Show all / Hide all

How could I log on with my account?

Symptoms:

• I could not download the data from the authenticated ftp folders.
• I don't see any place to enter my username/password.
• I see only data from 1992 in the folders/I don't see the data I'm interested in the folders, how can I get more recent data?

To access to data from authenticated ftp site, firstly you must have a personal account provided by the Aviso Users Services after you have properly filled the registration form available online.

Secondly, you have to connect with your personal account information by proceeding as follows:

With a web browser:

• Using Firefox: WRITE ftp://yourLOGIN:yourPASSWORD@ftp.aviso.oceanobs.com into the URL bar.
• OR using Internet Explorer: WRITE ftp://ftp.aviso.oceanobs.com, click on Page, choose 'Open FTP Site in Windows Explorer' then enter your login and password.

Then browse 'global' directory ... (DO NOT browse into the 'pub' folder, it leads you to the anonymous account where only data from October, 14th 1992 to December, 30th 1992 are available).

Otherwise, see below the other ways to log in:

• by using a FTP client (FileZilla for example): enter 'ftp.aviso.oceanobs.com' in the hostname, then your login and password in the appropriate boxes.
• By using Unix commands: write 'ftp ftp.aviso.oceanobs.com' (and then enter your login/password). You can also use the command line "wget --no-proxy" with the suitable options (--user=YOUR_LOGIN  --password=YOUR_PASSWORD).

Some parameters settings have maybe to be checked: passive mode, proxy settings...

Category: Data and processing

What is the difference between the two file formats available in the FTP folders for the DUACS Along-Track data?

Since the end of September 2011, the DUACS Along-track products in Delayed-Time are available in a new format, making the data easier to use. This new format follows the standard NetCDF CF-1.4 (Climate and Forecast).
Each variable in the file has an associated description of what it represents, including physical units if appropriate, and that each value can be located in space and time: a variable refers to data stored in the file as a vector.

Filename nomenclature, paths and access rights are the same: the new format CF files are stored in folders suffixed with "_cf" at the same levels as folders containing older format files. Example: in the FTP folder ‘global/dt/upd/sla/’, folders j2/ and j2_cf/ contain respectively data in the old format and new format CF for the Jason-2 mission.

Note that the DUACS Along-Track products with the old format will still be distributed until the end of January 2012, leaving the users time to adapt to the new format CF. Beyond that date, the new format CF will be the only available format.

Category: Data and processing

Why do I observe a bias between Rio05 and RioMed around each side of Gibraltar Strait ?

The difference in mean dynamic topograpy between the Mediterranean Sea and the Atlantic Ocean in the two different products of MDT that we distribute (RioMed and Rio05) is due to the fact that both surfaces have been computed starting from different large scale first guess :

• In the Rio05 MDT, the first guess is a merging between the Levitus climatology at low and mid-latitudes and, at higher latitude, a large scale MDT based on a Mean Sea Surface (CLS01) minus a geoid model (EIGEN-GRACE3S) filtered at 400km resolution. The average of the RIO05 MDT over the latitude range between 40°N and 40°S is equal to the average over the same area of the Levitus climatology.
• On the other hand, the RioMed MDT was constructed starting from the mean of the MFSTEP Mediterranean forecasting model . The average over the Mediterranean Sea of the RioMed MDT is thus equal to the average of the MFSTEP MDT.

There is no reason why these two constants should be the same, and that is why you observe a bias between the Rio05 MDT and the RioMed MDT.

To merge the two fields, you might consider that the two constants just east of Gibraltar are about the same that the seal level just west of Gibraltar, but I have no clue of how best to deal with this.

A solution will be given in the future by the use of Goce geoid data : then the MDT may be computed for the Atlantic Ocean and for the Mediterranean Sea starting from the same first guess (MSS-Goce, filtered at some spatial wavelength).

Category: Data and processing

What are the differences between Jason-1 GDR-b and GDR-c?

 The GDR-c has a different file naming convention and has improved corrections.

• The file extension .NASA and .CNES are not added to the GDR-c file names,
• There are two new parameters, pseudo_datation_bias_corr_ku and mdt,
• Better orbit, sea stat bias, ionospheric range, inverse barometer, and tidal corrections,
• Improved rain and ice flags.

An updated user handbook is provided:

• SMM-MU-M5-OP-13184-CN, Edition 4.1, October 2008 
  Aviso and PoDaac User Handbook - IGDR and GDR Jason-1 Products, pdf version (2 Mo).

Category: Data and processing

What tide models are used in the merged sea level anomaly data?

