Abstract
Background
TOPEX/POSEIDON Status
Jason-1 Planning
Status of Post Jason-1 Missions
Science Presentations
Splinter Meetings
Gravity Missions
Future Meetings
The TOPEX/POSEIDON and Jason-1 Science Working Teams (SWTs) met at the Keystone Conference Center, Keystone, Colorado, USA from October 13-15, 1998. The purpose of the meeting was to update the TOPEX/POSEIDON and Jason-1 Science Working Teams on the status of both TOPEX/POSEIDON and Jason-1 missions, and for the SWTs to advise the Projects on design of various aspects of the Jason-1 mission and to discuss progress of science activities undertaken by SWT members. The TOPEX/POSEIDON Satellite Altimeter Mission performance continues to provide high quality data on sea-level. Overall, instrument performance is excellent but systematic effects are beginning to show up at the 10 mm level.
Planning for the Jason-1 mission is proceeding well. The anticipated launch date is May 2000. The Jason-1 mission has more ambitious accuracy goals than TOPEX/POSEIDON. The orbit accuracy requirement is 2.5 cm rms with a goal of 1 cm. The Poseidon-2 altimeter is a two frequency altimeter and is to have a 1 s noise level of better than 1.7 cm (goal is 1.5 cm). The Jason Microwave Radiometer is to have a rss error of better than 1.2 cm.
A number of gravity missions are to be launched over the next few years and these potentially offer significant opportunities for Australian Science.
The next TOPEX/POSEIDON and Jason-1 Science Working Team meeting will be a Biarritz-style meeting. Possible dates are 11-15/10/99 or 18-22/10/99. The location will be Biarritz or Arcachon (near Bordeaux).
A summary of the current status and future options for the TOPEX/POSEIDON Project, Jason-1 and the gravity missions is reported here.
The TOPEX/POSEIDON satellite mission was designed specifically to make high quality global observations of sea surface height for improved estimates of the ocean circulation and for better understanding of the role of the ocean in the global climate system. Launched in 1992, the TOPEX/POSEIDON satellite altimeter mission is a key component of the World Ocean Circulation Experiment. It had an initial mission of three years but is now expected to continue to operate until 2001.
Following announcements of opportunity by NASA and CNES, a joint CSIRO and external collaborators team was selected in 1987 by NASA to be part of the TOPEX/POSEIDON Science Working Team (SWT). A TOPEX/POSEIDON extended mission team was subsequently selected (in 1996) by NASA to be part of the TOPEX/POSEIDON Extended Mission Science Working Team. Following the outstanding success of TOPEX/POSEIDON, NASA and CNES agreed to launch a follow-on mission, Jason-1, in May 2000. Again a Hobart and national and international collaborators team put together a set of projects that was selected by NASA to join the Jason-1 Science Working Team. Membership of the TOPEX/POSEIDON and Jason-1 Science Working Teams gives the team immediate access to the data streams and support for utilization of the altimeter data. It also confers on the team responsibility to contribute to the design of the mission, including through undertaking relevant science projects and through contributions at Science Working Team Meetings.
The purpose of the Keystone meeting was to update the TOPEX/POSEIDON and Jason-1 Science Working Teams (SWTs) on the status of both TOPEX/POSEIDON and Jason-1 missions, and for the SWTs to advise the Projects on design of various aspects of the Jason-1 mission (particularly through the splinter sessions) and to discuss progress of science activities undertaken by SWT members.
There were a total of five Australians at the meeting (John Church, Richard Coleman, Neil White, Ken Ridgway and Nathan Bindoff). This high attendance reflects the growing importance of altimeter data for climate and oceanographic research in Australia and parallels the rapid growth in the use of altimeter data internationally.
The agenda of the Meeting and the Splinter meetings is attached (Attachment A and B).
TOPEX/POSEIDON has now been operating successfully for over six years. Data for cycle 212 has been processed by NASA and cycle 215 by CNES. The most recent CDROM of the merged GDRs completed by CNES includes cycles 208 to 210. There is a difference, evident since cycle 190, between the TOPEX altimeter (ALT) and the Poseidon altimeter (SSALT).
