See below the latest Ssalto/Duacs product changes and updates.

Before 2009, see the archives of changes in Ssalto/Duacs system.

• Ssalto/Duacs product changes and updates
• 2010: MyOcean and change in Duacs data policy
• 2010: Duacs Delayed-Time reprocessing
• November 2009: Migration to new Aviso FTP hardware
• September 2009: Timeliness improvement on IGDR+OGDR derived map
• November 2009: The Ssalto/Duacs Delayed-time products updated
• July 2009: The Ssalto/Duacs Delayed-time products updated
• June 2009: The Ssalto/Duacs Delayed-time products updated
• April 2009: New NRT regional products on Black Sea
• April 2009: The Ssalto/Duacs Delayed-time products updated
• March 3, 2009 - NRT v9.0.0
• January 2009: The Ssalto/Duacs Delayed-time products updated
• January 2009: Duacs switches from Jason-1 to Jason-2 as a reference mission

What is MyOcean?

Here is an excerpt from the project description. Feel free to get more information on the project website.
http://www.myocean.eu.org/
MyOcean is the implementation project of the GMES Marine Core Service, aiming at deploying the first concerted and integrated pan-European capacity for Ocean Monitoring and Forecasting.
During years 2009-2011, thanks to FP7 co-funding, MyOcean will lead the setting up of this new European service, grown on past investments in research & development, system development and international collaborations.
MyOcean Service provides the best information available on the Ocean for the large scale (worldwide coverage) and regional scales (European seas), based on the combination of space and in situ observations, and their assimilation into 3D simulation models: temperature, salinity, currents, ice extent, sea level, primary ecosystems...

What is the link with Duacs?

The Cnes/CLS system Duacs is the backbone of MyOcean's Sea-Level Thematic Assembly Center (SL-TAC). The SL-TAC is a production unit which provides operational European forecasting centers with directly usable, mission-coherent and high-quality sea-level data from all past and present sensors.
In the MyOcean framework, Duacs is expected to provide Near Real time, offline and re-analyzed L3/L4 data to the project, to develop new products or to improve existing ones, but also to demonstrate its operational capability (global MyOcean system description, performance indicators, ...) and usefulness. The demonstration is the step towards a possible long term support from Europe to operational oceanography and the associated ocean observation capacity (e.g.: altimetry satellites).
As far as the end-user is concerned, Duacs will comply with two notable requirements from MyOcean:

• All MyOcean products are distributed free of charge for all applications (scientific or commercial) both offline or in near real time
• All MyOcean users must be identified by the project's user desk in order to map data flows from the original provider to the user and above all to try and improve its service towards them

Until 2009, Duacs data were divided into 2 categories: data older than 30 days (anonymous access, free of charge, all applications allowed), and near real time (<30 days) free of charge for scientific use or even commercial if enough scientific value is added to the product (e.g.: assimilation in a model).
The MyOcean data policy has the following notable impacts for Duacs users:

• The along-track data from Duacs are MyOcean products. Therefore the 30 day restriction to commercial application disappears. Anyone, including commercial applications, can now request an access to Near Real Time along-track data from Duacs (along-track SLA, SLA+ and so on). These data are jointly distributed by the MyOcean project and Aviso.
• The multi-mission maps are not in the MyOcean catalogue, but they are an internal validation dataset for project partners. These data are still available free of charge for scientific use. Commercial and non-scientific users can still get access to Duacs data provided that their request is accepted. These data are still distributed by Aviso.
• Anonymous data repositories will be closed on Aviso. A light-weight identification process is required. Generally speaking barely more than name, email and a couple of classical questions are necessary to subscribe for data access. 
• This data policy is applicable both for near real time and offline or delayed-time data (e.g.: level 3 derived from level 2 OGDR are considered as under the L3 data policy).
• Lastly, for all datasets delivered free of charge, the license requires explicit source credits in publications and science project reports derived from these data.

So in a nutshell: along-track SLA are free for all, maps are free for scientific use, and all products require a license agreement to get your data access setup. Barring the new identification process, this is basically transparent for all of you since a vast majority of Duacs users are either scientific or add a significant value to the original product that will justify unrestricted data access.
Note that the data policy does not take into account the status of the end-user: if a university is involved in a commercial project, its request will be processed like a private company's would be. Similarly if a scientific entity is granted free data access in response to a request for scientific use only, and if it then starts using these data in a profit-oriented project, then it can be considered a breach of license.

