Environmental Monitoring and Assessment (2006) 115: 23–37

DOI: 10.1007/s10661-006-5233-9 
c Springer 2006

MEASURING WATER STORAGE FLUCTUATIONS IN LAKE

DONGTING, CHINA, BY TOPEX/POSEIDON SATELLITE ALTIMETRY

JIQUN ZHANG1,2 , KAIQIN XU2,4,∗ , YONGHUI YANG2 ,

LIANHUI QI3 , SEIJI HAYASHI2 and MASATAKA WATANABE2
1
Water Resources Management Center, Ministry of Water Resources, Beijing, 100053, China;
2
National Institute for Environmental Studies, 16-2 Onogawa, Tsukuba 305-8506, Japan; 3 Beijing
University of Aeronautics and Astronautics, 37 Xue Yuan Rd. Haidian District Beijing
100083, China; 4 Department of Earth and Environmental Engineering, Columbia University,
New York, USA
(∗ author for correspondence, e-mail: [email protected]; [email protected])

(Received 30 December 2003; accepted 8 April 2005)

Abstract. Although satellite radar altimetry was developed and optimized for open oceans, it has
been used to monitor variations in the level of inland water-bodies such as lakes and rivers. Here, for
the first time, we have further used the altimetry-derived variation of water level for estimating the
fluctuation of water storage as an addition to the present in situ water storage estimation systems to
be used in remote areas and in emergency situation such as in the events flooding monitoring and for
studying the effect of climate change. Lake Dongting, the second largest lake in China, influenced
frequently by flooding, was, therefore, chosen to demonstrate the potential of the technique. By using
the concept of an “assumed reference point”, we converted Topex/Poseidon satellite altimetry data on
water level variations in Lake Dongting to “water level” data. The “water level” time-series data and
in situ water storage were used to establish a rating curve. From the rating curve, we converted data on
“water level” derived from seven years (1993–1999) of Topex/Poseidon data to actual water storage in
Lake Dongting. The result reveals that the seasonal and annual fluctuations of water storage occurred
during the 1990s with a more frequent flooding at the late 1990s’ especially the flooding in whole
catchment level in 1998 and 1999. The study supports the usefulness of satellite altimetry for dense
and continuous monitoring of the temporal variations in water dynamic in moderate to large lakes.

Keywords: lake dongting, satellite altimetry, water levels, water storage, Yangtze river

1. Introduction

Lake and wetland water storage and river discharge are critical terms in the sur-
face water balance, yet they are poorly observed globally, and the prospects for
improvement from in situ networks are bleak (e.g., Stokstad, 1999; Shiklomanov
et al., 2002). Measuring surface storage and river discharge will allow us to under-
stand the dynamics of the land surface branch of the global hydrologic cycle, and in
particular to predict the consequences of global change on water resources. How-
ever, the large-scale dynamics of water storage in lakes, reservoirs, and wetlands
are largely unknown (Alsdorf et al., 2003).
In order to determine variations in water storage, it is essential to estimate water
level first. Currently, water level is estimated by directly measuring in the field, and
24 J. ZHANG ET AL.

