These two meetings were held in Washington DC at the same time as the AGU Spring meeting.
They are both part of the lead-up to the launch of the Jason-1 mission. This launch was going to be in November 2000, and the timing of these meetings was based on that launch date. In late May it was announced that the launch date had been moved to 28 February 2001 but that it was hoped that it could be brought forward to December or January if one of the satellites due to be launched in the December or January slots was not ready - there is one launch per month from the launch site for this mission (Vandenberg Air Force Base). This delay in the launch was caused by the satellite not being ready.
Summary of the CALVAL meeting
The approximate timetable for Jason-1 CALVAL activities is:
The overall aim is to calibrate the Jason bias within 1cm and the relative bias between TOPEX/Poseidon and Jason within 5mm.
The calibration activities discussed at this meeting fell into three main categories:
1. Absolute bias measurements
There will be five or six groups measuring the absolute bias, that is, the absolute error in the altimeter sea-surface height (SSH) measurement. These measurements use some combination of tide gauges, GPS measurements and high- precision local geoids.
NASA and CNES each have one main calibration site making this type of measurement. The NASA site is the Harvest oil platform off the coast of California and the CNES site is in Corsica.
There will be two or three sites in the Mediterranean and one in the North Sea. Our measurements at Burnie will be the only such measurements in the SouthernHemisphere.
2. Bias drift measurements
Three groups (including ourselves) will be using data from a number of tide gauges, in conjunction with measurements of local land movement (from GPS or Doris) to estimate long-term drift in the altimeter bias.
The importance of this type of study was shown a few years ago when an algorithm error in the TOPEX/Poseidon software produced a substantial drift in the altimeter bias which was interpreted incorrectly by some as sea-level rise!
Vertical land movements are the largest source of error for this type of study and efforts are being made by the international community to install more GPS receivers near tide gauges. There are also efforts to install more Doris beacons near tide gauges and to move some of the existing Doris beacons closer to tide gauges.
3. Work on specific corrections
There are a number of groups working on either calibrating particular instruments on the satellite or improving algorithms. The most significant areas are:
Four groups will be working on validation for the Jason Microwave Radiometer (JMR). This work will mainly focus on comparisons with other sources of atmospheric water vapour data - radiosondes, satellites (e.g. ERS-2 and TOPEX/Poseidon) and water vapour measurements from GPS and VLBI.
Work continues on the EM Bias (or sea-state bias) correction. This corrects for the fact that the reflections of the radar pulse from wave troughs are stronger than those from the crests, producing an error in the measured sea-surface height. In the past wave height and wind speed have been the main inputs for this correction. It now seems that long wave condition is also important and it may be possible to use global wave model data for this. Work on this continues and it is hoped that more progress will have been made by the next SWT meeting (November 2000).
The initial aim for Precise Orbit Determination for the Jason mission is to retain the current TOPEX/Poseidon level of accuracy (~2cm) and to avoid introducing any systematic orbit differences. The longer term aim is to achieve 1cm accuracy. Improvements are needed in all inputs if this is to be achieved.
4. Data
reporting and standards
There has been some confusion over reporting of calibration data for TOPEX/Poseidon (for example, one group started, mid-mission, reporting altimeter biases in the opposite sign convention to what had been used in the past). Some effort is being made to make sure that everyone uses the same conventions. There are also some weaknesses in handling of random and systematic errors and data reporting that are being addressed.
Summary
of the algorithm review meeting
This meeting discussed some of the science algorithms and supporting
data sets that will be used by the Jason mission.
There will be a comprehensive review of all algorithms by the end of September 2000 and the formal review will be completed by the end of October. Geophysical models (MSSH, geoid etc) will be selected by the end of July.
Supporting
data sets
The Mean Sea Surface Height (MSSH) data set is mainly intended
to help users of the data remove some of the noise caused by cross-track
sea-surface height gradients. Three assessments of available
data sets have been made and the choice of data set to use seems
to be fairly clear.
There is agreement that the EGM96 geoid model is the best of the available models at this time.
Two candidate bathymetry models were put forward, but there is no clear recommendation at this time - this is partly because both models are fairly good and partly because this is not a very important data set.
A more important data set is the coastline and land/water mask. The best candidate for this is the data set that comes with the GMT plotting package! There was some controversy when it was stated that the current plan for adding an ice flag to the data is to get it from a climatology. Calculating this from the altimeter data is straightforward and is done on both TOPEX/Poseidon and the ERS satellites. I'm not sure whether this is going to be reviewed or not.
One of the most important data sets is the ocean tide correction. Eight candidate models were assessed by two different assessors and there seems to be consensus on which models are best. One proposal is that the GDR should have one empirical model (GOT99?) and one assimilation model (NAO99?).
Algorithms
An important, but less straightforward, issue is how to handle
the response of the ocean surface to atmospheric forcing. There
will be some small changes to the way the inverse barometer correction
is calculated (this correction is the first-order attempt to handle
this). There is also ongoing work on this area.
As the radiometer footprint is larger than the altimeter footprint the wet-troposphere correction can be contaminated near land, reducing the quality of otherwise good altimeter data. Some effort is going into partial solutions for this.
Work is continuing on wind speed and EM bias algorithms.