For Jason-1&2 as well as for Envisat data, the tide models are

• ocean tide is deduced from GOT00.V2 model (R. Ray code) (load tide is included)
• earth tide is deduced from the Cartwright tide potential
• polar tide is deduced  from equilibrium model (IERS conventions)

Each tide data is interpolated on the altimeter data position (time/space interpolation) and subtracted to the altimeter measurement. The same tidal solutions are applied to each mission in order to minimize inter-mission biases.   Independently, multi-mission inter-calibration process (crossover adjustment using Jason-2 as reference) is done considering the data corrected from all geophysical corrections (tides, tropospheric and ionospheric effects, high frequency signal aliasing effect, ...).

Category: Data and processing

How is Mean Dynamic Topography calculated ?

The MDT used for Aviso products is called CNES-CLS09_v1.1 and is computed combining several steps:

• a direct method estimates at large scale a MDT using the MSS and the EIGEN-GRGS.RL02.MEAN-FIELD (this model is based on 4.5 years of GRACE data).
• a synthetic method where sea level anomalies are subtracted from in-situ measurements of the ocean state (dynamic heights over the 1993-2007 period and geostrophic surface velocities over the 1993-2008 period) to compute synthetic estimates of the MDT and the corresponding geostrophic surface velocities. 
• Then, using a multivariate objective analysis, the synthetic estimates of heights and velocities are computed and used to improve the first guess of the MDT at large scale. It's the calculation of the CMDT.
  Finally, a validation is done by comparing the MDT_CNES-CLS09 with others MDT.

The MDT is used to calculate the Absolute Dynamic Topography (ADT) according to ADT = MDT + SLA.

Further information and distribution: Combined mean dynamic topography

Category: Data and processing

Mean Dynamic Topography, Mean Sea Surface, Mean Sea Level or mean profile?

There are several means used in altimetry processing or altimetry-based studies.
The MSS (Mean Sea Surface) is a mean of several years of altimetry data, eventually using several satellites. It is a reference surface (thus you have values on a lon/lat grid).
The MDT (Mean Dynamic Topography) is the previous surface minus the geoid. It is the ocean topography due to the mean currents. It is also a reference surface.
The Mean profile is the mean of an altimeter measurements along the satellite's tracks. It can be used as reference for computing SLA for the same satellite. It is an along-track data (not distributed)
The MSL (Mean Sea Level) is a trend (a number of mm/year), often given with respect to time as a curve.

Category: Data and processing

What are the conditions to get Aviso data?

All Aviso data are free of charge for non-commercial use, whether online (FTP, Opendap) or on media (DVD).

The request must be justified by a short text explaining how the user plan to use the data. This helps Aviso in refining the available data to be better tuned to the users' needs.

For a first level of education, the best is to work with online maps (gif format), e.g. the one available through the Argonautica project, or with the Live Access Server.

Category: Data and processing

How are geostrophic velocities computed at the Equator?

In the geostrophic velocity data distributed by Aviso today, current velocities at the equator are computed using second derivative methode inspired from Lagerloef et al. [1999]. This method is quite usual. In a 5° band around the Equator, a connection is computed to ensure continuity with classical geostrophy.

Lagerloef, G.S.E.; G.T. Mitchum, R.B. Lukas, P.P. Niiler, 1999: Tropical Pacific near-surface currents estimated from altimeter, wind, and drifter data, J. Geophys. Res., 104, 23313-23326

Category: Data and processing

What are the differences between ADT and SLA products?

Absolute Dynamic Topography (ADT) and Sea Level Anomaly (SLA) are defined by:
ADT=MDT+SLA
where MDT is Mean Dynamic Topography.

Category: Data and processing

Could you send me an explanation of the times associated with your gridded data?

The gridded data are computed for a given day. Each map represents the situation on the day indicated in its name. It is not a mean over the period between two files.
To generate these maps, computing methods based on objective analysis allow us to interpolate data in time and in space. In delayed time, the processing window used is centered, e.g. 6 weeks of data before and after the given date are taken into account. In near-real time, only the 6 weeks of data preceding the given day are considered.

Category: Data and processing

When I read two consecutive Ssalto/Duacs NRT SLA files, some fields are exactly the same but the SLA filed is different. Why ?

For example, in the two Ssalto/Duacs NRT Sea Level Anomaly NetCDF files, res_oer_j2_21549_21569.nc contains the data from 21549 julian day to 21569 julian day, while the file res_oer_j2_21550_21570.nc contains the data from 21550 julian day to 21570 julian day. The two numbers in the file name indicate the period covered by the file. That's why you can find the same track in the two files.

The different value you observe on the same point between the two files comes from the time window used for the data processing. The file res_oer_j2_21549_21569.nc was generated on 21570 julian day when only data until 21569 julian day were available. One day later (file res_oer_j2_21550_21570.nc), data for day 21570 were available and the time window was shifted by one day.