converted to water storage through the use of a water level – water storage rating
curve. Despite the importance of surface water dynamics information, setting up and
maintaining a comprehensive monitoring network faces numerous technological,
economic, and institutional obstacles. For many wetlands in the world, ground
measurements are either nonexistent or not available in a timely manner. This
is especially true in developing countries, for which the cost of establishing and
maintaining a dense network of gauge stations is prohibitive (Vörösmarty et al.,
1999). Thus, a method that uses remote data to determine surface water dynamics
information would be extremely valuable from an economic perspective and would
provide a much larger data network.
Many studies have used synthetic aperture radar (SAR) data or optical image
data to monitoring surface water dynamics (Smith, 1997; Alsdorf et al., 2000;
Prigent et al., 2001; Bjerklie et al., 2003). By using “ditch indices” derived from
Landsat TM images, Shepherd et al. (2000) determined surface water storage in the
north Kent marshes. While most studies monitor water dynamic changes through
surface extent, Alsdorf (2003) estimated water storage changes across the central
Amazon floodplain using mainly GIS and interferometric SAR images, which pro-
vide the information of variations in water level. So far, most satellite image data
are purchased from data suppliers, but the prices are expensive, thus restricting the
use of the data.
Although developed and optimized for open oceans, satellite altimetry also
works well for monitoring inland water bodies. Its free use policy (Topex/Poseidon
data) and coverage regardless of time of day, weather, vegetation, or canopy cover,
encourages its use over non-ocean areas. Several studies have evaluated the per-
formance of satellite-borne altimetry over large rivers or great lakes. Most have
focused on evaluating and calibrating the altimetry system. Koblinsky et al. (1993)
explored methods of estimating water level variations in large rivers. Other studies
(Morris and Gill, 1994a,b; Birkett, 1995, 1998, 2000; Mercier et al., 2002; Maheu
et al., 2003) derived lake level time-series data and discussed the advantages and
disadvantages of using radar altimetry in large lakes (≥100 km2 ). In these cases,
radar altimetry could provide relative lake water level changes with an accuracy
of less than 10 cm (even 3/4 cm) depending on the target size and surface rough-
ness. Although none of these approaches is ideal, in part because they all rely on
instruments and platforms designed for other purposes, the advances made by these
studies provide direction for instrument improvements (Alsdorf et al., 2003).
So far, remote sensing of land surfaces can provide only the water surface area,
or information on variations in water level; the volume of water storage cannot
be observed directly. In this study, after estimating water level variations in Lake
Dongting from T/P altimeter range data from January 1993 through December 1999,
we focused on estimating water storage in the lake. As a typical flood-regulating and
-storing lake of the “swallow and spit” type, Lake Dongting is much influenced by
the seasonal water flow of its catchment area. The difference in water levels between
flood and low-water periods from ground-based gauge measurement can be used
 MEASURING WATER STORAGE FLUCTUATIONS IN LAKE DONGTING 25

to infer the water storage of the lake. Thus, differences in water storage based on
both in situ and satellite observations were detectable on a monthly time-scale.
The objective of the paper is aimed to provide a new method to directly transform
satellite-derived information to water storage as an addition to the normal methods
such as monitoring lake surface extent and monitoring just water level change.

2. Data and Methodology

2.1. STUDY SITES

Lake Dongting (Figure 1), the second largest freshwater lake in China, with a
capacity of more than 17 billion m3 , is located in the middle reach of the Yangtze
River Basin (28◦ 40 –29◦ 25 N, 112◦ 00 –113◦ 15 E), and drains as one of the sub-
tributaries into the Yangtze River. Its large amount of sediment and fresh water is
fed by four tributaries including: Xiangjiang, Zishui, Yuanjiang, and Lishui rivers.
As a branch of water supply, its large capacity for water storage is important for
the regulation of flooding in the Yangtze River. It also has functions for shipping,
fisheries, providing irrigation water, and regulating the local climate. The area
surrounding the lake has long been a major population and cultural center in China.
The lake is affected by seasonal inflows from the major river system as a function to
regulate water in the flooding season, and drain back into the Yangtze River when
water level in the Yangtze River is low. During the flood season, the lake provides a
storage place for river floodwaters. During the dry season, it provides water to the
river, thereby enabling river transportation to continue without major interruption.
Owing to both natural siltation and human activities, such as littoral land recla-
mation, the surface water area of Lake Dongting has dropped from 6000 km2 to
2623 km2 during the past centuries (Figure 2, Table I). Comparative bathymetric
surveys verify that the Dongting Lake has lost almost 2/3 of its total in the past cen-
tury (Du et al., 2001). Because of serious silting up, the lake has been divided into
three major sub-lakes: Southern, Eastern, and Western Dongting as indicated in Fig-
ure 1 and Figure 2. Inevitably, these combined processes have altered the system’s
response to Yangtze floods. The flood modulation capacity of the lake had been re-
duced considerably from 293×108 m3 in 1949 to 174×108 m3 in 1978 (Tao, 1989).

TABLE I
Area and volume evolution of Dongting Lake

Year 1825 1896 1932 1949 1954 1958 1971 1978 1995

Water area (km2) 6000 5400 4700 4350 3915 3141 2820 2691 2623
Water storage (108 m3 ) 293 268 228 188 174 167
Note. Water storage was calculated when water level reached to 31.5 m at Chenglingji Hydrological
Station.
26 J. ZHANG ET AL.

