Lake level variations from satellite radaraltimetry with
retracking of multi-leading edge
Shirzad Roohi([email protected])
andNico Sneeuw
University of Stuttgart, GermanyInstitute of Geodesy
Geodätische Woche, Essen, Germany 2013
Why waveform retracking?
I Improve the quality of water level measurementsI Increase the number of reliable observations particularly in the
shoreline and shallow water
44.9 45 45.1 45.2 45.3 45.4 45.5 45.6 45.7 45.8 45.937
37.2
37.4
37.6
37.8
38
38.2
38.4Sub−satellite points
Longitude [deg]
Latit
ude
[deg
]
In−situ Gauge
tracks 371
tracks 178
2002 2004 2006 2008 2010 20121268
1270
1272
1274
1276Ascending tracks
Time [year]
Wat
er le
vel [
m]
2002 2004 2006 2008 2010 20121268
1270
1272
1274
1276
Residual = 42 cm
Ascending tracks
Time [year]
Wat
er le
vel [
m]
2002 2004 2006 2008 2010 20121268
1270
1272
1274
1276Descending tracks
Time [year]
Wat
er le
vel [
m]
Water level time series and fitting the trend to the all values
2002 2004 2006 2008 2010 20121268
1270
1272
1274
1276
Residual = 83 cm
Descending tracks
Time [year]
Wat
er le
vel [
m]
Lake level variations from satellite radar altimetry with retracking of multi-leading edge 1
RADAR principle
http://www.altimetry.info
Lake level variations from satellite radar altimetry with retracking of multi-leading edge 2
How can we have more precise water levelmeasurements?
I Increasing precision of range measurementsI Use more precise range correction, e.g. corrections included in
GDRsI waveform retracking, i.e. calculate another range correction from
SGDRs
∆Rretracking = (Gr −G0)×c2τ
Gr: retracked gate, G0: nominal retracking gate, c: light velocity, τ:pulse duration
Lake level variations from satellite radar altimetry with retracking of multi-leading edge 3
Waveform retracking techniques
I Conventional retrackers
I Onboard retracker (Ice-1/2 and Sea-ice)I Offset Center Of Gravity (OCOG)I ThresholdI β- parameters
I Unconventional retrackers
I Multi-leading edgeI Modified waveform
Lake level variations from satellite radar altimetry with retracking of multi-leading edge 4
Data and area of studyRA2 Geophysical and Sensor Data Records, i.e. RA2 GDRs and RA2
SGDRs of Envisat satellite altimetry from cycle 6 to cycle 113
Envisat satellite ground tracks from cycle 92
Lake level variations from satellite radar altimetry with retracking of multi-leading edge 5
Conventional retrackersI Onboard retrackers:
Water level from RA2 GDRs data using median values of waterlevel in each satellite over pass in Ice-1 retracker algorithm
2002 2003 2004 2005 2006 2007 2008 2009 2010 20111270
1270.5
1271
1271.5
1272
1272.5
1273
1273.5
1274
1274.5
1275Water level from median values of all satellite tracks and fitting a trend
Time [year]
Wat
er le
vel [
m]
Residual = 27 cm
h(ti) = a+ bti + ct2i + dsin
(2πT
ti
)+ ecos
(2πT
ti
)a, b, c, d and e are unknown parameters to be estimated. T is the
annual period and h is the observed water height.Lake level variations from satellite radar altimetry with retracking of multi-leading edge 6
Conventional retrackers
I OCOG
2002 2003 2004 2005 2006 2007 2008 2009 2010 20111270
1270.5
1271
1271.5
1272
1272.5
1273
1273.5
1274
1274.5
1275
Time [year]
Wat
er le
vel [
m]
Water level time series from combined ascending and descending tracks
Residual = 27 cm
I Threshold
2002 2003 2004 2005 2006 2007 2008 2009 2010 20111270
1270.5
1271
1271.5
1272
1272.5
1273
1273.5
1274
1274.5
1275 Water level time series from combined ascending and descending tracks
Time [year]
Wat
er le
vel [
m]
Residual = 15 cm
Lake level variations from satellite radar altimetry with retracking of multi-leading edge 7
Unconventional retrackersI Multi-leading edge
0 20 40 60 80 100 120 1400
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
Gate
Pow
er
Lake level variations from satellite radar altimetry with retracking of multi-leading edge 8
Unconventional retrackers
I Multi-leading edge
2002 2003 2004 2005 2006 2007 2008 2009 2010 20111270
1270.5
1271
1271.5
1272
1272.5
1273
1273.5
1274
1274.5
1275 Water level time series from combined ascending and descending tracks
Time [year]
Wat
er le
vel [
m]
Residual = 14 cm
Lake level variations from satellite radar altimetry with retracking of multi-leading edge 9
Comparing different retrackers
I Onboard retracker
2002 2004 2006 2008 2010 20121268
1270
1272
1274
1276Ascending tracks
Time [year]
Wat
er le
vel [
m]
2002 2004 2006 2008 2010 20121268
1270
1272
1274
1276
Residual = 42 cm
Ascending tracks
Time [year]
Wat
er le
vel [
m]
2002 2004 2006 2008 2010 20121268
1270
1272
1274
1276Descending tracks
Time [year]
Wat
er le
vel [
m]
Water level time series and fitting the trend to the all values
2002 2004 2006 2008 2010 20121268
1270
1272
1274
1276
Residual = 83 cm
Descending tracks
Time [year]
Wat
er le
vel [
m]
I Other retrackers
2002 2004 2006 2008 2010 20121268
1270
1272
1274
1276Ascending track
Time [year]
Wat
er le
vel [
m]
2002 2004 2006 2008 2010 20121268