This time window mainly impacts the correction of long wavelength signal such as the error on orbit determination and the residual error on high-frequency long-wavelength signal aliasing correction. The estimation of this correction for a given day is more accurate when data in the future (and of course also in the past) are available (i.e. when we are in similar condition to delayed time processing). In this way, we consider that data for a given day should better be read in the file starting from this day when possible.

 

Category: Data and processing

In the gridded data, to which point lat/lon refer to?

Each value has been computed exactly on the Lat/Lon values
defined by LatLonMin and LatLonStep , as specified in the metadata of the NetCDF file.

Explanation is given is section 4.3 of the handbook "Ssalto/Duacs User Handbook: (M)SLA and (M)ADT Near-Real Time and Delayed Time products" that you can find here

Category: Data and processing

Where can I find tables showing the correspondance date / cycle and pass number?

On-line tools on this website make the correspondance between hour and calendar date and cycle and pass number.

Category: Data and processing

How to convert measurement time in local time ?

• For T/P, 86400xTim_Moy_1 + 10^-3xTim_Moy_2 + 10^-6xTim_Moy_3 = time (in seconds)
• For Jason-1&2, 86400xTime_day + 10^-3xTime_sec + 10^-6xTim_microsec = time (in seconds)

Time is given in UTC (Universal Time Coordinate, linked to Greenwich meridian), date in Julian day, which is the elapsed time between a reference epoch (January 1^st 1950, 0h0min0s for Cnes Julian days, January 1^st 1958, 0h0min0s for Nasa Julian days) and measurement date. To convert UTC time in local time, you have to convert it in date, hours, minutes, seconds, then in solar time and finally in your local time. We haven't got any conversion procedure.

Category: Data and processing

What is the impact of missing data in the Duacs products?

When one satellite is missing for a while, due to an incident example, the impact on the products is twofold:

• the along-track files are missing if the incident lasts for more than one day

• the impact on the generation of the map depends on the gap in the data: to generate these maps, computing methods based on objective analysis allow us to interpolate data in time and in space. In delayed time, the processing window used is centered, e.g. 6 weeks of data before and after the given date are taken into account. In near-real time, only the 6 weeks of data preeceding the given day are considered. The maps of formal errors (in the directory "err") will give an overview of the error induced by the missing data in the generation of the maps.

Category: Data and processing

We haven't receive T/P cycles 118/431/432, Jason-1 cycles 178 and 375 to 500

There were no data processed for Topex/Poseidon cycles 118, 431, 432 and for Jason-1 cycle 178, due to incidents.

The cycles 375 to 500 for Jason-1 are missing because the mission ended the repeat orbit phase (until 374) and began geodetic phase from cycle 500 (the cycles have been recomputed).

Category: Data and processing

What are Jason1&2, Topex/Poseidon data resolution?

Spatially, along the track, data are available every 7 km (measurements averaged on 1 second, or "1 Hz data"). GDR 10 or 20 Hz data also exists (depending on the satellite), but those are more noisy. Between tracks, for Jason-1&2 and Topex/Poseidon, the distance is up to 315 km at the Equator; for ERS and Envisat, it is about 80 km.
Gridded Aviso data are available at 1/3° on a Mercator grid.

Category: Data and processing

What are the characteristics of the reference ellipsoid you are using?

The Aviso products are referenced to the Topex/Poseidon ellipsoid (see the references below). So the Sea Surface Height of the MSS, CorSSH products are referenced to T/P ellipsoid, even for ERS-1, ERS-2, Envisat and Cryosat products. The SLA and MSLA products are not concerned since they are anomalies.

The reference ellipsoid is the first-order definition of the non-spherical shape of the Earth. For the GDRs, concerning Jason-1&2 GFO and Topex/Poseidon, the reference ellipsoid is:

• radius : 6378136.3
• inverse Earth flattening coefficient : 298.257

For ERS-1, ERS-2, Envisat, Cryosat-2:

• radius : 6378137
• inverse Earth flattening coefficient : 298.257223563

Category: Data and processing

On which platform/Operating system can I use Aviso data?

You can use Aviso data on every platform and Operating system, but the reading software may not be available for your precise configuration (see the available software).
Note that T/P GDR-Ms are coded in little endian format and Jason-1 GDRs in big endian format.

Category: Data and processing

I'd like data over a precise area (data extraction)

- for gridded NetCDF data, an online extraction tool is available
- for Topex/Poseidon and Jason-1 GDR, Aviso can perform extraction for a short period of time on a small area
For the other dataset, users have to perform their own extraction from the global or regional datasets

 

Category: Data and processing

Where can I find NRT products before January 2011?

