The TOPEX/Poseidon Science Working Team has been meeting since 1989 and has had a significant input into the operation of the TOPEX/Poseidon mission and is now having a similar impact on Jason-1 (the follow-on mission to TOPEX/Poseidon). In addition, these meetings have provided very good opportunities for interaction amongst the science community.

I've tried to summarise the meeting briefly below. I have more detailed notes if anyone is interested in more information.

To anyone unfamiliar with satellite altimetry the explanatory note at the end may help.

TOPEX/Poseidon

This satellite was launched in 1992 and is still going strong. It is hoped to get 10+ years from it. The original plans were for a prime mission of 3 years and an extended mission of a further 2 years!

There have been a couple of problems along the way. The most recent was when it became clear that the characteristics of the main altimeter circuitry were changing with time, giving erroneous sea-surface heights (by about 1cm) and significant wave heights. It was decided about a year ago to try the redundant ("side B") altimeter channel. This was done in February this year. The characteristics of this previously unused circuitry were the same as those of side A had been when new, and side B will be used for the rest of the mission.

Another problem which became apparent about a year ago was a drift in the atmospheric water vapour measurements from the satellite's radiometer. The cause for this (an electrical or mechanical problem when the radiometer switches from looking at deep space for calibration to looking down at the ocean) is now understood and a correction has been formulated.

An update data set, to give investigators a seamless data set across the changeover of the altimeter circuitry and through the period when the water vapour readings were erroneous will be distributed early next year.

Jason-1

Jason-1 is the follow-on mission to TOPEX/Poseidon. It will have the same ground track as TOPEX/Poseidon (10 day repeat cycle, ground track repeating within +/- 1 kilometre, inclination 66 degrees).

It was to be launched in May 2000, but this has now been postponed. The new tentative launch window is September-November 2000. This delay has been necessitated by a number of things: - delays in the delivery of the star tracker - problems with the GPS receiver - problems with the solar panels - delays in the delivery of two sets of test bench software.

The schedule for launch of Jason-1 has always been tight, and this delay in the launch doesn't come as a great surprise.

The satellite is being assembled at Alcatel in Cannes, France. There was an excursion to the Alcatel plant on the Tuesday night of the meeting to see the satellite.

Apart from these problems the satellite promises to provide better quality data than TOPEX/Poseidon.

TOPEX/Poseidon and Jason-1 tandem mission

During the early part of the Jason-1 mission the two satellites will be flying along the same ground track with a time spacing of less than 10 minutes. This will allow direct comparisons between the two satellites as well as between Jason-1 and a number of calibration stations.

After this TOPEX/Poseidon will be moved onto a ground track which interleaves the Jason-1 ground track (Jason-1 will always be on the current TOPEX/Poseidon ground track). The exact spacing between the two ground tracks has not been decided yet, but it is most likely that the new TOPEX/Poseidon ground track will be half-way between Jason-1 ground tracks.

Jason-1 calibration

The first six months of the Jason-1 mission are seen as an "intense verification" phase, but it is also expected that verification activities will continue for the rest of the mission.

The CSIRO verification work using Burnie and other tide gauge sites in the region will be part of both the intense verification and also the routine verification through the rest of the mission. The National Tidal Facility are also taking part in this work.

Post Jason-1 altimetry

Jason-2 seems to be getting support within NASA and CNES, but no decisions have been made yet.

There is a proposal for a Wide Swath Altimeter, which would achieve a wider swath (about 200km) through interferometry. This could, for example, be added to Jason-2, in which case there would be two extra antennas, each about 7 metres from the main antenna. The biggest problems with this approach are the mechanical stability of the array and the roll of the platform. This is fairly speculative at this stage.

Another proposal is for a Ka band (37GHz) altimeter to complement Jason-style (Ku band (13.6GHz)) altimeters. This would be light enough (< 20kg) to fly on a micro-satellite or as a passenger on another mission. This proposal (called ALTIKA) is getting support within CNES. Such a satellite could be launched in about 2005.

Other Issues

One of the major problems with satellite altimetry is the inadequacy of the current geoids. We can expect significant improvement in this area with the GRACE (June 2001 launch) and GOCE (planned for 2004) missions going ahead.

Significant progress has been made on sea-state bias (a correction to the measured sea-surface height from the altimeter which is dependent on sea state). It is now known that this correction is highly correlated with wave slope (previous formulations have been in terms of significant wave height and wind speed). In addition, the theoretical modelling of this correction is now better.