A total of 9 satellite manouevers have been undertaken during the six year period to maintain the exact repeat track to better than 1 km. The most recent manoeuvres occurred over three years ago. It is anticipated that another manoeuvres may be required in early 1999.
Overall instrument performance is excellent but systematic effects are beginning to show up at the 10 mm level. The major instrumental problems are occurring with a drift of the TMR instrument used to get the water vapour correction and a drift of the TOPEX altimeter itself (both significant wave heights and sea level).
The extended AGC field has been added to the telemetry stream with software version 6.3 introduced and data have been reprocessed back to cycle 200.
Changes have been observed in SWH values since cycle 170. There is a need to revise the s 0 calibration for the GDRs. The T/P Microwave Radiometer (TMR) has drifted by about 1mm/yr equivalent sea level up to the end of 1996 due to a drift in the 18 GHz brightness temperature. Since cycle 190, TMR has exhibited unusual behaviour with changes at the 1 cm level.
There was significant discussion of the drift of the TOPEX altimeter associated with a linear trend in the point target response. This trend is giving errors in significant wave height and a change in the mean ocean height. If no action is taken then it is anticipated that by 2000, the errors associated with the linear trend will not be correctable. They are considering switching to side B of the TOPEX altimeter but this entails some risk.
NASA has committed funding through 2000 for continued support of the mission and the project is requesting funding through 2002. CNES has funding in place.
Like TOPEX/POSEIDON, Jason-1 is a joint USA/France NASA/CNES mission. It has a minimum design lifetime of 5 years. In any 12 month period, 95% of the data is required to be available from ocean areas. CNES is responsible for the satellite and the ground system and for the first two months of the mission CNES is responsible for the mission operations. After this period, JPL becomes responsible for the mission operations.
CNES is to supply the two frequency altimeter (Poseidon-2), the DORIS Tracking System and the Proteus satellite bus. NASA is to supply the Jason Microwave Radiometer (JMR), the Turbo Rogue Space Receiver (TRSR) for GPS orbits and the Laser Retroreflector Array (LRA). The total weight is to be about 473 kg (compared with 2500 kg for TOPEX/POSEIDON) and the total power requirements are about 450W.
For the verification phase, the Jason-1 orbit will be an identical ground track to TOPEX/POSEIDON and passing over the same point within 10 minutes of TOPEX/POSEIDON. This is to allow for cross-calibration between the altimeters.
Jason-1 is planned for launch in May 2000 on a US DELTA-2 launch vehicle (7920-10) from Vandenburg Air Force Base. Of 260 attempts this launch vehicle has had 242 successes (95% success rate). JASON will be the first payload out from the dual payload. Orbital parameters are a= 7704.0 km, e=0.0005, I=66.039° , w = 90° , and W and M will be optimised after launch. There is a 20 minute launch window to ensure Jason-1 has the same ground track as TOPEX/POSEIDON and that it passes over the same point within 10 minutes of TOPEX/POSEIDON. The Jason-1 orbit will be 10 km lower than the TOPEX/POSEIDON orbit.
There are to be three data products. These are:
The Jason-1 mission is driven by the science requirements and has more ambitious accuracy goals than TOPEX/POSEIDON. However the mission has limited system contingency margins. For the GDR, the orbit accuracy requirement is 2.5 cm rms with a goal of 1 cm. Orbits are to be computed by both NASA and CNES. For TOPEX/POSEIDON, the radial orbit error budget is 35 mm for the JGM-2 gravity model and 20 mm for the JGM-3 gravity model. The orbits will be computed in 10 day arcs with a radial orbit budget of static gravity (4 mm), earth & ocean tides (4 mm), temporal gravity (4 mm), surface forces (5 mm), data errors (4 mm), station location (4 mm) for a total rss error of 10 mm.
