ORIGINAL PAPER Monitoring Continental Surface Waters by Satellite Altimetry Ste ´phane Calmant Fre ´de ´rique Seyler Jean Franc ¸ois Cretaux Received: 30 March 2008 / Accepted: 9 December 2008 / Published online: 24 January 2009 Ó Springer Science+Business Media B.V. 2009 Abstract The monitoring of continental water stages is a requirement for meeting human needs and assessing ongoing climatic changes. However, regular gauging networks fail to provide the information needed for spatial coverage and timely delivery. Although the space missions discussed here were not primarily dedicated to hydrology, 18 years of satellite altimetry have furnished complementary data that can be used to create hydro- logical products, such as time series of stages, estimated discharges of rivers or volume change of lakes, river altitude profiles or leveling of in situ stations. Raw data still suffer uncertainties of one to several decimeters. These require specific reprocessing such as waveform retracking or geophysical correction editing; much work still remains to be done. Besides, measuring the flow velocity appears feasible owing to SAR interferometer techniques. Inundated surfaces, and the time variations of their extent, are currently almost routinely computed using satellite imagery. Thus, the compilation of the continuous efforts of the scientific community in these various investigative directions, such as recording from space the discharges of rivers or the change in water volume stored in lakes, can be foreseen in the near future. Keywords Satellite altimetry Hydrology Rivers and lakes 1 Introduction The radar altimeters employed on altimetry mission satellites transmit a short microwave pulse in the nadir direction, and the echo reflected by the surface is examined. The time for the pulse to be reflected back to the altimeter corresponds to the distance (or ‘‘range’’) travelled by the electromagnetic pulse between the satellite and the Earth’s surface, S. Calmant (&) J. F. Cretaux LEGOS, UMR CNRS/UT3/IRD/CNES, Universite ´ Paul Sabatier, UT3, 31400 Toulouse, France e-mail: [email protected]F. Seyler LMTG, UMR CNRS/UT3/IRD, Universite ´ Paul Sabatier, UT3, 31400 Toulouse, France 123 Surv Geophys (2008) 29:247–269 DOI 10.1007/s10712-008-9051-1
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
ORI GIN AL PA PER
Monitoring Continental Surface Waters by SatelliteAltimetry
Stephane Calmant Æ Frederique Seyler Æ Jean Francois Cretaux
Received: 30 March 2008 / Accepted: 9 December 2008 / Published online: 24 January 2009� Springer Science+Business Media B.V. 2009
Abstract The monitoring of continental water stages is a requirement for meeting human
needs and assessing ongoing climatic changes. However, regular gauging networks fail to
provide the information needed for spatial coverage and timely delivery. Although the
space missions discussed here were not primarily dedicated to hydrology, 18 years of
satellite altimetry have furnished complementary data that can be used to create hydro-
logical products, such as time series of stages, estimated discharges of rivers or volume
change of lakes, river altitude profiles or leveling of in situ stations. Raw data still suffer
uncertainties of one to several decimeters. These require specific reprocessing such as
waveform retracking or geophysical correction editing; much work still remains to be
done. Besides, measuring the flow velocity appears feasible owing to SAR interferometer
techniques. Inundated surfaces, and the time variations of their extent, are currently almost
routinely computed using satellite imagery. Thus, the compilation of the continuous efforts
of the scientific community in these various investigative directions, such as recording
from space the discharges of rivers or the change in water volume stored in lakes, can be
foreseen in the near future.
Keywords Satellite altimetry � Hydrology � Rivers and lakes
1 Introduction
The radar altimeters employed on altimetry mission satellites transmit a short microwave
pulse in the nadir direction, and the echo reflected by the surface is examined. The time for
the pulse to be reflected back to the altimeter corresponds to the distance (or ‘‘range’’)
travelled by the electromagnetic pulse between the satellite and the Earth’s surface,
S. Calmant (&) � J. F. CretauxLEGOS, UMR CNRS/UT3/IRD/CNES, Universite Paul Sabatier, UT3, 31400 Toulouse, Francee-mail: [email protected]
F. SeylerLMTG, UMR CNRS/UT3/IRD, Universite Paul Sabatier, UT3, 31400 Toulouse, France
Regular ‘‘ocean-type’’ trackers expect long tail shapes of energy distribution (view A in
Fig. 3) whereas the echoes bouncing off rivers are often specular (Guzkowska et al. 1990)
or a combination of specular echoes (view B in Fig. 3). Thus, in the best case, energy is
received but the range estimate is erroneous or not estimated (views C and D in Fig. 3); in
the worst case, the altimeter loses tracking and subsequent echoes are lost. The antenna-
reflector range is determined by fitting the waveform with a predefined analytical function
(called waveform ‘‘tracking’’). If the analytical function is not well suited to the waveform
shape, this tracking leads to wrong estimates of the height value (or even no estimate at
all). It is worth noting that the radar altimetry data collected by the ongoing ENVISAT
mission are nominally retracked with four algorithms. In turn, retracking the radar
waveforms collected by the ERS 1 & 2, T/P, Jason, and GFO missions requires that some
retracking procedures be conducted. Berry et al. (2005) showed that, by retracking multi-
mission altimeter data with tuned filters, high quality height data can be obtained from the
vast majority of lakes with surface areas greater than 500 km2. Results from the Amazon
Basin also show the great improvement in both quantity and quality of data obtained. The
retracking of the T/P data with ENVISAT algorithms that has been performed by CLS, a
subsidiary of the French Space agency CNES (Centre National d’Etudes Spatiales), for the
CASH project has provided promising results in terms of accuracy improvement and
recovery of data that are missing in the Topex/Poseidon data distributed to users as Merged
Geophysical Data Records (MGDRs, Mercier and Zanife 2006). More in-depth informa-
tion about this key point of tracking algorithms can be found in Calmant and Seyler (2006)
and the references given there.
The second issue is the problem of off-nadir reflections, in particular the hooking effect.