Practical consequences for Duacs users

Here are practical consequences:

• First of all, if you're not using the new FTP server (ftp.Aviso.oceanobs.com), you should get in touch with Aviso or MyOcean to get your access set up. The old repositories will be closed in the coming months.
• If you're using along-track data, you should register for free data access to Aviso or MyOcean. You'll have to sign up the license (e.g.: no redistribution unless formerly accepted by the project), and you should be able to access your data (free of charge),
• If you're using maps for science or non-profit projects, you should register for free data access to Aviso. You'll have to sign up a slightly more restrictive license (e.g. :  no commercial use) more or less in line with what was done for NRT data until 2009, and you should be able to access your data (free of charge).
• If you're using maps for commercial use, or if your application is not in line with the standard license agreement (e.g.:  web portal but limited to a small number of purely scientific projects), then you should get in touch with Aviso or CLS who will process your request. The (possible) fee associated to data delivery is discussed on a case by case basis.
• If you're using multiple data sets, you'll have to make multiple requests, and your unique data access will be configured accordingly.

Introduction

Last year started an ambitious reprocessing of all altimetry data since 1992 (cumulated total of about 55 years of 1 Hz data). Our historical altimetry database has been updated to the most recent reprocessed GDR (e.g.: Jason-1 GDR-C from 2009), but also to up-to-date standards (GDR-C or equivalent), on all missions. Duacs-DT itself has been improved here and there (new processing, better transition from TP to Jason, better mapping parameters) and this reprocessing implements all changes introduced in Near Real Time processing.

One should note that this reprocessing was able to benefit from data, algorithms or experience developed various Cnes projects and reprocessing. The synergy with Cal/Val and MSL activities helped with the long-term climate-oriented coherency, and a better handling of reference missions and regional MSL.  The SLOOP project provided a framework for the computation of new Mean Profiles for all missions, and SLOOP will produce better reference surfaces (Mean Sea Surface, and Mean Dynamic Topography) which are used by Duacs for drifting altimetry missions or the computation of Absolute Dynamic Topography.

However due to time or workload constraints, it was not possible to integrate all possible improvements associated to recent projects: Envisat GDR-C reprocessing not yet available, ERS reprocessed data from the REAPER project are not yet available, retracked T/P are still being analyzed, and only a fraction of the coastal improvement from Pistach or Coastalt projects were actually used due to the lack of high-resolution along-track data in Duacs.

The weekly resolution used so far for DT maps was insufficient wherever the time decorrelation scale is close to 15 days, especially when end-users want to perform a simple linear time interpolation between consecutive maps. However, switching to a daily resolution involves the computation of more than 6200 maps (some of them using three or four sensors). Not exactly a trifle. As a release date vs resolution tradeoff, the first batch of DT-2010 maps will be released by spring 2010 with a weekly temporal resolution, and the resolution should be increased to a daily resolution equivalent to the NRT resolution by summer 2010.

Note that the DT-2010 generation falls under the new data policy associated to MyOcean (See MyOcean and change in Duacs data policy). Consequently, it will not be distributed anonymously as older products were. When the new dataset is available on Aviso, you'll have to subscribe to a nominative (or company or lab) data access. As mentioned above, barring the light-weight identification needed, the change should be mostly transparent to all scientific and nonprofit users.

Here are some notable upgrades that the new generation Duacs-DT product will include (not exhaustive):