The Poseidon-2 altimeter is a two frequency altimeter (Ku band for measurement and C band for ionospheric correction) with ~ 9 msec travel time (two-way). The distance resolution is linked to the chirp bandwidth. For an ocean target, the return signal is affected by speckle – will use the Brown model. The pulses will be emitted in the ratio of 1 C band to 6 Ku band. The altimeter is to have a 1 s noise level of better than 1.7 cm (goal is 1.5 cm with 2.5 cm for OSDR) with a bias of less than 5 cm (or 0.5 cm?) and a stability of better than 1 mm/yr. The JMR has slightly different frequencies but very similar characteristics to the TMR. It is to have a rss error of better than 1.2 cm. The TRSR is a development on the TOPEX/POSEIDON receiver and quite different to the GPS system that recently failed on the GEOSAT Follow-On mission. The LRA is of GFO design and flight heritage.
To achieve the high POD requirements, a new set of standards are required – with both CNES and NASA using identical standards (algorithms), reference systems and processing strategies. There has been a transfer of the T/P POD processing to a new UNIX-based environment which will improve performance. To ensure consistency between TOPEX/POSEIDON and Jason-1, these new standards will require reprocessing of the complete TOPEX/POSEIDON data set. It is anticipated reprocessing of the two mission data sets will be required about the end of 2001. It was pointed out that the ultimate accuracy of TOPEX/POSEIDON data is limited by the available tracking information such that continual reprocessing of TOPEX/POSEIDON data could not be justified. The TOPEX/POSEIDON orbits would never get to the 1 cm level whereas the Jason-1 orbits may get to that accuracy level. POD accuracy from DORIS is 4 cm (1 s ) for the IGDR, 2.5 cm for GDR and 2.5 cm for the GPS orbits. There will be parallel data processing and data distribution by both agencies.
After launch there will be a three-week instrument verification phase with a commissioning phase from cycle 0 to the end of 2000.
The current CalVal plans call for cross calibration between TOPEX/POSEIDON and Jason-1 of 5 mm, determination of the absolute bias of Jason-1 to 1 cm and estimating the drift rate of Jason-1 to better than 1 mm/yr. The NASA CalVal site will continue to be Harvest (or a nearby site - there is concern that Chevron may dismantle the Harvest platform) and the CNES CalVal site will be Corsica. It is proposed to deploy about 10 GPS instruments at various tide gauges around the globe. It seems that these goals will need to be surpassed if Jason-1 is to contribute significantly to the determination of global sea-level change.
Status of Post Jason-1 Missions
There was a discussion of post Jason-1 satellite altimeter missions and the potential transfer of altimeter missions from a research to an operational mode. The USA operational system is the National Polar Orbiting Operational Environmental Satellite System (NPOESS). This system consists of three sun synchronous satellites in near circular 98.2° orbits at 833 km altitude with 10-35 day repeats. The likeliest launch date for an operational altimeter mission would be 2011, with a possible launch in 2007. The expected lifetime is 10-20 years. To achieve this operational status, the value of continued altimeter measurements to society (not the research community) would need to be demonstrated.
It was argued, particularly by Carl Wunsch, that it was premature for altimetry to go operational - it was still a developmental issue, that the highest possible quality data was needed for the foreseeable future and it was too early to compromise on mission requirements. After some discussion, the following resolution was agreed to
Recommendation on Post Jason-1 Satellite Altimetry Missions by the Joint TOPEX/Poseidon/Jason-1 Science Working Team October 15, 1998, Keystone, Colorado Recognizing,
- The demonstrated usefulness of TOPEX/Poseidon to the scientific investigations of seasonal to interannual climate variability (e.g., El Niño of 1997/98) and long term climate change (e.g., sea level rise).
- The emerging operational applications of satellite altimetry (such as climate prediction, naval and commercial risk mitigation, and fisheries management). Further scientific development is likely to broaden operational utilization.
- NASA's and CNES's capability to develop new technologies for altimetric measurements (e.g., Jason-1 is much lighter, cheaper, and developed faster than TOPEX/Poseidon with better accuracy). We expect this capability will lead to future altimetric missions with increased accuracy and resolution at lower cost.
