IEEE Geoscience and Remote Sensing Magazine - September 2021 - 34

have outlived their design lifetimes yet still have a positive
impact on forecasting skills.
GEOPHYSICAL CAPABILITIES OF
THE CONSTELLATION
The overall capability of the EOSC can be estimated in
terms of the extent of the geophysical variables collectively
observed and their spatial/temporal/quality attributes, as
listed in Table 1. Atmospheric temperature, for instance, is
usually retrieved from satellites with good accuracy and has
good spatiotemporal coverage and resolution. Atmospheric
moisture, in contrast, is more difficult to characterize because
of its higher temporal and spatial variability and its
range over several orders of magnitude. Winds estimated
through the motion of atmospheric tracers such as clouds
are termed atmospheric motion vectors (AMVs). AMVs are
obtained with good temporal and horizontal resolution
but with few observations in the mid-troposphere (750 to
450 hPa) and, until recently, with fewer observations in the
range of latitudes from 50 to 70°.
A source of AMV error is that these observations are
treated as point measurements when they are really representative
of a layer average, and it is difficult to assign
heights to these observations. An additional, emerging
type of AMV is derived from time sequences of the moisture
imagery determined from the retrieved moisture profiles.
These profiles of AMVs provide 3D wind information and
have the potential to fill in some of the gaps left by cloud
AMVs. The new Aeolus DWL adds a number of directly observed
line-of-sight winds [30]. The improved global wind
data at all altitudes was the highest priority need identified
in the Space Platform Requirements Working Group study
for new observations [7].
For many applications, such as for soil moisture and
vegetation fraction, current polar-orbiting sensors provide
good spatiotemporal coverage, albeit at modest resolution.
However, other purposes, such as observing the full diurnal
cycle of surface temperature or monitoring wildfires, require
the high frequency of observations available from GEO.
Ocean variables change over longer time scales than do
atmospheric variables and thus have different sampling
and temporal refresh requirements. The relatively slowly
varying SST, sea-surface salinity, and ocean color are generally
observed at global scales and at a high-enough frequency
relative to their natural variability. However, the current
sensors lack sufficient temporal and horizontal resolution
in the societally important coastal areas. The retrieval of
salinity is a relatively new capability, but the quality of salinity
products is rapidly maturing [31]. The more quickly
changing ocean surface wind vector and wave height are
mainly inferred from active sensors (scatterometers and altimeters,
respectively) from polar-orbiting platforms, and
sometimes the International Space Station, with generally
good quality, spatial coverage, and horizontal resolution.
Given the limited number of these sensors, the temporal
refresh is not ideal. Also, the altimeter generally observes
34
a pencil-shaped, nadir-only beam and therefore provides
only a low density of measurements, although the planned
French-U.S. SWOT mission would offer wider-swath altimetry
in the future [32]. In addition to retrieving wind speed
and direction from scatterometers, ocean surface wind
speed is retrieved from several MW radiometers, including
SSMIS, WindSat, and AMSR-2. The WindSat sensor, which
operates at lower frequencies and with a polarimetric signal,
can also be used to retrieve surface wind direction.
Trace gases, ozone, and aerosols observations are generally
made using IR and UV sensors. For example, hyperspectral
IR sensors, primarily designed for atmospheric
temperature and moisture profiles, can also be used to estimate
an increasing set of trace gas concentrations (e.g., ammonia)
[33]. Generally, for these variables, only vertically
integrated or very low vertical-resolution observations are
possible. However, detailed vertical resolution can sometimes
be obtained with limb sounding or solar-occultation
techniques. These data are useful for aerosol studies (e.g.,
in [34] and [35]). Some chemical species present extreme
difficulties for observing systems because of their short lifetimes,
small concentrations, and diurnal and seasonal variability.
Because chemical species are advected by the wind,
an NWP model output is essential in driving chemistry,
aerosol, and air-quality models. In addition, a time history
of chemistry, aerosols, and air quality contains information
on the wind. Already, a number of species are being assimilated
into the ECMWF system [36]. Future applications
might have more stringent requirements on some of these
observables, such as sea-surface salinity or trace gases.
GEOPHYSICAL CAPABILITIES OF THE EOSC
Figure 3 presents the overall capabilities of the different
types of satellite sensors driving the collective capabilities
of the EOSC. Figure 3 aims at capturing, in only a simplified
and nonexhaustive way, the major components of the
space constellation in terms of sensor types. This figure
shows the relevance of each type of instrument to provide
information for the geophysical variables or, conversely, for
each geophysical variable, the instrument types that contribute
to making the relevant measurements.
Figure 3 was developed from the OSCAR Space database
[1] and is the authors' collective, but subjective, assessment.
It is noted that, in the actual OSCAR Space database, the
granularity of this information extends to more than 600
individual instruments (from April 1960 to the present day
as well as to planned instruments up to 2040) and to 124
geophysical variables. OSCAR assigns each instrument to
one of the instrument types given by the rows of Figure 3.
Each column corresponds to several specific geophysical
variables, which are termed mission objectives in OSCAR.
Note that these capabilities are in terms of information
" directly " derivable from the observations-that is, independent
of their impact in a DA system. For example, the
impact of RO observations on tropical winds through DA
is not included here.
IEEE GEOSCIENCE AND REMOTE SENSING MAGAZINE SEPTEMBER 2021
IEEE Geoscience and Remote Sensing Magazine - September 2021

Table of Contents for the Digital Edition of IEEE Geoscience and Remote Sensing Magazine - September 2021