Explanatory Note

The TOPEX/Poseidon mission is a joint NASA/CNES mission. The satellite carries two altimeters. The main altimeter is the NASA-provided TOPEX altimeter which is a dual band Ku (13.6GHz) and C (5.3GHz) altimeter. The second altimeter is the CNES-provided solid-state, single-band (Ku) altimeter. The two altimeters share one antenna, so only one can be on at a time. The TOPEX altimeter has been on for approximately 90% of the TOPEX/Poseidon mission. The Poseidon altimeter was seen as experimental, but has performed well through the mission.

The TOPEX/Poseidon mission was launched from the CNES launch site at Kourou, French Guyana, from an Ariane vehicle. The satellite itself was a NASA satellite.

Jason-1 has one altimeter, which is a descendant of the Poseidon altimeter. It is a dual band (Ku and C) altimeter. The satellite is a CNES satellite using the Proteus 2 platform. It will be launched from Vandenberg Air Force base in California from a Delta launch vehicle. NASA is also providing some of the other instrumentation - the radiometer, GPS receiver and laser retro-reflector array.

TOPEX/Poseidon project status

· operational funding has been approved through September, 2002.
· the whole system is working well.
· OMM (Orbit Maintenance Manoeuvre) 10 was in December 1998. The next OMMs will be in October 1999 and March 2000. The frequency is increasing with the increase in the solar cycle. There was a gap of nearly 3 years (from March 1996) between manoeuvres.
· there will be a switch to Doris side 2 in December 1999. - tape recorder B was restarted in April 1999 and is being used in backup mode.
· the GPS rollover in August 1999 caused some problems.
· the SEU induced safehold in August 1999 resulted in the loss of 4.5 days of data.
· the satellite seems to be good for 10+ years of life
· this estimate seems to be based largely on the life of the opto-coupler and the battery degradation. These seem to be the critical items at this stage, but this wasn't clearly stated.
· Y2K testing is complete.
· the Poseidon altimeter has been operating nominally. There have been a significant number of SEUs.
· there has been an improvement in the CNES POI
· the Earth reference system has been updated to ITRF97 and this has resulted in laser residuals being reduced from 6mm to 2.5mm.
· The French team participated in the side B CALVAL.

TOPEX/Poseidon ALT B CALVAL

· a crossover solution for cycles 11 -> 251 gives results from comparisons of ALT A to Poseidon and ALT B to Poseidon which are indistinguishable.
· the ALT B SWH data are indistinguishable from ALT A values before the problems started. The jump from ALT A (end) to ALT B was about 13cm. The Ku band sigma-naught values from ALT B and early ALT A are consistent.
· the ionosphere corrections are consistent at the changeover.
· the impact of the changeover on the SSH values is a jump of about 1cm (down) from the end of ALT A to ALT B.
· the major goal of the calibration phase has been reached.
· ALT B is working well. The higher temperature of one processing chip isn't a cause for concern. A small acquisition problem was "cured" by the August safehold!
· there is a general increase of SEUs with the increase in the solar cycle.
· the PTR of ALT B is good - it is the same as the PTR of ALT A when it was new.
· SWH is within .1m of (early) ALT A.
· sigma-naught is within .05dB of (early) ALT A.
· there are various estimates of the SSH difference between ALT A and ALT B. Gary Mitchum's estimate is that ALT B's SSH is low by about 9mm. All of the estimates are within about 5mm.
· the SWH error for ALT A grew to about .3-.4m, giving an e.m. bias error of about 6mm (SSH too high).
· the ALT A data will be reprocessed and an update product (about 80kB/cycle) will be produced. This should be released in about February 2000.
· an "end of mission" data set will be produced.

Jason-1 Project Status

· May 2000 is no longer a realistic launch date!
· September 15 -> November 15, 2000 is now selected as a tentative launch window.
· all instruments are in Cannes and the Doris, JMR (radiometer) and TRSR (GPS) instruments are all integrated and tested.
· the altimeter is integrated and under testing.
· the antennae (altimeter and Doris) and the LRA (retroreflector) will soon be integrated. - the altimeter RF units are undergoing modification and there is a problem with the Mass memory and download
· everything will have to undergo some further testing.
· all flight equipment has been integrated with the exception of the star tracker, battery, solar array and the TTC antenna.
· areas of concern
· the star tracker delivery schedule has drifted.
· the GPS receiver has a low acquisition threshold and locks up occasionally.
· the acceptability and life time of the solar panel still has to be demonstrated. Backup procurement was initiated in May 1999. This should have no impact on the overall schedule.
· there has been a drift in the delivery of two sets of test bench software. A task force (including external experts) has been set up.
· there are no major issues related to the launch vehicle. The development of the TIMED satellite is on track.