- There is currently no provision in the U.S. or European space programs to provide scientific quality altimetry missions after Jason-1, even though many international programs (such as CLIVAR, GODAE) requiring global integrated observing systems count on their presence.
The Joint TOPEX/Poseidon/Jason-1 Science Working Team therefore urges NASA and CNES to commit to a scientific-quality altimetric mission that overlaps the end of Jason-1. This new mission would:
- Ensure the continuity of scientific-quality ocean topography measurements after Jason-1 and before the start of an envisioned operational program that will deliver same or better high accuracy data.
- Provide a stepping stone to a long-term operational system by demonstrating prototype elements of such a system.
- Continue the development of and provide the means to test new technologies.
Furthermore, the Joint TOPEX/Poseidon/Jason-1 Science Working Team is committed to finding solutions to the problems associated with a transition to future operational, scientific-quality, altimetric missions.
There were two invited science presentations and a poster session. A brief summary of the invited talks is
Toni Busalacchi - El Nino and the role of altimetry
Toni gave a description of the 1998 El Nino/La Nina transition. There was a peak in the NCEP SST in the eastern Pacific in Jan '98. A cool pool developed in the mid-Pacific in June '98 (~6° C drop in ~ 20 days ) and then the cold anomaly moved across the Pacific through to Sep-Oct '98. SSMI wind stress anomalies were a maximum in Jan (1 dyne cm-2) in the mid-Pacific. After May '98 there was a weakening of the wind field. T/P anomalies of 15-20 cm corresponded to the Jan '98 SST maximum. There was a big shoaling of the thermocline south of 10° S at ~ 160° W and upwelling in the equatorial region. NCEP Forecast is for a strengthening of the La Nina for early 1999. Most models have under-predicted the ENSO events, with ECMWF the only model that gave good results. In hindcast mode, models get much closer to reality if they initialise models with ALT or wind data – basically need a proxy for the sub-surface thermal structure. NASA has a new seasonal-to-interannual prediction project being led by Michele Rienecker and Dave Adamec. To date there is not much evidence of the impact of TOPEX/POSEIDON data on El Nino prediction but T/P has been important in the detection of El Nino. Altimeter data is much better at picking decadal scale height changes. The best ENSO predictions currently come from ECMWF and they do not currently use TOPEX/POSEIDON data. For seasonal to interannual prediction, use of TOPEX/POSEIDON data is currently in the research mode and a longer time series was required for it to be considered an operational mode.
Warren White - Air-Sea Coupled Planetary Waves
Warren gave an overview of recent work on the observation and theoretical development of coupled air/sea planetary waves in the equatorial, tropical and extra-tropical regions. He showed the coupling between species of Rossby waves – extra tropical biennial waves -> coupling yields faster speeds; tropical biennial waves -> coupling yields slower speeds and dispersion; equatorial ENSO waves -> coupling yields much slower speeds. The equatorial ENSO coupling derived from the sea level pressure response to SST anomalies, yielding wind stress curl anomalies that make the Rossby waves travel slower. There is a recent JPO paper summarising his findings.
Poster Session
The list of title of the posters is attached (Attachment C) and a set of abstract is available from the attendees at the meeting. The posters were split into groups on
Global and Basin Scale Ocean Circulation,
Mesoscale Ocean Circulation,
Tropical Ocean Circulation,
Tides,
Geodesy and Geodynamics,
Wind/Wave/Rain/Surface Effects,
Calibration/Validation/Instrumentation, and
Outreach Activities.
The first item discussed was the need to correctly estimate the S2 atmospheric tide for both the dry tropospheric delay and for the oceanic response to the atmospheric S2 tide. The atmospheric S2 tide is currently aliased in the 6-hourly atmospheric pressure fields obtained from ECMWF. Alternatives discussed were to obtain the atmospheric pressure fields at hourly intervals and to run a global barotropic model forced by global wind fields and atmospheric pressures. The amplitude of the S2 tide varies seasonally and interannually.