Actually, there is no Near-Real Time (NRT) product before a certain date (January 20th, 2011 for most of them), in the same way as there is no DT product after this date.

DT gridded products which are more precise than NRT products, substitute for NRT products (more explanations into the Ssalto/Duacs User Handbook (PDF) subsection '4. Delayed Time Component').

Note that the availability date of NRT and DT products changes after each new delivery of Delayed-Time products, as a result, the availability of these dates brings forward (see question '[...] Is there a time delay in data release?')

If you want to work with 'historical' time series, the Delayed-time data are what you should use, not the NRT that are only for operational use.

Category: Data and processing

After conducting MADT studies based on RIO05 MDT product, now new works based on the CNES-CLS09 MDT are different, is it normal?

Yes, this difference can be explained by the MDT reference change that took place in June 2010: new improved version of the Combined Mean Dynamic Topography called MDT_CNES-CLS09 replaces the old RIO05 MDT reference.

The CNES-CLS09 MDT was built using the same method than RIO05 MDT and compared to the previous RIO05 field, its main improvements are :

• a better resolution grid (1/4° instead of 1/2° for RIO05) due to the use of more buoys
• a recent geoid based on 4,5 years of GRACE data, where wavelength shorter than 300 km are filtered out
• an updated dataset of drifting buoy velocities and dynamic heights
• an improved Ekman model to extract the geostrophic component of the buoy velocities
• an improved processing method of the dynamic heights.

You will find more information via links MDT details and MDT description

Category: Data and processing

How get/exploit Aviso products available in NetCDF format (.nc)?

Firstly, to get Aviso data in NetCDF format, download the corresponding files via FTP client (for example FileZilla the Open Source FTP client filezilla-project.org). After downloading, in order to get the corresponding NetCDF files ('.nc'), extract the files with a 'gunzip' command if extension is '.gz' , use the 'tar -xzvf' commande for extensions '.tar.gz'

Aviso proposes two ways to directly work with Aviso NetCDF data :

1. Use Aviso data extraction tool: see Data access services.
2. Use OPeNDAP tool: refer to OPeNDAP details, in particular go to the OPeNDAP website or OPeNDAP support.

Category: Data and processing

Aviso only provides gridded Delayed-Time products (DT MSLA and MADT) until mid January 2011. Is there a time delay in data release?

Nowadays, the Ssalto/Duacs gridded products in Delayed-Time are updated several times a year and this operation produces files from October 14th 1992 to the later processed data (all valid data from all satellites are needed to produce a combined DT product).

The gridded DT data are dated until January 19th, 2011. Since April 2011, these DT maps have a daily temporal resolution while present ones were released with a weekly temporal resolution (gridded NRT products are already released with an equivalent daily resolution).

 

Category: Data and processing

Show all / Hide all

How could I log on with my account?

Symptoms:

• I could not download the data from the authenticated ftp folders.
• I don't see any place to enter my username/password.
• I see only data from 1992 in the folders/I don't see the data I'm interested in the folders, how can I get more recent data?

To access to data from authenticated ftp site, firstly you must have a personal account provided by the Aviso Users Services after you have properly filled the registration form available online.

Secondly, you have to connect with your personal account information by proceeding as follows:

With a web browser:

• Using Firefox: WRITE ftp://yourLOGIN:yourPASSWORD@ftp.aviso.oceanobs.com into the URL bar.
• OR using Internet Explorer: WRITE ftp://ftp.aviso.oceanobs.com, click on Page, choose 'Open FTP Site in Windows Explorer' then enter your login and password.

Then browse 'global' directory ... (DO NOT browse into the 'pub' folder, it leads you to the anonymous account where only data from October, 14th 1992 to December, 30th 1992 are available).

Otherwise, see below the other ways to log in:

• by using a FTP client (FileZilla for example): enter 'ftp.aviso.oceanobs.com' in the hostname, then your login and password in the appropriate boxes.
• By using Unix commands: write 'ftp ftp.aviso.oceanobs.com' (and then enter your login/password). You can also use the command line "wget --no-proxy" with the suitable options (--user=YOUR_LOGIN  --password=YOUR_PASSWORD).

Some parameters settings have maybe to be checked: passive mode, proxy settings...

Category: Data and processing

What is the difference between the two file formats available in the FTP folders for the DUACS Along-Track data?

Since the end of September 2011, the DUACS Along-track products in Delayed-Time are available in a new format, making the data easier to use. This new format follows the standard NetCDF CF-1.4 (Climate and Forecast).
Each variable in the file has an associated description of what it represents, including physical units if appropriate, and that each value can be located in space and time: a variable refers to data stored in the file as a vector.