This effect occurs when the ground area illuminated by the radar beam at the nadir of the
Fig. 1 Position of the pulse relative to the reflecting surface and energy received by the satellite. Thevertical axis is positioned at the time of emission of the pulse. The dotted line stands for the theoretical timegiving the correct two-way travel time to be used for the range (half amplitude of the peak)
252 Surv Geophys (2008) 29:247–269
123
antenna is bouncing back little energy in comparison to the energy bounced back by a
water surface at the edge of the radar ground footprint. Thus, the latter is dominant in the
echo returned although lessened by the antenna pattern. The tracking procedure will then
determine the range by fitting this return of energy that traversed a skewed path. Because
the range estimate assumes that the target is at the satellite nadir, this leads to an over-
estimated range, i.e., to an underestimate of the height of the reflecting water surface. This
is not uncommon in continental waters and has to be carefully taken into account because
these measurements can be erroneously interpreted (Frappart et al. 2006a, b). The
geometry of such a mis-measurement is depicted in Fig. 4 and an example of a real case is
given in Fig. 5.
The third geometrical effect impacting accuracy is the slope effect. Indeed, within a
radar footprint a few kilometers wide, the river height can change significantly. This
changes the waveform shape and affects range determination. Also, in braided reaches,
branches may have different heights, which will spread the return time of energy and
contribute to an erroneous range determination.
The width of the radar beam of ocean-oriented missions is several kilometers since a
surface average is needed over oceans to reduce as much as possible height measurement
corruption due to wind driven waves. The drawback for hydrology studies is that echoes
over rivers whose width is less than, say, 1 km—i.e., most of them—are polluted by energy
Fig. 2 Samples of Topex waveforms over rivers in the Amazon basin. a ‘‘Ocean-like’’ waveform processedby the onboard tracker, e.g., the algorithm implemented onboard the satellite to provide a range estimate.b Multi-peak waveform rejected by the tracker. c Specular waveform processed by the tracker. d Specularwaveform, similar to (c), rejected by the tracker
Surv Geophys (2008) 29:247–269 253
123
reflected by riverbanks or islets. The heterogeneity of the reflecting surface is estimated to
be the first cause of error in terms of magnitude (Rosmorduc et al. 2006).
Lastly, the air density, the amount of water vapor and the electrons in the ionosphere
modify the travel time of radar waves throughout the atmosphere. The electron content in
the ionosphere and air pressure are given by independent datasets. However, the amount of
water vapor is estimated using microwave radiometers employed onboard together with the
radar altimeters. Current microwave radiometers fail to estimate the atmospheric content of
water vapor over continents because the signature of the atmospheric water vapor is mixed
with that of ground wetness. Thus, this effect, ranging from a few centimeters to some tens
of centimeters, cannot be accurately corrected for over land waters and corrections can
only be estimated from large scale global datasets such as ECMWF or NCEP. Yet, these
Fig. 3 Waveforms over different water bodies: an example of waveforms over oceanic surfaces—showinga high similarity along the track—is presented in the left-hand figure. An example of waveforms collectedover the Amazon basin, at the confluence of the Solimoes (waters in blue in the false color Landsat image)and Negro river (waters in black in the Landsat image) is presented in the right-hand figure. This examplehighlights the high variability in the shape of the waveforms making necessary that echo retracking isperformed on a case basis. Redrawn after Mercier and Zanife (2006)
Fig. 4 Geometry of the hooking effect. When the satellite is at location A on its orbit, i.e., not yet flyingover the water body, radar energy reaches the water body on the borders of the footprint. If the energybounced back by this water body dominates the total energy returned to the satellite antenna (i.e., when theground at the nadir is poorly reflecting), the retracking procedure will estimate the two-way travel time byfitting this peak of energy, estimating a slant distance (h0) instead of a nadir one (h). The true height, i.e., theone at the nadir will be obtained only when the satellite is a location B on its orbit. The resulting along-trackheights exhibit a characteristic parabola pattern similar to the one presented in Fig. 5
254 Surv Geophys (2008) 29:247–269
123
model-based corrections—although satisfactory in most cases—can suffer from large
errors, in particular in the case of mountain lakes. Mercier and Zanife (2006) have shown
that the errors due to changes in the altitude of the reflecting surface (and thus the thickness
of the atmosphere column) are not taken into account. They also demonstrated that a
computation of the atmospheric corrections such as the DTC based on the use of a Digital
Elevation Model to estimate the altitude of the reflecting surface was not adequate. They
proposed a method where the altitude of the reflecting surface is deduced from the alti-
metric measurement itself. Cretaux et al. (2008) have shown in the case of Lake Issykul-
kul that the Dry Tropospheric Corrections delivered by the T/P, ENVISAT, JASON and
GFO Geophysical Data Records (GDRs) were wrong (Fig. 6), either because the altitude of
the lake was not accounted for in the computation of the reference air pressure or because
this altitude varies erroneously within the Lake surface in the DTM used for the compu-
tation of this reference air pressure, producing an artificial variation of the correction along
the tracks crossing the lake, and hence erroneous height variations over the lake surface. In
this case, the largest difference between corrections on a given day exceeds 60 cm. Such an
error is a major issue when merging water stages derived from different altimetry missions
in order to increase the sampling frequency.
From Eq. (1), it is clear that the quality of the orbit directly enters into the error budget
of the water height. Yet, orbit modeling has dramatically improved over the last decade
(Fig. 7). Thus, in view of the aforementioned error sources, this error can be considered of
less importance for continental waters. This fact is used by web services such as
Cropexplorer to propose NRT preliminary values of water height using interim orbits,
given that the water height finally computed with a more refined orbit will not differ
significantly in regard to the overall uncertainties.
Fig. 5 Example of the hooking effect in the altimetry data (ENVISAT) over a small river in the Amazonbasin (Igarape Nelson Pinheiro, Rio Negro sub-basin). Left view (a) Map of the crossing betweenN. Pinheiro river and ENVISAT track (blue dots). The green polygon stands for the window from whichmeasurements were extracted to be plotted in the right figure. The river appears in white in the JERS mosaic,i.e., highly energetic echoes, since the water echoes are coherently double bounced off of the tree trunkssurrounding the river. Right view (b) cross section displaying where the ENVISAT passes (each black line)over the river and adjacent banks. Note that the capture of echoes bounced by the river surface before andafter the satellite passes over the river itself (green polygon) turns into a parabolic apparent reflector underthe banks on both sides of the river. The data within the green polygon are marked as red dots in the left-hand view
Surv Geophys (2008) 29:247–269 255
123
Several studies have dealt with the estimation of the error budget in the altimetric series
of continental water stages. In particular, Birkett et al. (2002) compared T/P altimetric
series with in situ series. Discrepancies up to several meters have been reported in this
study. However, it should be pointed out that such comparisons suffer severe limitations.