There are a number of new tidal models - both hydrodynamic models (Le Provost and Tierney) and statistical models (Eanes). The hydrodynamic models have amplitudes too large by order 10 cm; i.e., there is not enough energy dissipation. When the altimeter data is assimilated in these models, the tidal amplitudes around New Zealand decrease by about 3 cm. The Eanes model is of somewhat greater resolution and somewhat more accurate. This model is to be released shortly on the Web.
Craig Tierney reported on the seasonal dependence of crossover residuals at high latitudes caused by variable ice cover.
Richard Ray showed that assuming a constant fraction to estimate the loading tide was a poor approximation. He showed examples of the effects of ocean ridges.
There is quite a lot of work occurring looking at the shallow water tides.
There is growing agreement between the long period tides from numerical models and TOPEX/POSEIDON solutions. It is likely that future GDR corrections will rely on one of these models rather than an assumption of equilibrium response. There are few gauges that can be used to test the validity of these long period tidal solutions.
The decision as to what tidal model to use for Jason-1 has not yet been decided and a tidal model intercomparison is planned.
Sea surface effects are presently the largest source of error in the TOPEX/POSEIDON data set. Theories for sea surface effects are based on wind-driven waves, but less than half of the waves are locally wind-driven. Hence the SSB term is difficult to quantify as a 'global' algorithm and for JASON perhaps regional correction schemes need to be developed.
Discussion focused on work done on non-parametric sea state bias (SSB) estimation and the problems for crossover solutions. Phillippe Gaspar gave a talk on his theoretical work involving kernel smoothing (see recent JGR paper) in which he simulated a set of crossover differences as
SSH1 – SSH2 = BM4(U1,SWH1) – BM4(U2, SWH2) + e
He found about 2 cm SSB errors at the boundaries of the domain of typical waveheights and wind speeds. The kernel approach was used to make the solution matrix sparse to speed up computation times. The solution depends on the bandwidth of the data and the adopted probability density function of the variable. A locally linear estimator gives an unbiased estimator. His findings indicated that the relative SSB for T/P should be closer to 3-4% rather than the adopted 2%. Errors are now at the 1 cm level. The crossover model used was a global fit to 100 cycles of data but in reality there will be regional and seasonal differences.
POD for TOPEX/POSEIDON has provided vast improvement from the initial mission goals and, although a difficult task, the goal of 1 cm should be able to be reached for JASON-1. There is a relative bias between ALT and SSALT of the order of 1 cm after cycle retracking of SSALT. This gives mean sea level differences between CNES and NASA orbits of typically a few mm, but after cycle 170, the differences go up to about 1 cm. This change is mainly seen as a Z-coordinate drift in the reference frame. Investigators have looked at crossover differences between ERS and T/P with D t (crossover time differences) of less than hour to investigate the problem. However, since most of the crossovers with this D t range are at high latitudes, the results are not conclusive. The main reason found for the orbit differences is that CNES and NASA use different reference systems and hence station coordinates and velocities are not the same. The reference systems are drifting apart mainly due to the different station velocities. The only remedy is to make the reference systems identical. The CNES and NASA orbit software have been cross-calibrated and verified at below the 1 cm level so this is not a problem.
During the TOPEX/POSEIDON-JASON tandem phase care is needed for merging of data since the data algorithms will be different. Could process JASON with T/P algorithms, or T/P with JASON or TOPEX/POSEIDON with both. No decision was made.
Geoid improvements have been tried using ocean models and inverse solutions to improve medium wavelength geoid estimates (paper by Ganachaud/Wunsch/Tapley et al paper in JPO?), but it can only be done commensurate with our hydrographic knowledge. The gravity field will be improved with the future gravity missions.
