spatiotemporally
continuous estimates of SWS from intermittent
satellite measurements [212].
Third, uncertainty inevitably exists in satellite observations
of SWS and differs greatly among products because
of the complexity of estimating and retrieving SWS. Therefore,
SWS validation and uncertainty assessment are crucial
requirements from the end user perspective on satellitederived
products. Lacking validation, any methods, algorithms,
and parameters derived from EO data cannot be
used with confidence, which may prevent the incorporation
of SWS estimation into decision-making processes and hydrological
applications. Presently, a quantitative evaluation
of derived SWS estimations is particularly difficult due to
the lack of reliable reference ground measurements as well
as recommended validation strategy. Thus, further research
on the uncertainty assessment of SWS satellite observations
should be carried out. First of all, ground observations of
SWS need to be strengthened. In situ measurements provide
a solid baseline to evaluate and quantify the quality of
a satellite-derived SWS product. In this regard, in situ data
collection efforts are indispensable, and we believe they are
expected to benefit from community-based campaigns by
raising awareness of existing ground observation practices
within different communities. This will help to provide
timely and continuous collection of reference data for SWS
research and applications.
In addition, an integrated validation strategy for remotely
sensed SWS products should be developed. This
IEEE GEOSCIENCE AND REMOTE SENSING MAGAZINE SEPTEMBER 2022
Multisatellite
Combination
* Data Consistency
* Data Complementarity
* Data Continuity
* Data Fusion
High-Accuracy
SWS Products
* Innovative Algorithms
for SWS Parameters
* Data Assimilation
IEEE Geoscience and Remote Sensing Magazine - September 2022

Table of Contents for the Digital Edition of IEEE Geoscience and Remote Sensing Magazine - September 2022