On the one hand, the satellite track usually fails to cross the river directly over a gauging
station. Comparing both series assumes that the water stage varies identically in both
places, but this is often not true. Indeed, water stage variations along a reach are affected
by changes in the cross section geometry. Also, flow routing should be performed to
account for the delay from the location of the gauge and that of the satellite track when the
distance between both locations is significant with respect to flow velocity and sampling
rate. Typically, the delay is 1 day—e.g., the sampling rate of gauging stations—for a flow
to cover 85 km with a 1 m/s wave. Given that water stages tend to vary by several
Fig. 6 Model-derived dry tropospheric correction at lake Issyk-kul (After Cretaux et al. 2008). The pinkband stands for the range of DTC expected from the records of a meteorological station on the lake shore
Fig. 7 Improvement in orbit error from GEOS-3 in 1975 to JASON and ENVISAT in 2002
256 Surv Geophys (2008) 29:247–269
123
centimeters or decimeters within a day, the error induced when the delay is not taken into
account can contribute significantly to the discrepancy between both series. On the other
hand, when the gauge is not leveled, scatter between altimetric and in situ series is
computed with reduced—zero mean series, and the resulting estimate of the discrepancy
does not take into account possible biases, such as the tracker-dependent bias. Roux et al.
(2008) investigated ways of estimating errors in virtual stations through the use of inter-
polated series, i.e., using more than one in situ gauging station and taking into account the
propagation time between the in situ and virtual gauges. They found that the random error
is roughly at the decimeter level except for low water, where the error can be *1 m or
more. This result is also consistent with the results presented by Bercher et al. (2006).
Alternatively, the accuracy of altimetric measurements can be assessed by means of a
comparison with GPS ground truth height measurements right under the satellite track.
Although this technique is widely used to assess the accuracy of altimetric data over
oceanic surfaces (Bonnefond et al. 2003, among others), this kind of fieldwork has seldom
been conducted over rivers and the results published to date, as in Frappart et al. (2006a,
b), are too limited to allow definite conclusions to be drawn as to the determination of the
error budget in altimetric series using this approach. Although intrinsic system errors of
radar measurements are the same over land waters and oceans, e.g., centimetric accuracy,
the overall uncertainty of altimetric measurements over continental waters is now a couple
of decimeters (Birkett et al. 2002; Frappart et al. 2006a, b).
6 Temporal Sampling
The time step of the altimetric series is given by the orbit repeat period. For the current
radar altimetric missions, this period ranges from 10 days for T/P and Jason to 35 days for
ERS-2 and ENVISAT. GFO has an intermediate repeat period of 17 days. In terms of
sampling rate, this is much lower than the time step of in situ measurements usually
collected once or twice a day, or even much more often, every 15 min in automated
networks of developed countries. Bercher et al. (2006) evaluated the amount of informa-
tion lost due to this under-sampling.
To improve this time step, multi-mission series can be constructed. This is often pos-
sible over large lakes that are frequently crossed by several tracks and/or several missions.
For example, JASON, ENVISAT and GFO altogether have overflown Lake Victoria about
20 times each month since 2002 (Fig. 8). Moreover, stage variations of most lake levels do
not vary significantly on a daily basis. Thus, taking into account all the available passes
should lead to a highly satisfactory time sampling of some dozens of the world’s largest
lakes. As far as rivers are concerned, the situation is more critical. In some ‘‘fortunate’’
places where satellite tracks intersect over a river reach, a multi-track ‘‘virtual station’’ can
be created. An example of such a series is given in Fig. 9. These combined series raise the
issue of the type of errors in the different data series. Indeed, the error budget of each
mission includes biases that must be accounted for prior to combining the data from
different missions.
Another method (Fig. 10) consists of interpolating satellite-based altimetry based on a
linear model exploiting data at a limited number of in situ limnimetric stations in order to
obtain time-series with a 1-day sampling period (Roux et al. 2008). The precision of
interpolation has been investigated in terms of methodology, sensitivity to model rapid
stage variations, robustness with regard to missing values and the effect of random noise.
An optimisation method based on a multi-objective criterion (OPT) furnishes the best
Surv Geophys (2008) 29:247–269 257
123
absolute results, as it is the method that is least sensitive to missing values and random
noise, two factors that systematically affect radar altimetry data. Results also show that
taking into account more than one in situ reference station significantly decreases the RMS
errors in the predicted stages. Taking into account time shifts between stations improves
the results too.
Fig. 8 a Satellite tracks crosscutting the surface of Lake Victoria, including GFO, T/P and JASON,ENVISAT and ICEsat. The color-coded surface in the background are the ellipsoidal height variation due togravity used to separate the geographical and temporal variations in the altimetry measurements along thesatellite tracks. b Temporal variations. The color coding of the dots is as follows: GFO in blue, T/P in pink,JASON-1 in black, ENVISAT in red, ICEsat in yellow. The green line stands for in situ measurements
Fig. 9 Example of multi-satellite time series at a crossover formed by the tracks of the ERS and T/Pmissions over the Rio Negro. Note that the three independent series have been adjusted for their relativebiases in order to produce a consistent series
258 Surv Geophys (2008) 29:247–269
123
7 Applications
A range of applications derived from the altimetric measurements of continental water
levels has been performed or is currently underway. A brief review of these applications is
now given.
7.1 Levelling of Hydrological Network Gauge Stations
A number of great river basins are located in remote geographical areas. These basins
include hydrological gauges, where topographic leveling has not been carried out or, in
some cases, leveling uncertainty is too high due to the difficulties in conducting the
conventional processes of terrestrial—spirit—leveling. By way of an example, for the
Amazon basin, most hydrological gauges from the ANA network (Brazilian National
Agency for Water) are located outside the topographic leveling routes of the IBGE
(Brazilian Institute for Geography and Statistics). Additionally, gauge stations within the
hydrographic basin of the Amazon and in the neighboring countries of Brazil are all
unleveled, apart from the Iquitos station in Peru. This is a major drawback for hydrody-
namical modeling of the basin, since these kinds of models require that the hydrographic
parameters of the river, such as bed slope, be entered into the model using a common
altitudinal reference. Cauhope (2004) leveled gauges in the Curuai Varzea and the adjacent
Amazon reach using ENVISAT time series. For the Tapajos River, a tributary of the
Amazon River, Calmant and Seyler (2004) showed that ICESat measurements are par-
ticularly well-suited for gauge leveling. Kosuth et al. (2006) established altimetric
levelling from Topex Poseidon for 97 hydrometric stations along 27,740 km of the
Fig. 10 Example of time series at an ENVISAT virtual station. The daily values are predicted using boththe ENVISAT series and readings at remote stations
Surv Geophys (2008) 29:247–269 259
123
Amazon hydrographic network. Validation has been undertaken for 23 stations, comparing
altimetric data with leveling values obtained from bi-frequency GPS positioning. However,
leveling of other gauge networks worldwide still has to be conducted.