CHAMP will improve the gravity field out to degree/order 40-50 (~ 800-1000 km). This mission is due for launch in Nov-Dec 1999 with a design lifetime of 5 years. Altitude for CHAMP is 450 km with 87° inclination and circular orbit. It will be able to measure secular variations of the gravity field and postglacial rebound; large polar ice sheet changes; global sea level change; variations and mass fluxes in the ocean, atmosphere and hydrology.
GRACE (Gravity Recovery and Climate Experiment) is designed to improve the long to medium wavelengths of the gravity field and will provide a (160,160) gravity field solution every 30 days. Hence we will get a mean field and a time variable field. GRACE uses 2 orbiters in tandem flying at about 475 km altitude with a K band ranger giving the range change between the orbiters. The inter-orbiter distance will be about 200-500 km and is a measure of the gravity field. Orbits will be done via GPS. GRACE is due for launch in June 2001 with a nominal mission lifetime of 3 years. The expected resolution of GRACE for the SNR of the ocean and geoid signal is at ~ degree 100-110 (i.e., wavelengths of ~ 360 km – best at 500-40,000 km).
GOCE (Gravity Field and Stead State Ocean Explorer) will resolve short-medium wavelength gravity field (see Delft web page – Pieter Visser) and is due for launch in 2003. Aim is to recover the gravity field to 2 mgal/2 cm at 100 km wavelengths. Orbits errors from GPS-based orbit simulation recovery indicate 2 cm radial errors, 15 cm along-track and 1 cm across-track. After about degree 50 in the gravity field expansion, GOCE does better than GRACE. GRACE is better at longer wavelengths. GOCE is good for 160-1500 km resolution.
Detlef Stammer gave a presentation where he advocated that during the tandem mission TOPEX/POSEIDON and Jason-1 should fly along parallel tracks separated in phasing by about 50 km. He argued, on the basis of analysis of model results, that this allowed the best resolution of both components of ocean surface currents. There was quite a bit of opposition to this argument with many advocating interleaved (equally spaced) TOPEX/POSEIDON and Jason-1 tracks.
Pierre-Yves Le Traon gave a presentation of the impacts of combining various altimeter data sets for determining height and velocity field. He argued that three satellite missions (ERS, TOPEX/POSEIDON and Jason-1) were necessary and advocated the interleaved scenario rather than the scenario advocated by Stammer.
The tidal specialists would like a very different phasing but their optimum choice of orbit is not practical.
In general, the discussion on this area was at a much higher level than at the equivalent session at Biarritz last year. Then, people were blithely putting up estimates of GSLR (Global Sea Level Rise) which varied substantially (e.g. between rise and fall) from minor changes in technique such as a change in the area-weighting scheme.
There is an offset between Topex and Poseidon SSH values, with Topex values about 1 cm higher.
There are differences of up to ~1cm between CNES and NASA orbits for cycles 180-210. The global difference is near 0, but the Northern and Southern Hemisphere means are different!
NASA and CNES orbits use different reference systems, which are drifting apart!
Gary Mitchum presented his latest results on the drift in the TOPEX/POSEIDON height measurements. A substantial amount of the drift disappears when the TMR drift numbers are put in. This also takes out some of the shorter wavelength signals.
It is clear that vertical land motion effects needs to be included in estimates of the altimeter bias drift.
Gary Mitchum thinks about 30 gauges with permanent GPS stations are needed to do a good calibration.
TMR Drift
Keihm, Zlotnicki and Ruf showed that the TMR has a time drift in 1992-1996. The drift stopped in 1997 and 1998 is under investigation. The 1992-1996 drift was caused by a drift in brightness temperatures in channel Tb18, and the stopped drift in 1997 also follows the Tb18.
The TMR has a scale error (an error proportional to path delay). The SCALE error is related to an old atmospheric absorption model.
Keihm, Ruf and Zlotnicki recommend a DRAFT correction formula.