• T/P : new non parametric Sea State bias solution from N. Tran and S.Labroue with up-to-date standards (e.g.: DAC or wind algorithm) and orbit solution
• T/P : new orbit solution, kindly provided by GSFC
• Jason-1 : GDR-C used from 2009 reprocessing
• Jason-1 : used as a reference mission instead of T/P for 2002-2005
• Jason-2 : all GDR used from the mission beginning
• Jason-2 : used as a reference mission from 2008
• Envisat :  used in place of ERS from cycle 9, except for missing cycle 15 (still using ERS2 for this 35-day period)
• Envisat : new non parametric SSB solution from S.Labroue
• Envisat : GDR-C orbits used (after Cal/Val certification)
• GFO : new orbit solution, kindly provided by GSFC. The new solution not being available during complex periods, the GDR orbit is also used with empirical coherency restoration in between
• GFO : new non parametric Sea State bias solution from N. Tran and S.Labroue with up-to-date standards and orbit solution (computed after orbit error minimization on calibrated reference missions)
• ERS : minimization of brightness temperature drift before wet tropo computation (NN)
• All missions: GDR-C or better standards applied whenever necessary (GOT4v7, High Resolution DAC correction,...)
• All missions: new mean profiles computed in coherency with the new standards and algorithms. New algorithms were applied to correct for time-averaging period discrepancies (e.g.: T/P != Jason-1) or ocean variability. The new mean profiles are still referenced over 1993-1999 for the sake of coherency, but they integrate a precious information from the most recent missions and datasets.
• All missions: minor upgrades of the editing process and complete re-editing based on new standards and corrections.
• All missions: complete re-computation of the empirical cross-calibration and homogenization processes (orbit error reduction using a reference mission, and long wavelength error minimization)
• Reference missions: revisited bias from T/P to Jason-1 and transition from Jason-1 to Jason-2 in order to correct for geographical discrepancies affecting regional MSL studies carried out on Duacs multi-mission maps. Only the inconsistencies which are explained and understood are minimized (update to 2009 knowledge). No geographical minimization is performed between TP-A and TP-B, no waveform leakage correction is applied in this processing (empirical minimization process not fully certified yet).
• Mapping process: updated suboptimal OI process to minimize transition artefacts (increases map computation time by a factor of 10)
  Mapping process: map coverage extended at high latitudes (previously dismissed due to lack of confidence in Mean Profile)
• Mapping process: time resolution increased (first release still on 1 map per week basis, to be increased to 2 or 7 maps per week by mid-2010). 

Last summer we initiated a migration of the old Aviso FTP server to a more robust architecture. This migration started with the offline data (Delayed Time or DT-class), followed by a duplication of Near Real Time data on old and new hardware. The old repositories will be closed in a few months. If you are not already using ftp.aviso.oceanobs.com , feel free to check out the Duacs documentation and/or to contact the Aviso user service. The Aviso team will provide you the help you need in the migration to the new server.

Note that there is also a product nomenclature change on the new server for NRT data. The historical nomenclature is abandoned in favor of a clearer, albeit slightly more verbose, nomenclature in line with the one used for Delayed Time products. Further details are given in the Ssalto/Duacs User Handbook (See 5.3 Nomenclature, page 24).

Duacs has been running a second daily production for more than one year. This product generation uses both IGDR and OGDR (or FD GDR from Envisat) data in order to exploit the most recent datasets available. Although still tagged as experimental as a conservative precaution, it has been running nominally long enough. This subsystem will replace the historical IGDR-based NRT sub-system in the coming months. This should be transparent in terms of data access.

 If you are currently using the IGDR-based product generation, you should observe two differences:

• Along-track data: thanks to the availability of OGDR data in the second product generation, you'll see up to 48 hours of additional data in the along-track SLA datasets (~50 passes for each mission which is an average of ~300k to 450k valid points total).
• Gridded data (maps): in the historical IGDR-based production, multi-mission maps have a 6-day delay. This availability delay was the historical tradeoff in terms of timeliness vs quality. Since last summer the IGDR+OGDR-based production is computing 3 maps every day:
• A 0-day map, which is the preliminary or draft version of the multi-mission map,
• A 3-day map, which is the intermediate version of the multi-mission map,
• And a 6-day map, which is the final version of the NRT multi-mission multi-map

In other words, each and every NRT map is computed at d0 (day of the map) and update twice (at d0+3 and d0+6). Further information and illustration in the Ssalto/Duacs User Handbook (See 5.1.1 Temporal availibility, page 17).

All the delayed-time Sea Level Anomalies products (both versions "Reference" and "Updated") have been updated. Latest SLA maps and files are dated the 22^th of July 2009 on the servers. Have a look on :

• DT-MSLA for gridded products, 
• DT-SLA for along-track products.

All the delayed-time Sea Level Anomalies products (both versions "Reference" and "Updated") have been updated. Latest SLA maps and files are dated the 18^th of February 2009 on the servers. Have a look on :

• DT-MSLA for gridded products, 
• DT-SLA for along-track products.