1270
1272
1274
1276Ascending track
Residual = 59 cm
Time [year]
Wat
er le
vel [
m]
2002 2004 2006 2008 2010 20121268
1270
1272
1274
1276Descending track
Time [year]
Wat
er le
vel [
m]
Water level time series based on Threshold 10 % retracker
2002 2004 2006 2008 2010 20121268
1270
1272
1274
1276Dscending track
Residual = 74 cm
Time [year]
Wat
er le
vel [
m]
Lake level variations from satellite radar altimetry with retracking of multi-leading edge 10
Comparing different retrackers
Water level standard deviation from different retrackersretracker standard deviation (cm) improvement
Ice-1 27 –OCOG 27 0 %
Threshold 10 18 33 %Threshold 20 15 44 %Threshold 50 17 37 %
Multi-leading edge 14 48 %
Improvement =σIce−1 − σRet
σIce−1× 100 %
Lake level variations from satellite radar altimetry with retracking of multi-leading edge 11
Along track waveform variations
First ascending pass-Jun 2002
Latitude [deg]
Gat
e
37.2 37.4 37.6 37.8 38 38.2
20
40
60
80
100
120
0
0.5
1
1.5
2
2.5
3
x 104
Last ascending pass- Sep 2010
Latitude [deg]
Gat
e
37.2 37.4 37.6 37.8 38 38.2
20
40
60
80
100
120
0
500
1000
1500
2000
2500
3000
3500
4000
Lake level variations from satellite radar altimetry with retracking of multi-leading edge 12
Validation
I OCOG
2002 2002.5 2003 2003.5 2004 2004.5 2005−0.8
−0.6
−0.4
−0.2
0
0.2
0.4
0.6 Water level anomaly from satellite and in−situ gauge measurements
Time [year]
Wat
er le
vel a
nom
aly
[m]
RMS = 41 cm
SatelliteIn−situ gauge
I Threshold
2002 2002.5 2003 2003.5 2004 2004.5 2005−0.8
−0.6
−0.4
−0.2
0
0.2
0.4
0.6 Water level anomaly from satellite and in−situ gauge measurements
Time [year]
Wat
er le
vel a
nom
aly
[m]
RMS = 23 cm
SatelliteIn−situ gauge
Lake level variations from satellite radar altimetry with retracking of multi-leading edge 13
Validation
I Multi-leading edge
2002 2002.5 2003 2003.5 2004 2004.5 2005−0.8
−0.6
−0.4
−0.2
0
0.2
0.4
0.6 Water level anomaly from satellite and in−situ gauge measurements
Time [year]
Wat
er le
vel a
nom
aly
[m]
RMS = 26 cm
SatelliteIn−situ gauge
Lake level variations from satellite radar altimetry with retracking of multi-leading edge 14
Conclusion
I Obviously waveform retracking techniques can improve the qualityof altimetry data.
I Due to the land and environmental effects on the return echoes tothe altimeter particularly in the shoreline the waveform retrackingis necessary.
I The quality of water level is dependent on the waveform retrackingtechniques.
I According to the result of data processing using both RA2 GDRand RA2 MWS (SGDRs) of Envisat, multi-leading edge andthreshold 20 % retrackers outperform the other retackers todetermine water level variations of Urmia lake.
Lake level variations from satellite radar altimetry with retracking of multi-leading edge 15
Works under way
I Continuing waveform retracking using:I β–parameterI Modified waveform
Lake level variations from satellite radar altimetry with retracking of multi-leading edge 16
Thank you for your attention
Lake level variations from satellite radar altimetry with retracking of multi-leading edge 17
References
F. Frappart, S. Calmant, M. Cauhopé, F. Seyler and A. Cazenave, 2006,Preliminary results of Envisat RA-2 derived water levels validationover the Amazon basin. Remote sensing of environment 100, 2,252–264.
G. J. Yun, C. Xiaotao, G. Y. Gang, S.Jialong and H. C. Way, 2009, Lakelevel variations monitored with satellite altimetry waveform retracking,Ieee journal of selected topics in applied earth observations andremote sensing, 2, 2, 80–86.
G. J. Yun, G. Y. Gang, H. C. Way, and S. J. Long, 2009, Amulti-subwaveform parametric retracker of the radar satellite altimetricwaveform and recovery of gravity anomalies over coastal oceans,Science China, Earth Sciences, 53, 4, 610–616.
G. J. Yun, H. C. Way, C. Xiaotao and L. Yuting, 2006, Improved thresholdretracker for satellite altimeter waveform retracking over coastal sea,Progress in natural science, 16, 7.
Lake level variations from satellite radar altimetry with retracking of multi-leading edge 18
References
H. Lee, C.K. Shum, K. H. Tseng, J. Y. Guo and C. Y. Kuo, 2010,Present-day lake level variation from Envisat altimetry over thenortheastern Qinghai-Tibetan plateau: links with precipitation andtemperature, Terr. Atmos. Ocean. Sci 22, 2, 169–175.
J. F. Crétaux and C. Birkett, 2006, Lake study from satellite radaraltimetry, International Geophysics (Applied Geophysics) 338, 14,1098–1112.
J. S. Silva, S. Calmant, F. Seyler, O. Corrêa, R. Filho, G. Cochonneauand W. J. Mansur, 2010, Water levels in the Amazon basin derivedfrom the ERS 2 and Envisat radar altimetry missions, Remotesensing of environment 114, 10, 2160–2181.
Y. Yuchan, A. V. Kouraev, C.K. Shum, V. S. Vuglinsky, J. F. Crétaux andS. Calmanti, 2012, The performance of altimetry retrackers at lakeBaikal, Terr. Atmos. Ocean. Sci 24, 4.
Lake level variations from satellite radar altimetry with retracking of multi-leading edge 19