Filename nomenclature, paths and access rights are the same: the new format CF files are stored in folders suffixed with "_cf" at the same levels as folders containing older format files. Example: in the FTP folder ‘global/dt/upd/sla/’, folders j2/ and j2_cf/ contain respectively data in the old format and new format CF for the Jason-2 mission.

Note that the DUACS Along-Track products with the old format will still be distributed until the end of January 2012, leaving the users time to adapt to the new format CF. Beyond that date, the new format CF will be the only available format.

Category: Data and processing

Why do I observe a bias between Rio05 and RioMed around each side of Gibraltar Strait ?

The difference in mean dynamic topograpy between the Mediterranean Sea and the Atlantic Ocean in the two different products of MDT that we distribute (RioMed and Rio05) is due to the fact that both surfaces have been computed starting from different large scale first guess :

• In the Rio05 MDT, the first guess is a merging between the Levitus climatology at low and mid-latitudes and, at higher latitude, a large scale MDT based on a Mean Sea Surface (CLS01) minus a geoid model (EIGEN-GRACE3S) filtered at 400km resolution. The average of the RIO05 MDT over the latitude range between 40°N and 40°S is equal to the average over the same area of the Levitus climatology.
• On the other hand, the RioMed MDT was constructed starting from the mean of the MFSTEP Mediterranean forecasting model . The average over the Mediterranean Sea of the RioMed MDT is thus equal to the average of the MFSTEP MDT.

There is no reason why these two constants should be the same, and that is why you observe a bias between the Rio05 MDT and the RioMed MDT.

To merge the two fields, you might consider that the two constants just east of Gibraltar are about the same that the seal level just west of Gibraltar, but I have no clue of how best to deal with this.

A solution will be given in the future by the use of Goce geoid data : then the MDT may be computed for the Atlantic Ocean and for the Mediterranean Sea starting from the same first guess (MSS-Goce, filtered at some spatial wavelength).

Category: Data and processing

What are the differences between Jason-1 GDR-b and GDR-c?

 The GDR-c has a different file naming convention and has improved corrections.

• The file extension .NASA and .CNES are not added to the GDR-c file names,
• There are two new parameters, pseudo_datation_bias_corr_ku and mdt,
• Better orbit, sea stat bias, ionospheric range, inverse barometer, and tidal corrections,
• Improved rain and ice flags.

An updated user handbook is provided:

• SMM-MU-M5-OP-13184-CN, Edition 4.1, October 2008 
  Aviso and PoDaac User Handbook - IGDR and GDR Jason-1 Products, pdf version (2 Mo).

Category: Data and processing

What tide models are used in the merged sea level anomaly data?

For Jason-1&2 as well as for Envisat data, the tide models are

• ocean tide is deduced from GOT00.V2 model (R. Ray code) (load tide is included)
• earth tide is deduced from the Cartwright tide potential
• polar tide is deduced  from equilibrium model (IERS conventions)

Each tide data is interpolated on the altimeter data position (time/space interpolation) and subtracted to the altimeter measurement. The same tidal solutions are applied to each mission in order to minimize inter-mission biases.   Independently, multi-mission inter-calibration process (crossover adjustment using Jason-2 as reference) is done considering the data corrected from all geophysical corrections (tides, tropospheric and ionospheric effects, high frequency signal aliasing effect, ...).

Category: Data and processing

How is Mean Dynamic Topography calculated ?

The MDT used for Aviso products is called CNES-CLS09_v1.1 and is computed combining several steps:

• a direct method estimates at large scale a MDT using the MSS and the EIGEN-GRGS.RL02.MEAN-FIELD (this model is based on 4.5 years of GRACE data).
• a synthetic method where sea level anomalies are subtracted from in-situ measurements of the ocean state (dynamic heights over the 1993-2007 period and geostrophic surface velocities over the 1993-2008 period) to compute synthetic estimates of the MDT and the corresponding geostrophic surface velocities. 
• Then, using a multivariate objective analysis, the synthetic estimates of heights and velocities are computed and used to improve the first guess of the MDT at large scale. It's the calculation of the CMDT.
  Finally, a validation is done by comparing the MDT_CNES-CLS09 with others MDT.

The MDT is used to calculate the Absolute Dynamic Topography (ADT) according to ADT = MDT + SLA.

Further information and distribution: Combined mean dynamic topography

Category: Data and processing

Mean Dynamic Topography, Mean Sea Surface, Mean Sea Level or mean profile?