7.2 Estimation of Discharge from Stage Altimetric Measurements
Dense stage and discharge estimations have many uses. For example, discharge values are
essential for water management, extreme flow prediction, and hydrological modeling. Rain
gauges within remote watersheds are often sparsely distributed, although the spatially
averaged estimation of rain is an important parameter for Global Climate Models. As
previously pointed out, a better distribution of the integrated response of the basin to the
incoming rain would allow a better validation of these models. Better-distributed discharge
values could also be used to constrain models of weathering processes and carbon flux
estimations.
Several works have examined the ability of spatial data such as altimetry or imagery to
retrieve river discharge. Recently, Bjerklie et al. (2005) has estimated in-bank river dis-
charge on the basis of hydraulic relationships constrained with remotely sensed width
information and channel slope obtained from topographic maps. Coe and Birkett (2004)
estimated the mean monthly discharge of the Chari River at N’Djamena, Chad. They used
T/P surface water stages upstream from the gauging station calibrated with the ground-
based gauge height and discharge data using simple empirical regression techniques.
Kouraev et al. (2004) estimated discharge for the Ob river (Siberia) along two T/P tracks
crosscutting the river in the vicinity of the Salekhard gauging station. T/P measurements
were found to provide reliable water level (H) time series that could then be used to
estimate water discharge (Q) from the rating curve between H and Q at Salekhard, located
65–70 km away from the T/P tracks. Zakharova et al. (2006) used the same method to
derive discharge at T/P virtual stations of the Amazon main stream. These studies suggest
that remotely sensed river hydraulic data could be used to estimate the discharge at a
specific location directly, if nearby ground-based discharge measurements are used to
develop discharge ratings in conjunction with the remotely observed variable(s). As
pointed out by Bjerklie et al. (2005), discharge ratings developed from ground-based flow
measurements and remotely sensed hydraulic information are site specific. Leon et al.
(2008) developed a model based on a diffusion-cum-dynamic wave propagation assump-
tion, using in situ discharges and radar altimetry data to estimate rating curves at the
satellite track crossings in the Negro River Basin, Amazon. The calibration phase led to
differences of less than 4% between measured and estimated outflows and validation has
yielded less than 10% errors.
By estimating discharges by combining altimetric water stage and remote in situ dis-
charges, denser stage-discharge rating curves within a basin can be obtained. For example,
out of the 571 gauges listed by ANA, 46 are located in the Negro River sub-basin and 25
have complete records covering the last 20 years. Along the T/P tracks, water level time
series were built for 88 T/P crossings with river and floodplains (Frappart et al. 2005).
Based on altimetric heights, 3.5 times more measurement points are available in the Negro
River basin. These points are evenly distributed within the basin, making possible a
regionalization of the water fluxes. Yet, it is worth noting that these rating curves share
with the in situ curves the need for regular updates to account for possible changes in the
hydrological characteristics of the stem.
In addition, SAR interferometry appears to be the most promising technique to retrieve
surface flow velocities (Goldstein and Zebker 1987; Romeiser and Runge 2007; Romeiser
260 Surv Geophys (2008) 29:247–269
123
et al. 2007 among others). Yet, this technique still suffers from technical limitations, such
as being able to estimate the velocity component only in the line-of-sight direction. Thus,
under the assumption that the velocity vector is mostly parallel to the river channel, and
that a global coverage implies that the spacecraft orbits the Earth at large inclination, the
flow velocity in many East–West trending river segments is likely to be poorly resolved.
7.3 Estimation of Spatial and Temporal Variations of Water Storage: Rivers
and Wetlands
As pointed out by Alsdorf et al. (2003), for the past 100 years our understanding of the
hydraulic characteristics and hydrological mass-balances of surface water runoff have
largely been derived from discharge measurements at in-channel gauging stations. Mea-
surement of in-channel discharge unfortunately does not provide the information necessary
for understanding flow and storage in off-river-channel environments, such as wetlands,
floodplains, and anabranches (e.g., braided channels); these environments are increasingly
recognized for their importance in the biogeochemical cycling of waterborne constituents.
Using radar interferometry, Alsdorf et al. (2001) estimated water height changes in an
Amazon lake to be about 12 ± 2.4 cm, and a volume change of 280 106 m3 during the
44 days between the two JERS images used. These authors reviewed a T/P crossing of the
lake and estimated the change in water stage to be of 20 ± 10 cm during the same period.
The T/P nadir measurements might appear less accurate than interferometry but, so far,
only these measurements offer 10-day periodic estimates over more than a decade. In an
extensive study of T/P measurements over the Amazon basin, Birkett et al. (2002) have
successfully distinguished rivers from floodplains in a number of cases. A small phase
offset of a few days in stage variations between river and nearby floodplain has occa-
sionally been observed.
Frappart et al. (2005), determined spatio-temporal variations of water volume over the
main stream and floodplain located in the Negro River basin, using area variation estimates
for a seasonal cycle captured by the Synthetic Aperture Radar (SAR) onboard the Japanese
Earth Resources Satellite (JERS-1), and changes in water level from the T/P altimetry,
combined with in situ hydrographic stations. Similarly, Frappart et al. (2006a, b) moni-
tored the flood propagation along the Mekong River by combining satellite altimetry data
and imagery. A volume variation of 331 km3 was estimated for the whole Negro sub-basin,
enhancing the complex relationship between the volume potentially stored, the inundated
area and the volume flow during the same period. Altimetry data is useful to study the
hydrological cycle of rivers. Also using the T/P data, Maheu et al. (2003) mapped the flood
propagation along strike the La Plata basin, from the Pantanal wetland to the mouth in the
South Atlantic Ocean.