Figure 1. Location map of the Lake Dongting in the Yangtze River.

2.2. R ESEARCH DATA

2.2.1. Ground-Measured Data

Five years (1995–1999) of water level data from Chenglingji Hydrological Station,
located at jointing point from Lake Dongting to the Yangtze River, were collected.
Two years (1998–1999) of water level data from Nanzui Hydrological Station,
located at the center of the lake several kilometers away from the T/P ground track,
were used to verify the T/P-derived “water level” of the lake. A water level-storage
 MEASURING WATER STORAGE FLUCTUATIONS IN LAKE DONGTING 27

Figure 2. Long-term variations of surface water area of Lake Dongting.

look-up table from Chenglingji Hydrological Station, built in 1995, was provided
by the Yangtze River Water Resources Commission of China.

2.2.2. Topex/Poseidon Data

Topex/Poseidon is a joint NASA/CNES (Centre National d’Etudes Spatials, France)
altimetry mission devoted to studying ocean dynamics. T/P was launched in August
1992 on a 66◦ inclined orbit at 1336 km altitude. The satellite carries two altimeters:
TOPEX and Poseidon. The orbital repeat period is 9.916 days, which corresponds
to an intertrack spacing of 315 km at the equator. T/P Merged Geophysical Data
Record (MGDR-B) data were ordered from the NASA Physical Oceanography
Data Center. The MGDR-B data include corrections of geographical and instrument
range or height corrections and data quality flags, discussed in detail by Benada
(1997). Considering the availability of water level and MGDR-B data, we analyzed
7 years of T/P data from January 1993 through December 1999 (Cycles 011–267)
using the only track over Lake Dongting (Figure 1).

2.3. A PPROACH TO ESTIMATING WATER STORAGE

Water level – storage rating curves are routinely used to estimate water storage
in lakes or reservoirs from continuous recordings of water level. To assess the
28 J. ZHANG ET AL.

feasibility of satellite-based measurements, we developed rating curves between

T/P-derived information and ground-measured water storage.
Altimetry can be used to measure the instantaneous surface water height relative
to the Earth’s center of mass. A basic altimeter measurement is the difference
between the elevation of the satellite relative to the Earth’s center of mass derived
from precision orbit determination and the height of the satellite relative to the
surface as determined from the round-trip travel time of a radar pulse from the
spacecraft. Birkett (1998) provides a detailed description of the T/P mission and its
measurements. On the basis of previous studies of estimating water level variation
from radar altimeter data, we designed the following steps to determine lake water
storage.

Step 1: Estimate time-series water level variations in Lake Dongting from T/P data.
Step 2: Convert water level variation data to “water level” data after assuming a
common reference point. The altimetry data are not actual elevation but
elevation variations with respect to a reference point. The reference point
is arbitrary but the variation from the reference point is real.
Step 3: Construct rating curve between “water level” data and corresponding
ground- measured water storage in Lake Dongting.
Step 4: Estimate water storage directly from satellite altimetry measurements via
the rating curve.

3. Results

3.1. WATER LEVEL TIME - SERIES CONSTRUCTION AND ANALYSIS

The T/P ground-track over Lake Dongting and its adjacent area is pass 12 (Fig-
ure 1). The main output of MGDR-B is the altimeter range, which is the distance
from the satellite to the surface of the Earth. The reported range of the MGDR-B
data was already corrected for instrument effects. The range was then corrected for
atmospheric effects and the nature of the reflecting ground surface. The corrections
include wet and dry troposphere correction, ionosphere correction, and electromag-
netic bias. A full discussion of the derivation of altimeter heights and associated
errors can be found elsewhere (in general, Koblinsky et al., 1993; Benade, 1997,
Birkett, 1998). The general procedure can be summarized as follows:

H = Sat Alt − RangeCORR + Tide + Slope, (1)