All the delayed-time Sea Level Anomalies products (both versions "Reference" and "Updated") have been updated. Latest SLA maps and files are dated the 31^th of December 2008 on the servers. Have a look on :

• DT-MSLA for gridded products, 
• DT-SLA for along-track products.

New NRT and RT regional dataset on Black Sea : Along-track and combined maps with regional-specific parameters for the Black Sea. These data are generated with the same processing and parameters as for the offline (DT) and re-analysis (RAN) performed by Duacs last year. Access to the NRT and RT regional Black Sea products.

All the delayed-time Sea Level Anomalies products (both versions "Reference" and "Updated") have been updated. Latest SLA maps and files are dated the 03^th of December 2008 on the servers. Have a look on :

• DT-MSLA for gridded products,
• DT-SLA for along-track products.

The latest version of the Duacs-NRTsoftware (v9.0.0) has been installed and activated on the production servers.

New operational dataset : Re-integration of Jason-1 on the new interleaved ground track. After a few months with only two satellites, Duacs benefits from the excellent sampling provided by a coherent Jason tandem. Science parameters used to process Jason-1 are a combination of the parameters used for Jason-1 on its original ground track, and old parameters used to process the historical TP interleaved dataset. Interleaved Jason-1 will be available on all along-track data (NRT, NRT+RT, global, regional, SLA, ADT), and these data will be merged with Jason-2 and Envisat in all combined maps.

New demonstration dataset : Along-track and combined maps with regional-specific parameters for the Black Sea. These data are generated with the same processing and parameters as for the offline (DT) and re-analysis (RAN) performed by Duacs last year. The demonstration production will be carried in Near Real Time and in NRT+RT mode. Access to the NRT regional Black Sea products.

Improved processing : the Mediterranean Sea regional products have been upgraded to better match the quality of the offline dataset (Duacs-DT)

Improved processing : Upgrade of the OGDR (aka Real Time or FDGDR or OSDR) data editing and overall processing to better handle altimeter flows with degraded quality or unusual status (e.g. : man?uver)

Software upgrade : New processing engine used for regional-specific products (Mediterranean Sea, Black Sea...). The software upgrade does not alter the product content.

Software upgrade : Improved robustness and control on various sub-systems.

Operational change : Significant upgrade of the Quality Control and operational metrics on Real Time data used in the experimental NRT+DT product generation.

Operational hardware upgrade : New production server (hot swappable).

Hardware testing : Test initiated on a new FTP service (on production side in the coming days, and on user side in the coming months).

All the delayed-time Sea Level Anomalies products (both versions "Reference" and "Updated") have been updated. Latest SLA maps and files are dated the 11^th of June 2008 on the servers. Have a look on :

• DT-MSLA for gridded products,
• DT-SLA for along-track products.

Following the recommendations from the OSTST meeting in Nice last November, and the conclusions of the latest Joint Steering Group meeting (JSG), Jason-1 will be moved to a new orbit from January 26.

The new ground track and pass numbers are the same the ones used for Topex/Poseidon during the interleaved phase. A significant difference is that the repeat cycles of Jason-1 and Jason-2 will differ by approximately 5 days. Several maneuvers will be performed beginning on January 26, and the new orbit achieved by early February. During the drifting phase, Jason-1 data will not be available. 

In order to anticipate this change, the Duacs system will switch from Jason-1 to Jason-2 next week (tentative date: January 21). Then Jason-2 will be used as the reference mission for the orbit error reduction scheme. When Jason-1 has reached its new orbit, we'll re-integrate it in the multisatellite system as soon as Calibration/Validation metrics are considered nominal (tentative date: February 16).

After the transition period, Duacs will routinely process three satellites: Jason-2, Jason-1 and Envisat. The same changes will be carried out on the demonstration product merging NRT+RT data (a.k.a IGDR+OGDR) generated every day. Jason-2 and Jason-1 NRT+OGDR data will be processed in addition to Envisat's Fast Delivery data.  

The processing parameters will be similar to the ones used last summer for Jason2, and the transition artefacts should be minimal. The product distribution (directories and filenames) will be in line with what is being done for Jason-1 (just replace j1 by j2).