There are several means used in altimetry processing or altimetry-based studies.
The MSS (Mean Sea Surface) is a mean of several years of altimetry data, eventually using several satellites. It is a reference surface (thus you have values on a lon/lat grid).
The MDT (Mean Dynamic Topography) is the previous surface minus the geoid. It is the ocean topography due to the mean currents. It is also a reference surface.
The Mean profile is the mean of an altimeter measurements along the satellite's tracks. It can be used as reference for computing SLA for the same satellite. It is an along-track data (not distributed)
The MSL (Mean Sea Level) is a trend (a number of mm/year), often given with respect to time as a curve.

Category: Data and processing

What are the conditions to get Aviso data?

All Aviso data are free of charge for non-commercial use, whether online (FTP, Opendap) or on media (DVD).

The request must be justified by a short text explaining how the user plan to use the data. This helps Aviso in refining the available data to be better tuned to the users' needs.

For a first level of education, the best is to work with online maps (gif format), e.g. the one available through the Argonautica project, or with the Live Access Server.

Category: Data and processing

How are geostrophic velocities computed at the Equator?

In the geostrophic velocity data distributed by Aviso today, current velocities at the equator are computed using second derivative methode inspired from Lagerloef et al. [1999]. This method is quite usual. In a 5° band around the Equator, a connection is computed to ensure continuity with classical geostrophy.

Lagerloef, G.S.E.; G.T. Mitchum, R.B. Lukas, P.P. Niiler, 1999: Tropical Pacific near-surface currents estimated from altimeter, wind, and drifter data, J. Geophys. Res., 104, 23313-23326

Category: Data and processing

What are the differences between ADT and SLA products?

Absolute Dynamic Topography (ADT) and Sea Level Anomaly (SLA) are defined by:
ADT=MDT+SLA
where MDT is Mean Dynamic Topography.

Category: Data and processing

Could you send me an explanation of the times associated with your gridded data?

The gridded data are computed for a given day. Each map represents the situation on the day indicated in its name. It is not a mean over the period between two files.
To generate these maps, computing methods based on objective analysis allow us to interpolate data in time and in space. In delayed time, the processing window used is centered, e.g. 6 weeks of data before and after the given date are taken into account. In near-real time, only the 6 weeks of data preceding the given day are considered.

Category: Data and processing

When I read two consecutive Ssalto/Duacs NRT SLA files, some fields are exactly the same but the SLA filed is different. Why ?

For example, in the two Ssalto/Duacs NRT Sea Level Anomaly NetCDF files, res_oer_j2_21549_21569.nc contains the data from 21549 julian day to 21569 julian day, while the file res_oer_j2_21550_21570.nc contains the data from 21550 julian day to 21570 julian day. The two numbers in the file name indicate the period covered by the file. That's why you can find the same track in the two files.

The different value you observe on the same point between the two files comes from the time window used for the data processing. The file res_oer_j2_21549_21569.nc was generated on 21570 julian day when only data until 21569 julian day were available. One day later (file res_oer_j2_21550_21570.nc), data for day 21570 were available and the time window was shifted by one day.

This time window mainly impacts the correction of long wavelength signal such as the error on orbit determination and the residual error on high-frequency long-wavelength signal aliasing correction. The estimation of this correction for a given day is more accurate when data in the future (and of course also in the past) are available (i.e. when we are in similar condition to delayed time processing). In this way, we consider that data for a given day should better be read in the file starting from this day when possible.

 

Category: Data and processing

In the gridded data, to which point lat/lon refer to?

Each value has been computed exactly on the Lat/Lon values
defined by LatLonMin and LatLonStep , as specified in the metadata of the NetCDF file.

Explanation is given is section 4.3 of the handbook "Ssalto/Duacs User Handbook: (M)SLA and (M)ADT Near-Real Time and Delayed Time products" that you can find here

Category: Data and processing

Where can I find tables showing the correspondance date / cycle and pass number?

On-line tools on this website make the correspondance between hour and calendar date and cycle and pass number.

Category: Data and processing

How to convert measurement time in local time ?

• For T/P, 86400xTim_Moy_1 + 10^-3xTim_Moy_2 + 10^-6xTim_Moy_3 = time (in seconds)
• For Jason-1&2, 86400xTime_day + 10^-3xTime_sec + 10^-6xTim_microsec = time (in seconds)

Time is given in UTC (Universal Time Coordinate, linked to Greenwich meridian), date in Julian day, which is the elapsed time between a reference epoch (January 1^st 1950, 0h0min0s for Cnes Julian days, January 1^st 1958, 0h0min0s for Nasa Julian days) and measurement date. To convert UTC time in local time, you have to convert it in date, hours, minutes, seconds, then in solar time and finally in your local time. We haven't got any conversion procedure.