RangeCORR = H Alt + Wet Corr + Dry Corr + Iono Corr + EM bias, (2)

where, H is the height of the water surface derived from altimeter; Sat Alt denotes
the satellite altitude above the reference ellipsoid; H Alt is the altimetry distance
 MEASURING WATER STORAGE FLUCTUATIONS IN LAKE DONGTING 29

between the altimeter and the surface; Wet Corr, Dry Corr, and Iono Corr denote
atmospheric (wet, dry tropospheric, and ionospheric) corrections; Tide is the tidal
correction; EM bias is the electromagnetic bias correction; Slope is a correction
to account for topographic gradient. Slope correction was done by averaging the
altimeter range whose observation points are inside the lake. The effect of tide was
ignored rather than giving an average value to the tide height.
The T/P radar altimeter measured the water surface height every ∼10 days
along the track that crosses Lake Dongting. The along-track spacing of altimetric
elevation values is about 300 km in our study site. To further improve accuracy, we
averaged all of the values whose footprint lay in the center of the lake (a precise
digital elevation model (DEM) against which to check the locations, as some of
the flags, a parameter to indicate whether or not the data point is over land, may
not change immediately for some water-to-land transitions), where tides or seiche
effects are minimal, to provide the final water level variation. After correcting and
averaging, we selected the minimum value of the time-series data on water level
variation retrieved from T/P as a reference point (its “water level” value was set
as 1 m), and the difference between a water level and the minimum value was
regarded as the T/P-derived “water level”. Figure 3 shows the “water level” time-
series in Lake Dongting derived from T/P between January 1993 (orbit cycle 11)
and December 1999 (orbit cycle 267). It shows a clearly seasonal and annual cycle.

Figure 3. “Water level ± SE (Standard error)” derived from Topex/Poseidon data between January
1993 (orbit cycle 11) and December 1999 (orbit cycle 267) in comparison with ground-based water
level measurement in 1998–1999 in Lake Dongting.
30 J. ZHANG ET AL.

TABLE II
Comparison between in situ and T/P-derived water level in Lake Dongting

95% Confidence Interval

of the difference
Std. Std.
Mean N Deviation Error Mean Lower Upper Correlation

T/P 3.60 57 2.16 0.29

Ground 3.82 57 2.64 0.35
Differences −0.21 57 0.64 0.08 −0.38 −0.04 0.99
Note. N is the number of days with both T/P derived water level data and ground-based water level
data.

The highest water level in Lake Dongting increased slightly by year caused by
the big floods in 1998 and 1999 in the whole Yangtze Catchment, but the lowest
water level showed no systematic trend during the study period. In 1994, there is
large area in Yangtze Catchment was influenced by serious drought, and therefore,
the amplitude of water level fluctuation was low. As a typical flood-regulating and
-storing lake of the “swallow and spit” type, Lake Dongting is much influenced
by the seasonal water flow of its catchment area. The difference in water levels
between flood and low-water periods is very large. The water level begins to rise
continuously in April, reaching its peak in July to August, and starts to decline in
September, reaching its lowest in December to February.
Corresponding ground-measured water level time-series data from 1998 to 1999
at Nanzui Hydrological Station were collected. The difference between ground-
measured water value and T/P derived “water level” on the first data was used to
convert the ground-based water level value to similar water level from T/P altimetry.
Figure 3 also shows the comparison of the time-series “water level” obtained from
T/P with in situ water level records (with a two-year overlap). We computed the
root-mean-squares (RMS) of the water level differences over the two-year time
span. The RMS differences amount to 0.08 m (about 2.21% of the range of values
in the time-series). The RMS differences provide an estimate of the accuracy of
T/P-derived water levels. Table II gives a detailed comparison. Correlation analysis
of the two datasets shows that the T/P derived height of water is significantly
correlated with ground-measured water level (P < 0.01). Paired t test of the two
datasets shows that the T/P derived water level is not significantly different from
ground-based water level measurement (P < 0.01).

3.2. WATER STORAGE CALCULATION AND COMPARISON

To infer the water storage in Lake Dongting, we developed a rating curve between
ground-measured water level and water storage in Yangtze River at Chenglingji
Hydrological Station from the water level-storage look-up table for Chenglingji
 MEASURING WATER STORAGE FLUCTUATIONS IN LAKE DONGTING 31

Figure 4. Water level – water storage rating curve for Lake Dongting.