Jason-1 system and mission status

· there was a project system design and status review in November 1998. There are 3 action items still open from this:
· description of safehold recovery procedure.
· precise user interface definition.
· POD - knowledge of the centre of gravity behaviour needs to be improved and provided to the POD team.
· some satellite statistics:
· 480 kg total mass, 358 Watts, 26kg of fuel (equivalent to 111m/sec velocity change).
· 40A/h power available.
· life - qualified for 3 years, designed for 5 years.
· downlink - 763 kbits/sec.
· pointing - 0.1 degrees half cone.
· attitude restitution - 0.035 degrees (typical).
· on-orbit mass knowledge - 0.1%.
· COG knowledge - +/- 5mm on each axis (requirement).
· +/- 1mm on z-axis (objective).
· orbit - time difference between T/P and Jason
· between 1 and 10 minutes.
· it will take up to 20 days to achieve proper position.
· phases - launch - 1 day.
· assessment - 2 months.
· verification - 6 months (overlaps assessment phase).
· observation - 3+ years.

Altimeter performance status (Jason-1)

· the range noise/SWH ratio goes from 2.5cm at SWH=2m -> 5.4cm at SWH=8m.
· SWH noise is 50cm or 10% of SWH.
· sigma-naught noise is 0.7dB.
· emitted peak power is around 7W (Ku) and 16W (C).
· testing is continuing - there is no major non-compliance.

Doris performance status (Jason-1)

· all equipment tested and delivered to Alcatel
· the integration tests have been completed.
· there are some minor software anomalies
· will be upgraded in November 1999.
· no non-compliance has been detected.
· Doris orbit determination for IGDR - 4cm radial is achievable.
· for the GDR 2.5cm radial is achievable. The 1cm RMS challenge is a "real challenge".
· more beacons are being added and old ones are being upgraded.
· there are also some improvements in beacon coordinates through, for example, collocation with VLBI, SLR and GPS stations, and inclusion of beacon velocity.
· there have also been improvements in the onboard USO and beacon USO - a 20-30% improvement may be possible.

JMR/TRSR/Laser performance

· JMR - no real change on the information that was presented at Keystone last year.
· RSS total error = 1.2cm - goal is 1cm. The main problem is the physical model, rather than the instrument.
· this instrument has been delivered to Alcatel and meets all performance specs.
· TRSR (GPS) - will support 1.0-1.5cm POD
· two complete single-string flight units - each has its own antenna.
· several improvements over the T/P receiver.
· the gravity model has also been improved.
· final software delivery has been scheduled for February 2000.
· LRA - exact copy of the unit on GFO.
· meets all requirements - awaiting mechanical integration.

Jason ground system

· the ground terminals to be used are:
· Poker Flat (Alaska).
· Wallops (backup).
· Aussaguel (Toulouse).
· Hartebeesthoek (South Africa) during assessment phase.
· CNES will control the satellite until the end of the assessment phase, then JPL will take over.
· CNES will do POD for the whole mission.
· all should be ready by February/March 2000.

Splinter Groups

1. Tides
· up to 5 new models will be assessed.
· evaluation of FES98 and CSR4.0 is under way. Other models will be evaluated in the coming months. The deadline for selection is the end of March, 2000 (TBC).
· Christian le Provost will provide data from a set of deep ocean tide gauges for testing.
· there is a lack of independent data as the existing data (tide gauges and T/P) has been used in the models.
· there was some discussion of what fields we might be able to get from ECMWF. The current 6-hour frequency causes problems with the radiative tides (e.g. S1 and S2).
· some of the radiative tides may be getting corrected twice.

2. Other HF effects
· there are a variety of signals (IB departures, quasi-geostrophic signals, internal waves, equatorially trapped modes, effects in coastal and shallow seas).
· there was a lot of discussion on barotropic modes, the question being whether someone should provide a model product for subtraction.
· both wind- and pressure-driven diagnostic signals are important - a number of people are working on modelling approaches to this problem.

3. CALVAL
Gary Mitchum
· now allows temporal and spatial lags between the tide gauges and the altimeter data - significantly reduces noise.
· sees a jump from ALT A (end) to ALTB B of 10mm.
· he uses the Wallops internal calibration correction. If he leaves this out he gets the opposite curvature, but a larger linear trend. He thinks it is best to use the Wallops correction.
· he has improved his technique, but still doesn't account for the movement of the tide gauges in a very satisfactory way.