7.4 Water Profiles and Geodynamical Implications
River free surface slope is an important parameter in floodwave propagation models and
sediment transport calculations. For example, slope values of a few centimeters per kilo-
meter were evaluated for the Amazon main stem using barometric estimates of elevation
performed at some gauging stations (Salati and Marques 1984; Sioli 1984; Nordin and
Meade 1986; Meade et al. 1991). Guzkowska et al. (1990) and Cudlip et al. (1992) used 15
of the 32 crossings of the Amazon River by the SEASAT altimeter to provide an estimate
of the elevation profile of the Amazon, whereas Mertes et al. (1996) and Dunne et al.
(1998) used the SEASAT falling stage measurements to calculate 14 gradient values for
Surv Geophys (2008) 29:247–269 261
123
the Amazon main stem. Birkett et al. (2002) used T/P measurements along the main stem
of the Amazon to estimate the spatial and temporal variation of the gradient values.
The slope of the riverbed is an important parameter for modeling river hydrodynamics.
Recently, Leon et al. (2006a) have proposed a methodology to derive stream profiles from
the riverbed’s height and slope. This method takes advantage of the fact that altimetry data
all have a common reference, whereas in situ measurements are only referred to a local
origin, united from one gauge to the other. The height of the river bed at virtual stations is
determined as that height at which discharge vanishes in rating curves established by
combining times series of water height by satellite altimetry with discharges predicted by
routing the flow recorded at remote gauges. To give an example of the potential of
altimetric data to provide river slopes, even for rivers located in semi-arid regions, the
slope of the Godavari river is presented in Fig. 11. The Godavari river is a sacred river that
runs through India. It originates at 1,620 m above mean sea level, runs eastwards for about
1,500 km and empties into the Bay of Bengal. It runs through the Godavari graben, within
steep banks in the upper part of its course. With a mean annual discharge of 3,200 m3/s
(Singh and Swamy 2006) that ranges far below the Amazon basin with its 209,000 m3/s
mean annual discharge, the Godavari river is also an example of the capabilities of satellite
altimetry in retrieving heights for medium-size rivers, at least partly.
7.5 Hydrologic Regime of Ungauged or Poorly Gauged Basins
A very promising application of radar altimetry concerns the characterization of hydrologic
regime of poorly gauged tropical rivers. Leon et al. (2006b) have studied the 200,000 km2
basin of Rio Caqueta, the most important river basin within Colombian Amazonia. Over 32
gauging stations installed in the Rio Caqueta basin, only six stations are active today. This
point is mainly due to political problems in the FARC controlled region. Only one station,
namely Vila Betancourt, near the border of Brazil, has a full history of discharge esti-
mations in the period matching that of the altimetric data. Along the 1,270 km length, 13
virtual stations have been determined, 12 for ENVISAT crossings and only one with T/P.
Together with the river stages along the river, the discharge at the virtual station has been
estimated by hydrodynamic modelling. Hence, stage discharge relationships have been
computed, leading also to estimates of the river depth at the virtual station and to the slope
of the free surface river between virtual stations. The Caqueta river is characterized by a
hydrological regime highly variable in space and time, and an average width of about
Fig. 11 Example of height profile determined by satellite altimetry for a non-levelled river (Godavari,India)
262 Surv Geophys (2008) 29:247–269
123
1.1 km at high stage. Despite these factors that impact the quality of the altimetric data,
discharges have been estimated with a mean error of only 15% and the river depth with a
mean difference of less than 1.1 m. Moreover, a 15% mean error in discharge estimates is
interesting since discharge computed using in situ measurements often has a worse
accuracy since it is extrapolated from rating curves.
7.6 Monitoring Lakes for Climate Change Assessment and the Impact of Human
Activity
The monitoring of levels by satellite altimetry is much better for lakes than it is for rivers.
As stated previously, lake surfaces can be cross cut by several missions and, if large
enough, by several tracks of some missions. Merging these measurements together enables
one to produce high sampling rate time series of the level of the Lake, as displayed in
Fig. 8 for the case of Lake Victoria. Such a procedure requires that the curvature of the
lake surface be computed beforehand in order to have a common height reference all over
the surface of the lake (Fig. 8). In the present case of Lake Victoria, the rms difference
between the altimetry series and the in situ series is 8 cm, after the mean of each series has
been removed.
The assessment of the lake water balance could provide improved knowledge of
regional and global climate change and a quantification of the human stress on water
resources across all continents, as studied by Birkett (2000) in the case of Lake Chad and
adjacent wetlands or by Mercier et al. (2001) and Birkett et al. (1999) to highlight the links
between the levels in the African Great Lakes and oceanic climatic indices. Hostetler
(1995) noted that deep and steep sided lakes are good proxies for high amplitude-low
frequency changes, while shallow water basins are better targets for rapid-low amplitude
changes. Central Asia is a place where many lakes of both types, e.g., large bodies in flat
areas such as the Aral sea and narrow bodies in steep mountainous areas such as the
Karakul and Togtogul lakes, are encountered. Very few of these lakes are monitored with
in situ data. However, many of them can be monitored by satellite altimetry (Fig. 12).
Shallow lakes are also extremely sensitive for revealing decreased water input—pos-
sibly of human origin—and rising evaporation. The Aral sea is typical of such a case and it
has been extensively studied, including using altimetry (Cretaux et al. 2005). The Aral sea
is located in an arid zone characterised by marked differences between summer and winter
temperatures and has low precipitation all year round. Evaporation is approximately ten
times greater than precipitation and the sea being maintained at equilibrium by the
inflowing waters of the Amu Darya and Syr Darya. Around 1960, a decision was made to
develop an intensive cotton and rice economy in its vicinity. In such an arid zone, irrigation
provided the means to reach the planned agricultural objectives of the Soviet Union
government. Large-scale development of ground infrastructure (irrigation channels, res-
ervoirs) started in the 1960s and the volume of water utilised for irrigation increased to
around 100 km3/year, overtaking the annual Amu Daria and Syr Daria inflows. As a result,
the level of the Aral Sea dropped by 13 m between 1960 and 1989 (Fig. 13). This led to the
split of the Aral sea into so-called Small Aral in the north and Big Aral in the south.
Satellite altimetry tracks crosscut both lakes and have measured the level variations since
1992. During that period, Big Aral shrank at a rate of 60–80 cm/year (Cretaux et al. 2005).
The corresponding decrease in surface extent and volume was 67,000 km2 and 1,083 km3
in 1960 (Bortnik 1999) to 16,000 km2 and 100 km3 in 2004 (Cretaux et al. 2005). The
difference between evaporation and precipitation for Big Aral represents an average loss of
25–30 km3/year over the last decade.