Hydrological Station, built in 1995 by the Yangtze River Water Resources Com-
mission of China (Figure 4). Since 1995 right in the middle of the study period, we
think that the look-up table should be suitable for the whole study period, since the
changes both in area of the lake and the lake storages between 1978 and 1995 were
relative small (according to Table I).
Because our analysis showed that the correlation of water levels between Nanzui
Station and Chenglingji is very strong and significant with a correlation coefficiency
of 0.960 (See Figure 5), on the basis of the measured water level data at Nanzui
Hydrological Station and the water storage calculated through the rating curve, we

Figure 5. Comparison of water level between Chenglingji Station and Nanzui Station. The correlation
coefficiency is 0.960 (P < 0.01).
32 J. ZHANG ET AL.

Figure 6. T/P-derived “water level” – water storage rating curve for Lake Dongting.

created a time-series water storage corresponding to the T/P-derived “water level”

for 1998 to 1999. We also established a rating curve between T/P-derived “water
level” and in situ water storage (Figure 6). The rating curve has the following
equation:

W = −1.5148H3 + 22.4368H2 − 58.5753H + 62.0749(R 2 = 0.94), (3)

where H is T/P-derived “water level” and W is water storage in Lake Dongting.

From equation (3), the other five years (1993–1997) of T/P-derived “water level”
data were also converted to water storage data. We compared altimeter measure-
ments and conventional river gauge observations for the period from January 1995
through December 1999 (Figure 7). The statistical results are listed in Table III.
The correlation coefficient between the in-situ and T/P derived water storage is
0.96 with a significance of P < 0.01. It shows the effectiveness of using T/P al-
timetry data for water storage estimation. Paired samples T test shows that there
are not significant differences between the in-situ and T/P derived water storages.
The methodology of using T/P data for estimation of lake water storage is therefore
acceptable.

3.3. WATER STORAGE TIME - SERIES ANALYSIS

The range of the annual signal varied significantly from year to year. The water
storage showed a strong seasonal periodicity, with maximum values generally oc-
 MEASURING WATER STORAGE FLUCTUATIONS IN LAKE DONGTING 33

TABLE III
Comparison between in situ and T/P-derived water storage in Lake Dongting

95% Confidence interval

of the difference
Std. Std.
Mean N Deviation Error Mean Lower Upper Correlation

Ground 70.27 136 71.46 6.13

Satellite 67.97 136 67.77 5.81
Differences 2.30 136 20.84 1.79 −1.23 5.84 0.96

Figure 7. Comparison between ground measured and T/P-derived time-series of water storage in
Lake Dongting.

curring between July and September and minimum values between December and
February of the following year.
By using an averaging or interpolation method, we derived monthly average
water storage data (Figure 8). The T/P-derived water storage time-series data al-
lowed us to analyze the water dynamics in Lake Dongting. The difference in water
storage between the flood season (May to October) and the dry season (November
to April) is obvious. The maximum monthly water storage occurred in July and
was about five to six times the minimum value in February. Comparison of water
storage data showed that 1994 was a dry year, further demonstrated the value of
using T/P derived data for water storage estimation, since ground-based water level
data in 1993–1994 was not available for our study (though they are measured).
34 J. ZHANG ET AL.

Figure 8. T/P-derived monthly average water storage in Lake Dongting.

4. Discussion

So far, remote sensing of land surfaces can provide only the water surface area, or
information on variations in water level. The volume of water in a surface storage
is an important variable, reflecting both atmospheric (precipitation, evaporation
energy), hydrologic (surface water recharge, discharge, and ground water tables)
conditions and human control. In lakes and wetland areas where the human ef-
fects are not serious, water storage is strongly driven by climate conditions and
important for assessing net climate effects over time. However, it is often difficult
to obtain ground gauge data for many of the world’s large inland water bodies,
owing to poor data distribution or difficult terrain. Alsdorf (2003) tried to estimate
water storage changes across the central Amazon floodplain using mainly GIS and
interferometric SAR images. Though the meaning of water storage is not the same
as water storage in water bodies, Rodell et al. (2004) first substantiated GRACE
(Gravity Recovery and Climate Experiment) water storage (including soil mois-
ture, groundwater, and surface water) changes by comparing with results from a
land surface model and a combined atmospheric-terrestrial water budget approach.
Sweson and Wahr (2005) developed techniques that optimally convert the GRACE
gravity data into regionally-averaged water storage estimates. Applying these tech-
niques to recently-released gravity fields, they have derived regional estimates of
water storage variations for specific areas of the world. For the first time, water
storage for the Lake Dongting, a water body, was estimated directly through con-
verting altimetry water level data to water storage based on water level – water
storage rating equation. For a large lake like Lake Dongting, since there are always
 MEASURING WATER STORAGE FLUCTUATIONS IN LAKE DONGTING 35