Steve Nerem
· sees a trend of about 3mm/year over the mission. This is after using Gary Mitchum's calibration.
· feels that we need about 10 years of T/P standard altimeter data before meaningful estimates of sea-level rise can be made.

Bruce Haines
· the Harvest platform is being refurbished for the Jason mission.
· the GPS data shows platform movement of about 8mm/year (platform going down).

TMR drift update
· there was a drift from September 1992 through December 1996 due to a drift in the 18GHz brightness temperature.
· the curve then seems to level out, although this change of curvature is fairly subjective, and is a subject of debate.
· from December 1997 to June 1998 there is a 2-3mm path delay jump compared to SSMI.
· the drift is caused by a mechanical and/or electrical problem when the TMR switches from looking at deep space (for calibration) to looking down at the ocean.

Jason CALVAL Plan review

· comments on the draft should be in by the 1st of February, 2000.
· the plan will be updated in March 2000.
· there will be a kick-off meeting probably in May 2000, possibly linked to the AGU meeting.
· intense verification - first 6 months.
· IGDR available within 2 days.
· monthly progress reports.
· one mid-term meeting.
· routine verification during operational phase.
· reports every 6 months?

ISSUES
· standards for information exchange, e.g.
· sign convention for bias and drift.
· ensure same corrections are used by different groups.
· reporting of error bars.

4. POD
· there is no need to reprocess the orbits.
· the change to ITRF97 fixed the CNES orbits reference system problems.
· NASA reference system still consistent with ITRF97.
· use of current models will continue until 2001 or so.
· there are no critical issues related to the SLR tracking.
· Jason POD development is on track.

5. Geoid
· we can expect significant improvement with GRACE (June 2001 launch) and GOCE (planned for 2004) going ahead.
· input from oceanographers is needed for the selection of the geoid for the Jason-1 GDR.

6. Outreach
· the outreach team can help SWT members get their stuff into suitable form for outreach activities.
· they are keen to be fed stories and information to keep them up to date.

7. Algorithms/products
· the comments and questions from the Boston meeting have been analysed and taken into account for about 90% of the algorithms.
· there are still some items being worked on.
· geophysical models (e.g. MSS, geoid, tides, bathymetry, coastline, land/sea map, e.m. bias) to be decided by March 2000.
· sample (I)GDR data will be made available by mid-November 1999.
· there was some discussion about whether other products (e.g. CORSSH/SLA-type products) should be made available.

8. Multi-satellite altimetry
· interleaved orbits with the new T/P ground track half way between the Jason ground tracks is the recommendation, but this is not cast in concrete
· a science group is continuing to work on this.
· there is a need to continue to work on science objectives and sampling and data assimilation issues for the tandem mission.

9. Sea-surface effects
· recommendation 1: move to new wind speed estimation model
· this model has been defined and a journal paper is to be submitted. It won't be ready for launch.
· recommendation 2: better model (derived from T/P) now available
· this is a non-parametric model (i.e. a lookup table).
· significant progress has been made on the sea-state bias - it is highly correlated with wave slope. Unfortunately we can't measure wave slope from a satellite!
· theoretical modelling of the sea-state bias is now better.

Post-Jason altimetry

Jason-2
· high priority within NASA
· also being supported by CNES, but no decision has been made by CNES.

WSOA
· Wide Swath altimeter - uses interferometry with two extra antennae about 7m from the main antenna.
· ~200km swath. - big problem is stability - array has to be very stable mechanically.
· biggest problem is the roll of the platform - could be taken out using a crossover solution?
· this could be flown on Jason-2, but is fairly speculative at this stage.

ALTIKA
· Ka band (37GHz) altimetry to complement Jason missions.
· conventional altimetry, on a micro-satellite or as a passenger on another mission.
· sun-synchronous.
· Ka band, dual frequency (24/37GHz).
· long repeat cycle (e.g. 35 days) and altitude of 450-850 km.
· better for ice.
· rain causes problems at this frequency - however, it is felt that a good data return can be achieved.
· smaller footprint - better for sampling heterogeneous surfaces (e.g. ice, coastal zones, sea ice boundaries).
· larger bandwidth -> better vertical resolution.
· range accuracy better than Poseidon 2.
· mass < 20Kg!
· this is getting support within CNES.
· could be launched in 2005?

Next SWT
· will be at launch time, and, hopefully, SWT attenders will be able to watch the launch.
· will probably be at Santa Barbara, California
· nearest town of any size to Vandenberg Air Force Base - about a 45 minute bus ride.