PDtrue = PDgdr - 1.2*DTime + 0.068*(PDgdr+294) for DTime < (12/31/96-9/1/92)
PDtrue = PDgdr - 5.2 + 0.068*(PDgdr+294) for DTime < (12/31/97-9/1/92) AND DTime >=(12/31/96-9/1/92)
where: PDtrue, PDgdr are in negative millimeters (corrections with the TOPEX/POSEIDON convention), DTime is the difference between the time of the observation minus 9/1/92, measured in years. Depending on responses from reviewers, this formula may be modified. The DRIFT correction makes TMR measure MORE vapour than it used to see, as time increases; the drift STOPPED in 1997 and remained constant throughout the year; the correction is not valid for 1998; Keihm, Ruf and Zlotnicki are working on the 1998 behaviour. The SCALE correction makes TMR measure MORE vapour than it used to measure ABOVE 294 mm of path delay, and measure LESS vapour than it used to for lower path delays.
Jacques Stum showed comparisons between TMR and the ERS radiometer consistent with the drift above. Bruce Haines showed a comparison between TMR and columnar vapour from GPS, also consistent with the drift above.
SWH and TOPEX Altimeter Drift
Peter Challenor reported on their comparisons of different measurents of SWH. In brief, the Topex SWH is drifting by about 0.4mm/day from about cycle 130 (March 1996). The drift is constant across the wave height range.
George Hayne described the "CAL1" calibration mode and the change in return waveforms (in particular the increase in side-lobe strength) since before the start of the mission. This change produces a correctable error in the Topex SWH and mean sea level. This is described at length in the report which is available from the Wallops WWW site. He argued that the SWH drift has been continuous (at an exponential rate) rather than a drift occurring after a certain date.
Phil Callahan described the options for changing to side B (never previously used) of the satellite system. There is some risk associated with this, and there will be further discussion before a final decision is made.
There is minimal drift at the Harvest platform calibration site - very comparable to the results at Burnie.
The new CNES calibration site at Corsica was described.
Mike Parke described the calibration work in the Gulf of Mexico, in particular, the point where three satellites (T/P, ERS2 and GFO) cross over.
The Poseidon 2 internal calibration modes were described.
Bruce Haines and Yves Menard are preparing the Jason-1 CALVAL Plan. They require two page summaries from each investigator contributing to CALVAL activities. These should be sent by the 1st of November, and should include objectives, research plan and the expected outcomes. Since the Keystone meeting, there has been a more detailed request for this information.
There is a recommendation to use hourly atmospheric pressure fields for the dry troposphere correction
The land mask should identify the surface (land/lake/ice/ocean etc)
Tournadre's (?) rain flag should be used(?)
The next Algorithms meeting will be at the Spring AGU in Boston in May 1999.
Peter Niiler gave a nice talk on surface velocity field determined from TOPEX/POSEIDON and surface drifters. When care is taken to compare these data sets in well sampled and energetic regions, there is good agreement in the cross-track component of velocities. He argued that 100 km was the best length scale to smooth over to get velocities from the TOPEX/POSEIDON data. This result is similar to the Strub result. He stated that ellipses obtained by TOPEX/POSEIDON data alone could not be believed. He argued that in low kinetic energy regions (e.g., eastern central Pacific) that it is not possible to estimate kinetic energy from TOPEX/POSEIDON data. He then presented some very nice analyses of his drifter data for the North Pacific.
Bruce Cournelle gave a nice talk on the Gilson, Roemmich and Cournelle paper on the comparison of a trans-Pacific XBT section with altimeter data. I have a preprint of this paper.
Shiro Imawaki presented his work on monitoring of the Kuroshio transport using satellite altimetry and the calibration of transport estimates using in situ data. He presented results showing that the surface drifters preferentially sampled the high velocity regions of the Kuroshio and that they over estimated the mean by up to 30% (50 cm/s).
Gary Lagerloef reported on his work in which he can separate the wind-driven flow and the geostrophic flow in the equatorial and tropical regions.
There were several other presentations in this session.
There are several gravity missions either approved or under discussion. More details are in the POD/Geoid section above.
This is an approved mission for launch in November, 1999. It will improve the long wavelength gravity field out to degree 40-50.
