Absolute Calibration of Jason1 and TOPEXPoseidon Altimeters in Corsica

1
Absolute Calibration ofJason-1 and
TOPEX/Poseidon Altimetersin Corsica

• P. Bonnefond(1), P. Exertier(1), O. Laurain(1),
  Y. Ménard(2), A. Orsoni(3), E. Jeansou(4), G.
  Jan(4)
• (1)OCA/CERGA, avenue N. Copernic, F-06130 Grasse
• (2)CNES, avenue E. Belin, F-31055 Toulouse
• (3)IGN-CNES, avenue E. Belin, F-31055 Toulouse
• (4)NOVELTIS, Parc Technologique du Canal, 2 av.
  de l'Europe, F-31520 Ramonville

Jason-1 SWT, New Orleans, October 2002
2
INTRODUCTION
The Absolute calibration site in Corsica is based
on a double configuration A geodetic site at
Ajaccio (FTLRS has been settled from January to
September 2002) An in-situ site at Senetosa
cape under the track N85 The Senetosa site
allows to perform altimeter calibration from tide
gauges as well as from a GPS buoy.
Definition of altimeter bias calibration
sea height bias altimeter sea height - in situ
sea height  
Sea height bias lt 0 meaning the altimetric sea
height being too low (or the altimeter measuring
too long)Sea height bias gt 0 meaning the
altimetric sea height being too high (or the
altimeter measuring too short)
3
TOPEX/POSEIDON MGDR
TOPEX/Poseidon altimeters (ALT-A and ALT-B) have
been calibrated from 1998. Results show a great
coherence between both altimeters 2.8 4.5 mm
and 1.0 3.1 mm for ALT-A and ALT-B
respectively. Moreover, results are very
consistent with those obtained from the Harvest
Platform (difference of 6 and 2 mm respectively
for ALT-A and ALT-B).
4
GPSBUOY
Since 2000, a GPS buoy is also used in the
calibration process. Results show a good
consistency between tide gauges and GPS buoy
altimeter calibration even if GPS results have
not the same statistical significance due to the
lower number of determination.
5
WAVEHEIGHT
GPS buoy measurements also provide the sea height
variations due to waves. Because GPS buoy is
drifting during the calibration pass filtered sea
height is removed to avoid sea height variations
due to geoid slope (?shr). (?shr)2 (?gps)2
(?wave)2     where ?gps 2.6cm (GPS error)
?wave (waves) So, ?wave v((?shr)2 -
(?gps)2) SWH (or H1/3) is then deduced from the
formula below SWHbuoy 4.?wave   
While the correlation with T/P remains relatively
good (92) over the common observed period, the
correlation with Jason-1 is lower (77). No clear
bias has been observed but the SWH coming from
Jason-1 have higher standard deviation than T/P
ones (36cm and 26cm respectively) and the
correlation of the two sets is about 60. Because
the SWH is used to compute the Sea State Bias, a
"poor" determination may affect it and then the
deduced sea height this effect can reach 20mm in
the bias determination.
6
WETTROPOSPHERE

• Using GPS data from the geodetic reference point
  (Lighthouse) the wet troposphere path delay is
  computed
• Computation of a one per hour troposphere path
  delay using GAMIT software in relax mode
• The dry contribution is computed using local
  meteorological data

Concerning the common period between T/P and
Jason-1, correlations with GPS determination are
respectively 96 and 94. GPS - JMR -20.1
mm GPS - TMR -8.1 mm The resulting 12mm
difference between JMR and TMR means that JMR
"under-estimates" the wet troposphere path delay
Jason-1 bias should then be higher by about 12mm.
7
WINDSPEED
Wind Speed from meteorology station has been
compared to T/P and Jason-1 wind speed values in
IGDR. Results shows a good agreement with T/P and
Jason-1 wind speed of the IGDR products the main
part of differences may be linked to the location
of the meteorology station (altitude 90m near
the light house) where winds can be different
than off-shore. Correlations are 70 and 75
respectively for Jason-1 and T/P with very
similar time series between both satellite.
8
IGDRPRODUCTS

• TOPEX/Poseidon IGDR and GDRT from PO.DAAC are
  modified using AVISO standards
• Pole tide is computed for IGDR
• Sea state bias from TGS is replaced by the BM4
  model
• Wallops correction is included
• Wind Speed from Witter and Chelton (1991)
  formulation
• Range bias of 15mm is applied (TOPEX measuring
  too short)
• No significant differences between GDRT and M-GDR
  (TOPEX/Poseidon) after applying these
  modifications.