Surv Geophys (2008) 29:247–269 263
123
Cretaux et al. (2005) also showed that the reduction of lake volume compared to the
hydrological budget deduced from gauges implied an underground water inflow of
5 ± 3 km3/year to the Big Aral. During the last 2 years, the level and volume of the Big
Aral has continuously shrunk with a rate as high as about 10 km3/year in water loss and
1 m/year in decrease of the sea level (Fig. 13). Since 1989, the Small Aral experienced a
different situation since it has continued to be fed by the Syr Darya River and therefore has
Fig. 12 Central Asian lakes and reservoirs monitored by satellite altimetry (http://www.legos.obs-mip.fr/soa/hydrologie/hydroweb/). The image is from a mosaic of Landsat images taken in 1990
Fig. 13 Big Aral sea volume variations taken from in situ measurements and from altimetry data (afterCretaux et al. 2005)
dried up less than the Big Aral (Fig. 14). There are two main explanations: first, the area of
the basin is much smaller and the effect of evaporation is reduced, and, secondly, during
the years 1992 to 1999 a dam was built in the Berg’s strait in order to separate both lakes
and stop the loss of water from the Syr Darya into the desert. This dam was destroyed (and
rebuilt) three times during this period. Aladin et al. (2005) demonstrated that during the
period of 1993 to 1999 the existence of the dam contributed to the restoration of the Small
Aral. In August 2005, a solid dam, funded by World Bank, has been built in the Berg’s
Strait separating the Small and the Big Aral. It induced a 2-m increase in the sea level of
Small Aral since then and, every spring, the release of some water through open gate in the
dam has allowed the regulation the level of the Small Aral to an approximate level of 42 m
(Fig. 13).
8 Conclusions
Altimetric data offer many possibilities for the monitoring of continental waters. Among
them is the ability of tracking either in-channel fluxes or flow and storage in off-river-
channel environments, such as wetlands, floodplains, and anabranches. Also, that all the
series are naturally leveled is a great benefit of satellite altimetry with respect to in situ
gauges. These are critical for surface water balance. The development of methods to
estimate the river discharge using remotely sensed data would provide the means to
increase the streamflow measurement network globally. Typically, in situ data collection
and management activities are undertaken at the national level, where there is a need for
regionally coordinated systems and actions. Worldwide coverage and near-real time
availability of measurement is definitely another major benefit of satellite altimetry with
respect to local measurements.
Fig. 14 Variation of volume of small Aral taken from satellite altimetry (black stars) and from hydrologicalin situ measurements (orange triangles)
Surv Geophys (2008) 29:247–269 265
123
The major drawback in the use of altimetric height for water stage monitoring is the
temporal sampling rate. Clearly, the 10-day period of T/P and Jason and the 35-day period
for ENVISAT cannot compete with observations made daily or twice a day at most gauges
around the world. In some applications like flood events, even more frequent data, every
few minutes, are required. On the other hand, over lakes that can be overflown by several
missions and several tracks of each mission, the time sampling can be much better, up to
three to four times per week and then compare with the frequency of in situ sampling.
Another limitation is measurement uncertainty that has been seen to vary from a few
centimeters to some meters or more in the worst cases. With respect to radar altimetry,
accuracy can to a certain extent be improved by adapting the processing of echo wave-
forms to the continental case. This uncertainty is due to the ground point target size which
ranges from kilometers for T/P to a few hundred meters for ENVISAT and 70 m for
ICESat. Roughly speaking, uncertainty decreases with the size of the footprint, as the
combination of water and vegetation or the merging of different water bodies in a single
footprint becomes less likely. When it comes to identifying and separating peaks of energy
reflected by small water bodies, the sensor and echo processing capability becomes a major
issue.
Higher resolution is foreseen for the future missions. It should reach a few hundred
meters for AltiKa, since the footprint in the Ka band radar is smaller than it is in the Ku band
and Cryosat-2 owing to footprint slicing by virtue of interferometry. Lastly, again owing to
interferometry techniques, a full coverage of the continental domain, a resolution of a few
tens of meters and a slope accuracy of 10 lrad along river courses are expected with Surface
Water Ocean Topography (SWOT), the first mission specifically dedicated to the moni-
toring of continental waters (SWOT Homepage: www.geology.ohio-state.edu/water).
The prospects offered by future missions will definitely enhance the ability of spatial
data to be included in models and change the management and monitoring of water
resources. Clearly, satellite altimetry is not likely to replace in situ measurements in the
near future, but a combination of both systems will certainly enhance our ability to monitor
the cycle of surface water from the regional scale to the global scale.
Acknowledgments The authors thank P. Bates, an anonymous reviewer and the Editor for their in-depthreviews that greatly helped to improve this article.