This is an approved USA/German mission with a planned launch date of June 2001 and planned lifetime of 5 years. It will consist of two satellites with satellite to satellite ranging. It will improve the long to medium wavelength gravity field. The errors in the implied geoid should be less than the energy in the ocean field for spherical harmonics of degree less than 100. It will be used to measure the time varying gravity fields (30 day solutions). The following description is adapted from an email from Carl Wunsch announcing a meeting to be held in London (April 13-15, 1999) to discuss the oceanographic implication of GRACE.
Based upon the best engineering estimates, GRACE should be able to detect mass equivalent changes approaching 1 mm of water when averaged over regions of order 300 km across on time scales of order two weeks and longer. From an oceanographer's point of view, GRACE would give us the unique capability of measuring what is effectively the ocean bottom pressure variability, rendering the ocean transparent from space in this particular sense. But actually achieving the GRACE capability, and then using the results, raises a host of questions about both the ocean and atmosphere.
The full list of issues is very long. GRACE will be able to make one or two estimates per month, so shorter period signals will cause aliases, but it will also be extremely sensitive to oceanic motions at the diurnal and semi-diurnal periods (resonant responses) and even shorter periods will generate significant aliases. We thus need to know how well atmospheric tide components S1 and S2 are known and what the skill level is of the ECMWF, NCEP etc. products at all periods from hours out to the longest periods we can determine. Surface pressure is particularly important to GRACE. But other atmospheric variables, including surface winds which also redistribute oceanic mass, and geopotential heights (needed to represent the atmosphere's vertical density profile) are also relevant. What is the more general skill of atmospheric p(0) at all periods from hours out to the longest periods we can determine?
If there were a perfect inverted barometer response, and if all the atmospheric mass were concentrated at the surface, then knowledge of the atmosphere would be irrelevant. We know the inverted barometer breaks down at high frequencies, and possibly more generally in specific places; but a much more quantitative statement is required.
If we run barotropic ocean models, to de-alias the measurements, out to what period can one go before the baroclinic models are required? Do we understand the effects of coarse resolution on barotropic models? Do we understand how topography should be smoothed? (we suspect the answers are all "no" or "unknown").
All of the ocean models appear to show two regions of extremely intense barotropic motions in the Southern Ocean, one west of Chile, the other east of the Kerguelan Plateau. These appear (clearly for the first, probably for the second) to be regions of closed f/h contours resonating with a direct wind forcing? What does analytical theory say about the response in such places and to what extent do the models (barotropic and baroclinic) behave properly? If we were resolving the baroclinic deformation radius, would the model behavior likely be very different? Does the Circumpolar Current generate additional variability?
What role do these intense barotropic motions play in the physics of the Southern Ocean? Or are they just a "noise" process?
What is the information content of bottom pressure if assimilated into oceanic GCMs at any latitude or position?
Is there a measurable bottom pressure signature of deep western boundary currents?
What in situ measurements would best calibrate GRACE and our models? For example, is bottom pressure easier/better than vertical strings of moored current meters? Are there groups who would be interested in a collaboration with the mission?
There are potential opportunities for CMR/ACRC with GRACE. Carl Wunsch and Victor Zlotnicki have both approached us with a view to measuring full water column currents and bottom pressures in the Australian-Antarctic Basin.
Knowledge of bottom pressures together with the altimeter measurements of surface heights gives rather strong constraints on ocean circulation fields. CMR/ACRC should consider what potential opportunities this offers.
See the comments in the POD/Geoid section.
There were suggestions that future meetings should be four days long with more poster viewing time and more discussion time in the splinter meetings.
The next meeting will be a Biarritz-style meeting. Possible dates are 11-15/10/99 or 18-22/10/99. The location will be Biarritz or Arcachon (near Bordeaux).
Acknowledgement
Partial support for travel to and attendance at the TOPEX/POSEIDON/Jason-1 Science Working Team meeting was provided by the Bilateral Science and Technology Collaboration Program of the Commonwealth Department of Industry Science and Tourism.