9
20HzDATAPROBLEM
The 20 Hz altimetric data on IGDR are not
properly corrected of the Doppler effect. This
problem was identified before the previous SWT
(June, Biarritz) and P. Vincent and S. Desai have
given a formulation for the correction. However,
our recent analysis of Jason-1 data has shown
that this formulation introduces a drift on the
altimeter bias determination. Indeed, a trend of
-1.3 mm/cycle appears between bias determination
using 1 Hz or 20 Hz data.
This trend has been identified as an error on the
formulation to correct 20 Hz data and is in
agreement with trend observed in the correction
that is supposed to be almost constant indeed
this correction produces an increase of the
altimeter range and then a decrease in the bias
determination. The high rate delta range
parameters to be found in the IGDRs are in error
because in fact corrected for the USO drift, the
CoG and the internal path delay. Then, applying
the correction method referred to in "June"
message resulted in applying twice the USO
correction. New formulation will be soon given
by the project but for this analysis, only 1 Hz
data will be used.
10
CORRECTIONS ANALYSIS
This graph shows the corrections applied to
altimetric data for Jason-1 and T/P, from cycle 2
to 22 (respectively 344 to 365), where Jason-1
IGDR are consistent.
Tropospheric correction Dry No significant
constant biases between T/P and Jason-1 has been
identified for this correction. However the
relative trend of about 0.8 mm/cycle appears to
be linked to a different behavior in the model
interpolation before and after cycle 12 between
Sardegna and Corsica. Wet For cycle 2 to 22
Jason-1 wet tropospheric path delay is shorter by
about 16mm and the behavior of the two set show a
small relative trend of about -0.5 mm/cycle. Iono
spheric correction No significant differences
have been evidenced. Jason-1 ionospheric path
delay is shorter by 8 mm and there is no
significant relative trend induces by this
correction.
Sea State Bias The interim SSB model developed by
Sylvie Labroue (CLS) has been used in place of
the BM4 given in the Jason-1 IGDR products. The
use of this model increases the Jason-1 altimeter
bias by 32mm but this fact is not significant
because it probably reveal better the
instrumental part of the bias.
11
BIASESON IGDR
Results shows a good agreement between Senetosa
and Harvest calibration sites in terms of
absolute altimeter biases even for M4 tide gauge
that has been settled in June 2002. Results
differ by about 32mm from those presented in the
previous SWT (Biarritz, June 2002) mainly due to
the use of the interim SSB model. Jason-1
(Poseidon-2) 85.6 10.8 mm TOPEX/Poseidon
(ALT-B) -1.4 10.6 mm (for comparison, ALT-B
bias determination from cycle 208 to 350 using
M-GDR is 1.0 3.1 mm) The relative bias
between T/P and Jason-1 using a combined solution
of M3, M4 and M5 determinations is Poseidon-2
ALT-B 84.1 8.1 mm
12
FTLRSCAMPAIGN
The FTLRS has been settled at Ajaccio geodetic
site (7848) from January to September 2002. Its
tracking location allows quasi-zenithal
observations of Senetosa calibration pass
N85. Very accurate short-arc orbits can then be
computed and used in the altimeter calibration
process. Moreover, the FTLRS and also the Grasse
station have the capability to switch from
Jason-1 to T/P during passes.
13
IMPACT OF THE ORBIT

• This example shows Jason-1 bias determination
  with 5 kind of orbits
• laser-based short arc orbits
• DORISSLR POE
• DORIS MOE (IGDR products)
• GPS MOE
• GPSDORIS reduced dynamic orbits.

Except for GPSDORIS reduced dynamic orbits the
values of Jason-1 bias are very close together
(1-2mm). However, the lower error bar is
obtained when using our short-arc orbit solutions
(11mm). Reduced dynamic orbits from JPL also
have a low error bar (12mm) and for the other the
error bars are at the level of 13 to 15mm.
14
CONCLUSION
The Senetosa calibration site has permitted to
calibrate T/P altimeters since 1998 ALT-A 2.8
4.5 mm ALT-B 1.0 3.1 mm Since 2000, thanks to
GPS buoy experiment the calibration process has
been improved. It also provide additional
information on SWH and wet troposphere. All
available Jason-1 cycles have been processed and
compared to T/P ALT-B -1.4 10.6 mm POSEIDON-2
83.6 10.8 mm POSEIDON-2-ALT-B 84.1 8.1 mm