References
Agencia National das Aguas. http://hidroweb.ana.gov.br/doc/WRMB/part1.htm. Accessed Nov 2008Aladin NV, Cretaux J-F, Plotnikov IS, Kouraev AV, Smurov AO, Cazenave A, Egorov AN, Papa F (2005)
Modern hydro-biological state of the small Aral sea. Environmetric 16:1–18. doi:10.1002/env.709Alsdorf DE, Birkett CM, Dunne T, Melack J, Hess L (2001) Water level changes in a large Amazon lake
measured with spaceborne radar interferometry and altimetry. Geophys Res Lett 28(14):2671–2674Alsdorf D, Lettenmaier D, Vorosmarty C et al (2003) The need for global, satellite-based observations of
terrestrial surface waters. EOS Trans 84(29):269–276Bercher N, Kosuth P, Bruniquel J (2006) Quality of river water level time series issued from satellite radar
altimetry: influence of river hydrology and satellite measurement accuracy and frequency. Presented atEGU General Assembly, Vienna, April 2006
Berry PAM, Garlick JD, Freeman JA, Mathers EL (2005) Global inland water monitoring from multi-mission altimetry. Geophys Res Lett 32:L16401. doi:10.1029/2005GL022814
Birkett CM (1995a) The contribution of Topex/Poseidon to the global monitoring of climatically sensitivelakes. J Geophys Res 100(C12):25179–25204
Birkett CM (1995b) The global remote sensing of lakes, wetlands and rivers for hydrological and climateresearch. Geoscience and Remote Sensing Symposium, IGARSS’ 95, Quantitative Remote Sensing forScience and Applications, vol 3. pp 1979–1981
Birkett CM (1998) Contribution of the Topex NASA radar altimeter to the global monitoring of large riversand wetlands. Water Resour Res 34(5):1223–1239
Birkett CM (2000) Synergistic remote sensing of Lake Chad: variability of basin inundation. Remote SensEnviron 72:218–236
Birkett CM, Murtugudde R, Allan T (1999) Indian Ocean climate event brings floods to east Africa’s lakesand the Sudd marsh. Geophys Res Lett 26:1031–1034
Birkett CM, Mertes LAK, Dunne T, Costa M, Jasinski J (2002) Altimetric remote sensing of the Amazon:application of satellite radar altimetry. J Geophys Res 107(D20):8059. doi:10.1029/2001JD000609
Bjerklie DM, Moller D, Smith LC, Dingman SL (2005) Estimating discharge in rivers using remotely sensedhydraulic information. J Hydrol 309:191–209
Bonnefond P, Exertier P, Laurain O, Menard Y, Orsoni A, Jan G, Jansou E (2003) Absolute calibration ofJason-1 and TOPEX/Poseidon altimeters in Corsica. Mar Geod 26:261–284
Bortnik VN (1999) Alteration of water level and salinity of the Aral sea, creeping environmental problemsand sustainable development in the Aral sea basin. Cambridge University Press, Cambridge, UK, pp47–65
Brooks RL (1982) Lake elevation from satellite radar altimetry from a validation area in Canada. Report,Geoscience Research Corporation, Salibury, Maryland, USA
Brown GS (1977) Skylab S-193 radar experiment analysis and results. NASA CR-2763, February 1977Calmant S, Seyler F (2004) Tapajos hydraulic slope at the confluence with the Amazon from combined
satellite altimetric data. EGU, Nice, April 2004Calmant S, Seyler F (2006) Continental surface waters from satellite altimetry. C R Geosciences 338:1113–
1122Cauhope M (2004) Hauteurs d’eau d’une plaine d’inondation amazonienne par altimetrie spatiale. Rapport
de stage de DEA Sciences de la Terre et l’Environnement. IMFT, Toulouse, France, p 30 (in French)Cazenave A, Bonnefond P, DoMinh K (1997) Caspian sea level from Topex/Poseidon altimetry: level now
falling. Geophys Res Lett 24:881–884Coe MT (2000) Modeling terrestrial hydrological systems at the continental scale: testing the accuracy of an
atmospheric GCM. J Climatol 13:686–704Coe MT, Birkett CM (2004) Calculation of river discharge and prediction of lake height from satellite radar
altimetry: example for the Lake Chad basin. Water Resour Res 40:W10205. doi:10.1029/2003WR002543Cretaux J-F, Kouraev AV, Papa F, Berge Nguyen M, Cazenave A, Aladin NV, Plotnikov IS (2005) Water
balance of the big Aral sea from satellite remote sensing and in situ observations. J Great Lakes Res31(4):520–534
Cretaux J-F, Calmant S, Romanovski V, Shibuyin A, Lyard F, Berge-Nguyen M, Cazenave A, Hernandez F(2008) Implementation of a new absolute calibration site for radar altimeter in the continental area:lake Issykkul in Central Asia. J Geod (in press)
Cudlip W, Ridley JK, Rapley CG (1992) The use of satellite radar altimetry for monitoring wetlands. In:Remote sensing and global change. Proceedings of the 16th annual conference of Remote SensingSociety, London, UK, pp 207–216
De Oliveira Campos I, Mercier F, Maheu C, Cochonneau G, Kosuth P, Blitzkow D, Cazenave A (2001)Temporal variations of river basin waters from Topex/Poseidon satellite altimetry. Application to theAmazon basin. CR Acad Sci Paris 333:633–643
Dunne T, Mertes LAK, Meade RH, Richey JE, Forsberg BR (1998) Exchanges of sediment between thefloodplain and channel of the Amazon river in Brazil. GSA Bull 110(4):450–467
Frappart F, Martinez JM, Seyler F, Leon JG, Cazenave A (2005) Determination of the water volumevariation in the Negro river sub-basin by combination of remote sensing and in-situ data. Remote SensEnviron 99:387–399
Frappart F, Calmant S, Cauhope M, Seyler F, Cazenave A (2006a) Results of ENVISAT RA-2 derivedlevels, validation over the Amazon basin. Remote Sens Environ 100:252–264
Frappart F, Dominh K, Lhermitte J, Ramilllien G, Cazenave A, LeToan T (2006b) Water volume change inthe lower MEKONG basin from satellite altimetry and other remote sensing data. Geophys J Int167:570–584. doi:10.1111/j.1365-246X.2006.03184.x
Fu L, Cazenave A (2001) Satellite altimetry and earth sciences: a handbook of techniques and applications.Academic Press, London (UK), 464 p
Goldstein RM, Zebker HA (1987) Interferometric radar measurement of ocean surface currents. Nature328(6132):707–709
Guzkowska MAJ, Rapley CG, Rideley JK, Cudlip W, Birkett CM, Scott RF (1990) Developments in inlandwater and land altimetry. ESA contract report 78391881FIFL
Hostetler SW (1995) Hydrological and thermal response of lakes to climate: description and modeling. In:Physics and chemistry of lakes. Springer-Verlag, Berlin, Germany
Kaula WM (1969) The terrestrial environment: solid earth and ocean physics. NASA report study atWilliamstown, MA, NASA CR-1579, August 1969