ground-based water level measurements, altimetry data could not provide more
information than ground-based data. However, our study shows the possibility to
get rather acceptable estimation of water storage changes in other sites without
detailed ground-based water level measurement if a water storage and water level
curve is developed. This will be helpful to the decision makers on the top level to
get information directly through remote sensing data, not only T/P derived data.
However, satellite radar altimetry has its limitation both from temporal and
spatial resolution. T/P satellite data have a temporal resolution of ∼10 days and
can be used for rivers wider than 1 km. The T/P precision orbit is delivered 30 days
after the data, since the accurate satellite altimetry data are dependent on a precise
ephemeris, which currently requires post-processing of the satellite-tracking data.
Other T/P datasets can be delivered within 1–3 days when the requirement for
accuracy of elevation measurements is not so high. The European Space Agency’s
Remote Sensing Satellite (ERS) missions have a better potential to be used in
rivers about 0.5 km wide. For operational purposes (flood warning, operation of
hydropower), the information must be made available in near real time if it is to be
useful. To improve the temporal and spatial resolution, results from ERS-1, ERS-2,
Topex/Poseidon, Jason-1, and ENVISAT can be used in combination, which will
be a direction for future study to make the methodology more applicable.

5. Conclusions

So far, the measuring of water level variations by satellite altimetry has been devel-
oped and optimized for open oceans. This study demonstrates the potential use of
space-borne altimetry for inferring inland water storage. By using the “water level”
time-series data and in situ water storage rating curve, we converted data on “wa-
ter level” derived from seven years (1993–1999) of Topex/Poseidon data to actual
water storage in Lake Dongting. The satellite-derived water storage exhibited the
natural water storage changes for instance the dry year in 1994 and flooding years
in 1998 and 1999 and the regular changes of water storage in seasonal pattern.
The study supports the usefulness of satellite altimetry for dense and continuous
monitoring of temporal variations in water dynamics from moderate to large lakes
and enclosed seas. Studies of water dynamics in large lakes will not only have im-
portant implications in practice, but also provide good experience for water storage
studies in other large lakes or reservoirs to accumulate a data set for the synthesis
and integration of global wetland.

Acknowledgments

We would like to thank the anonymous referees who kindly reviewed the earlier
version of this manuscript and provided many valuable suggestions and comments.
36 J. ZHANG ET AL.

This research was supported by the Asia-Pacific Environmental Innovation Strat-

egy Project of Japan. The NASA Physical Oceanography Data Center (PODAAC)
provided the Topex/Poseidon MGDR-B data.