Category: Data and processing

What is the impact of missing data in the Duacs products?

When one satellite is missing for a while, due to an incident example, the impact on the products is twofold:

• the along-track files are missing if the incident lasts for more than one day

• the impact on the generation of the map depends on the gap in the data: to generate these maps, computing methods based on objective analysis allow us to interpolate data in time and in space. In delayed time, the processing window used is centered, e.g. 6 weeks of data before and after the given date are taken into account. In near-real time, only the 6 weeks of data preeceding the given day are considered. The maps of formal errors (in the directory "err") will give an overview of the error induced by the missing data in the generation of the maps.

Category: Data and processing

We haven't receive T/P cycles 118/431/432, Jason-1 cycles 178 and 375 to 500

There were no data processed for Topex/Poseidon cycles 118, 431, 432 and for Jason-1 cycle 178, due to incidents.

The cycles 375 to 500 for Jason-1 are missing because the mission ended the repeat orbit phase (until 374) and began geodetic phase from cycle 500 (the cycles have been recomputed).

Category: Data and processing

What are Jason1&2, Topex/Poseidon data resolution?

Spatially, along the track, data are available every 7 km (measurements averaged on 1 second, or "1 Hz data"). GDR 10 or 20 Hz data also exists (depending on the satellite), but those are more noisy. Between tracks, for Jason-1&2 and Topex/Poseidon, the distance is up to 315 km at the Equator; for ERS and Envisat, it is about 80 km.
Gridded Aviso data are available at 1/3° on a Mercator grid.

Category: Data and processing

What are the characteristics of the reference ellipsoid you are using?

The Aviso products are referenced to the Topex/Poseidon ellipsoid (see the references below). So the Sea Surface Height of the MSS, CorSSH products are referenced to T/P ellipsoid, even for ERS-1, ERS-2, Envisat and Cryosat products. The SLA and MSLA products are not concerned since they are anomalies.

The reference ellipsoid is the first-order definition of the non-spherical shape of the Earth. For the GDRs, concerning Jason-1&2 GFO and Topex/Poseidon, the reference ellipsoid is:

• radius : 6378136.3
• inverse Earth flattening coefficient : 298.257

For ERS-1, ERS-2, Envisat, Cryosat-2:

• radius : 6378137
• inverse Earth flattening coefficient : 298.257223563

Category: Data and processing

On which platform/Operating system can I use Aviso data?

You can use Aviso data on every platform and Operating system, but the reading software may not be available for your precise configuration (see the available software).
Note that T/P GDR-Ms are coded in little endian format and Jason-1 GDRs in big endian format.

Category: Data and processing

I'd like data over a precise area (data extraction)

- for gridded NetCDF data, an online extraction tool is available
- for Topex/Poseidon and Jason-1 GDR, Aviso can perform extraction for a short period of time on a small area
For the other dataset, users have to perform their own extraction from the global or regional datasets

 

Category: Data and processing

Where can I find NRT products before January 2011?

Actually, there is no Near-Real Time (NRT) product before a certain date (January 20th, 2011 for most of them), in the same way as there is no DT product after this date.

DT gridded products which are more precise than NRT products, substitute for NRT products (more explanations into the Ssalto/Duacs User Handbook (PDF) subsection '4. Delayed Time Component').

Note that the availability date of NRT and DT products changes after each new delivery of Delayed-Time products, as a result, the availability of these dates brings forward (see question '[...] Is there a time delay in data release?')

If you want to work with 'historical' time series, the Delayed-time data are what you should use, not the NRT that are only for operational use.

Category: Data and processing

After conducting MADT studies based on RIO05 MDT product, now new works based on the CNES-CLS09 MDT are different, is it normal?

Yes, this difference can be explained by the MDT reference change that took place in June 2010: new improved version of the Combined Mean Dynamic Topography called MDT_CNES-CLS09 replaces the old RIO05 MDT reference.

The CNES-CLS09 MDT was built using the same method than RIO05 MDT and compared to the previous RIO05 field, its main improvements are :

• a better resolution grid (1/4° instead of 1/2° for RIO05) due to the use of more buoys
• a recent geoid based on 4,5 years of GRACE data, where wavelength shorter than 300 km are filtered out
• an updated dataset of drifting buoy velocities and dynamic heights
• an improved Ekman model to extract the geostrophic component of the buoy velocities
• an improved processing method of the dynamic heights.

You will find more information via links MDT details and MDT description

Category: Data and processing

How get/exploit Aviso products available in NetCDF format (.nc)?