Koblinsky CJ, Clarke RT, Brenner AC, Frey H (1993) Measurements of river level variations with satellitealtimetry. Water Resour Res 29(6):1839–1848
Koster RD, Houser PR, Engman ET, Kustas WP (1999) Remote sensing may provide unprecedentedhydrological data. American Geophysical Union. http://www.agu.org/eos_elec
Kosuth P, Blitzkow D, Cochonneau G (2006) Establishment of an altimetric reference network over theAmazon basin using satellite radar altimetry (Topex Poseidon). Proceedings of the symposium on15 years of progress in radar altimetry, Venice, Italy, 13–18 March 2006
Kouraev A, Zakharova E, Samain O, Mognard N, Cazenave A (2004) ‘‘Ob’’ river discharge from TOPEX/Poseidon satellite altimetry (1992–2002). Remote Sens Environ 93:238–245
Leon JG, Calmant S, Seyler F, Bonnet M-P, Cauhope M, Frappart F, Filizola N (2006a) Rating curves andestimation of average water depth at the upper Rio Negro river based on satellite altimeter data andmodeled discharges. J Hydrol 328:481–496
Leon JG, Seyler F, Calmant S, Bonnet M, Cauhope M (2006b) Hydrological parameter estimation forungauged basin based on satellite altimeter data and discharge modeling. A simulation for the Caquetariver (Amazonian basin, Colombia). Hydrology and Earth System Sciences. SRef-ID: 1812-2116/hessd/2006-3-3023
Leon JG, Bonnet MP, Seyler F, Calmant S, Cauhope M (2008) Distributed water flow estimates of the upperNegro river by a Muskingum-Cunge routing model using altimetric spatial data. J Hydrol
Le Traon PY, Gaspar P, Bouyssel F, Makhmara H (1995) Using Topex/Poseidon data to enhance ERS1 data.J Atmos Ocean Technol 12(1):161–170
Maheu C, Cazenave A, Mechoso CR (2003) Water level fluctuations in the Plata basin (South America)from Topex/Poseidon satellite altimetry. Geophys Res Lett 30(3):1143–1146
Mason IM, Rapley CG, Street-Perrott FA, Guzkowska M (1985) ERS-1 observations of lakes for climateresearch. Proceedings of the EARSeL/ESA symposium on ‘‘European remote sensing opportunities’’,Strasbourg, 31 March-3 April 1985
Meade RH, Rayol JM, da Conceicao SC, Natividade JRG (1991) Backwater effects in the Amazon riverbasin of Brazil. Environ Geol Water Sci 18(2):105–114
Mercier F (2001) Altimetrie spatiale sur les eaux continentales: apport des missions Topex/Poseidon etERS1&2 a l’etude des lacs, mers interieures et bassins fluviaux. These Univ. Toulouse III-PaulSabatier, 9/11/2001, p 190
Mercier F, Zanife O-Z (2006) Improvement of the Topex/Poseidon altimetric data processing for hydro-logical purposes (CASH project). Proceedings of the symposium on 15 years of progress in radaraltimetry, Venice, Italy, 13–18 March 2006
Mercier F, Cazenave A, Maheu C (2002) Interannual lake level fluctuations (1993–1999) in Africa fromTopex/Poseidon: connections with ocean–atmosphere interactions over the Indian ocean. Glob PlanetChanges 32:141–163
Mertes LAK, Dunne T, Martinelli LA (1996) Channel-floodplain geomorphology along the Solimoes-Amazon river, Brazil. GSA Bull 108(9):1089–1107
Miller LS (1979) Topographic and backscatter characteristics of GEOS 3 overland data. J Geophys Res 84-B8:4045–4054
Morris CS, Gill SK (1994a) Variation of Great Lakes waters from geosat altimetry. Water Resour Res30:1009–1017
Morris CS, Gill SK (1994b) Evaluation of the Topex/Poseidon altimeter system over the Great Lakes.J Geophys Res 99(C12):24527–24539
Nordin CF Jr, Meade RH (1986) The Amazon and the Orinoco in McGraw Hill Yearbook of Sciences andTechnology. McGraw Hill, New York, USA, pp 385–390
Ponchaut F, Cazenave A (1998) Continental lake level variations from Topex/Poseidon (1993–1996). C RAcad Sci Paris 326:13–20
Population Reference Bureau (1997) http://www.prb.org/Content/NavigationMenu/Other_reports/1997-1999/WorldPopMoreThanNos_Eng.pdf)
Roads J, Lawford R, Bainto E, Berbery E, Chen S, Fekete B, Gallo K, Grundstein A, Higgins W, KanamitsuM, Krajewski W, Lakshmi V, Leathers D, Lettenmaier D, Luo L, Maurer E, Meyers T, Miller D,Mirchell K, Mote T, Pinker R, Reichler T, Robinson D, Robock A, Smith J, Srinivasan G, Verdin K,Vinnikov K, Vonder Haar T, Vorosmarty C, Williams S, Yarosh E (2003) GCIP water and energybudget synthesis (WEBS). J Geophys Res 108(D16):8609. doi:10.1029/2002JD002583
Romeiser R, Runge H (2007) Theoretical evaluation of several possible along-track inSAR modes ofTerraSAR-X for ocean current measurements. IEEE Trans Geosci Remote Sens 45(1):21–35
Romeiser R, Runge H, Suchandt S, Sprenger J, Weibeer H, Sohrmann A, Stammer D (2007) Currentmeasurements in rivers by spaceborne along-track InSAR. IEEE Trans Geosci Remote Sens 45(12):4019–4031
Rosmorduc V, Benveniste J, Lauret O, Milagro M, Picot N (2006) In: Benveniste J, Picot N (ed) Radaraltimetry tutorial. http://www.altimetry.info
Roux E, Cauhope M, Bonnet M-P, Calmant Seyler F (2008) Daily water stage estimated from satellitealtimetric data for large river basin monitoring. Hydrol Sci J—Journal des Sciences Hydrologiques 53-1:81–99
Salati E, Marques J (1984) Climatology of the Amazon region. In: Sioli H (ed) The Amazon, limnology andlandscape ecology of a mighty tropical river and its basin. Monographs in biology, vol 56. KluwerAcademics, Norwell, MA, pp 85–126
Singh B, Swamy ASR (2006) Delta sedimentation: east coast of India. Technology Publication, Dehradun,India, iv, p 400
Sioli H (1984) The Amazon and its main affluents. In: Sioli H (ed) The Amazon, limnology and landscapeecology of a mighty tropical river and its basin. Monographs in biology, vol 56. Kluwer Academics,Norwell, MA, pp 127–165
Smith LC (1997) Satellite remote sensing of river inundation area, stage, and discharge: a review. HydrolProcess 11:1427–1439
WMO (2003) Report GTOS32. HWRP/GCOS/GTOS Expert meeting on hydrological data for globalstudies. Report, Toronto, Canada, 18–20 Nov 2002, GCOS 84, GTOS 32, WMO/TD—N�1156
WMO (2004) WMO statement on the status of the global climate. WMO Tech Rep 966, 22pWorld Resources Institute (2000) http://www.wri.org/wri/wrZakharova EA, Kouraev AV, Cazenave A, Seyler F (2006) Amazon river discharge estimated from TOPEX/
Poseidon altimetry. Geosciences Comptes Rendus (French Academy of Sciences) 338(3):188–196