References

Alsdorf, D. E.: 2003, ‘Water storage of the central Amazon floodplain measured with GIS and remote
sensing imagery’, Ann. Assoc. Am. Geographers 93(1), 55–66.
Alsdorf, D. E., Melack, J. M., Dunne, T., Mertes, L. A. K., Hess, L. L. and Smith, L. C.: 2000,
‘Interferometric radar measurements of water level changes on the Amazon floodplain’, Nature
404, 174–177.
Alsdorf, D. E., Lettenmaier, D., Vörösmarty, C. and the NASA Surface Water Working Group: 2003,
‘The need for global satellite based observations of terrestrial surface waters’, EOS 84(29), 269–
280.
Benada, J. R.: 1997, Topex/Poseidon Users’ Handbook, http://www.nodc.noaa.gov/woce V2/disk12/
topex/docs/html/usr toc.htm.
Birkett, C. M.: 1995, ‘The contribution of TOPEX/POSEIDON to the global monitoring of climatically
sensitive lakes’, J. Geophys. Res – Oceans 100(C12), 25179–25204.
Birkett, C. M.: 1998, ‘The contribution of the TOPEX (NRA) radar altimeter to the global monitoring
of large rivers and wetlands’, Water Resources Res. 34, 1223–1240.
Birkett, C. M.: 2000, ‘Synergistic remote sensing of Lake Chad: Variability of basin inundation’,
Remote Sens. Environ. 72(2), 218–236.
Bjerklie, D. M., Dingman, S. L., Vörösmarty, C. J., Bolster, C. H. and Congalton, R. G.: 2003
‘Evaluating the potential for measuring river discharge from space’, J. Hydrol. 278(1–4), 17–38.
Du, Y., Cai, S. M., Zhang, X. Y. and Zhao, Y.: 2001, ‘Interpretation of the environmental change of
Dongting Lake, middle reach of Yangtze River, China, by 210 Pb measurement and satellite image
analysis’, Geomorphology 41, 171–181.
Koblinsky, C. J., Clarke, R. T., Brenner, A. C. and Frey, H.: 1993, ‘Measurement of river level
variations with satellite altimetry’, Water Resources Res. 29, 1839–1848.
Maheu, C., Cazenave, A. and Mechoso, C. R.: 2003, ‘Water level fluctuations in the Plata basin (South
America) from Topex/Poseidon Satellite Altimetry’, Geophys. Res. Lett. 30(3), 1143.
Mason, I. M., Guzkowsak, M. A. J., Rapley, C. G. and Street-Perrott, F. A.: 1994, ‘The response of
lake levels and areas to climatic change’, Climatic Change 27, 161–197.
Mercier, F., Cazenave, A. and Maheu, C.: 2002, ‘Interannual lake level fluctuations (1993–1999) in
Africa from Topex/Poseidon: Connections with ocean-atmosphere interactions over the Indian
Ocean’, Global Planet. Change 32(2–3), 141–163.
Morris, C. S. and Gill, S. K.: 1994a, ‘Variation of Great Lakes water levels described from Geosat
altimetry’, Water Resources Res. 30, 1009–1017.
Morris, C. S. and Gill, S. K.: 1994b, ‘Evaluation of the Topex/Poseidon altimeter system over the
Great Lakes’, J. Geophys. Res. – Oceans 99(C12), 24527–24539.
Prigent, C., Matthews, E., Aires, F. and Rossow, W. B.: 2001, ‘Remote sensing of global wetland
dynamics with multiple satellite data sets’, Geophys. Res. Lett. 28(24), 4631–4634.
Rodell, M., Famiglietti, J. S., Chen, J., Seneviratne, S. I., Viterbo, P., Holl, S. and Wilson, C. R.: 2004,
‘Basin scale estimates of evapotranspiration using GRACE and other observations’, Geophys.
Res. Lett. 31(20) L20504.
Shepherd, I., Wilkinson, G. and Thompson, J.: 2000, ‘Monitoring surface water storage in the north
Kent marshes using Landsat TM images’, Int. J. Remote Sens. 21(9), 1843–1865.
Shiklomanov, A. I., Lammers, R. B. and Vörösmarty, C. J.: 2002, ‘Widespread decline in hydrological
monitoring threatens Pan-Arctic research’, EOS T. AGU 83, 13–16.
 MEASURING WATER STORAGE FLUCTUATIONS IN LAKE DONGTING 37

Smith, L. C.: 1997, ‘Satellite remote sensing of river inundation area, stage, and discharge: A review’,
Hydrol. Process. 11, 1427–1439.
Stokstad, E.: 1999, ‘Scarcity of rain, stream gages threatens forecasts’, Science 285, 1199.
Swenson, S. and Wahr, J.: 2005, ‘Regional terrestrial water storage estimated from GRACE satellite
gravimetry and macroscale hydrologic models’, Geophys. Res. Abstracts 7, 03737.
Tao, J. Y.: 1989, ‘The influence of straightening the Jingjiang River and the Dongting Lake’, J. Central
China Normal Univ., Nature Sci. 23, 263–267 (in Chinese, with English abstract).
Vörösmarty, C., Birkett, C., Dingman, L., Lettenmaier, D. P., Kim, Y., Rodrigues, E. and Emmitt,
G. D.: 1999, ‘NASA post 2002 Land Surface Hydrology Mission Component for Surface Water
Monitoring, HYDRA-SAT’, Report from the NASA Post 2002 LSHP Planning Workshop, USA.