Firstly, to get Aviso data in NetCDF format, download the corresponding files via FTP client (for example FileZilla the Open Source FTP client filezilla-project.org). After downloading, in order to get the corresponding NetCDF files ('.nc'), extract the files with a 'gunzip' command if extension is '.gz' , use the 'tar -xzvf' commande for extensions '.tar.gz'

Aviso proposes two ways to directly work with Aviso NetCDF data :

1. Use Aviso data extraction tool: see Data access services.
2. Use OPeNDAP tool: refer to OPeNDAP details, in particular go to the OPeNDAP website or OPeNDAP support.

Category: Data and processing

Aviso only provides gridded Delayed-Time products (DT MSLA and MADT) until mid January 2011. Is there a time delay in data release?

Nowadays, the Ssalto/Duacs gridded products in Delayed-Time are updated several times a year and this operation produces files from October 14th 1992 to the later processed data (all valid data from all satellites are needed to produce a combined DT product).

The gridded DT data are dated until January 19th, 2011. Since April 2011, these DT maps have a daily temporal resolution while present ones were released with a weekly temporal resolution (gridded NRT products are already released with an equivalent daily resolution).

 

Category: Data and processing

Show all / Hide all

How could I log on with my account?

Symptoms:

• I could not download the data from the authenticated ftp folders.
• I don't see any place to enter my username/password.
• I see only data from 1992 in the folders/I don't see the data I'm interested in the folders, how can I get more recent data?

To access to data from authenticated ftp site, firstly you must have a personal account provided by the Aviso Users Services after you have properly filled the registration form available online.

Secondly, you have to connect with your personal account information by proceeding as follows:

With a web browser:

• Using Firefox: WRITE ftp://yourLOGIN:yourPASSWORD@ftp.aviso.oceanobs.com into the URL bar.
• OR using Internet Explorer: WRITE ftp://ftp.aviso.oceanobs.com, click on Page, choose 'Open FTP Site in Windows Explorer' then enter your login and password.

Then browse 'global' directory ... (DO NOT browse into the 'pub' folder, it leads you to the anonymous account where only data from October, 14th 1992 to December, 30th 1992 are available).

Otherwise, see below the other ways to log in:

• by using a FTP client (FileZilla for example): enter 'ftp.aviso.oceanobs.com' in the hostname, then your login and password in the appropriate boxes.
• By using Unix commands: write 'ftp ftp.aviso.oceanobs.com' (and then enter your login/password). You can also use the command line "wget --no-proxy" with the suitable options (--user=YOUR_LOGIN  --password=YOUR_PASSWORD).

Some parameters settings have maybe to be checked: passive mode, proxy settings...

Category: Data and processing

What is the difference between the two file formats available in the FTP folders for the DUACS Along-Track data?

Since the end of September 2011, the DUACS Along-track products in Delayed-Time are available in a new format, making the data easier to use. This new format follows the standard NetCDF CF-1.4 (Climate and Forecast).
Each variable in the file has an associated description of what it represents, including physical units if appropriate, and that each value can be located in space and time: a variable refers to data stored in the file as a vector.

Filename nomenclature, paths and access rights are the same: the new format CF files are stored in folders suffixed with "_cf" at the same levels as folders containing older format files. Example: in the FTP folder ‘global/dt/upd/sla/’, folders j2/ and j2_cf/ contain respectively data in the old format and new format CF for the Jason-2 mission.

Note that the DUACS Along-Track products with the old format will still be distributed until the end of January 2012, leaving the users time to adapt to the new format CF. Beyond that date, the new format CF will be the only available format.

Category: Data and processing

Why do I observe a bias between Rio05 and RioMed around each side of Gibraltar Strait ?

The difference in mean dynamic topograpy between the Mediterranean Sea and the Atlantic Ocean in the two different products of MDT that we distribute (RioMed and Rio05) is due to the fact that both surfaces have been computed starting from different large scale first guess :

• In the Rio05 MDT, the first guess is a merging between the Levitus climatology at low and mid-latitudes and, at higher latitude, a large scale MDT based on a Mean Sea Surface (CLS01) minus a geoid model (EIGEN-GRACE3S) filtered at 400km resolution. The average of the RIO05 MDT over the latitude range between 40°N and 40°S is equal to the average over the same area of the Levitus climatology.
• On the other hand, the RioMed MDT was constructed starting from the mean of the MFSTEP Mediterranean forecasting model . The average over the Mediterranean Sea of the RioMed MDT is thus equal to the average of the MFSTEP MDT.

There is no reason why these two constants should be the same, and that is why you observe a bias between the Rio05 MDT and the RioMed MDT.

To merge the two fields, you might consider that the two constants just east of Gibraltar are about the same that the seal level just west of Gibraltar, but I have no clue of how best to deal with this.

A solution will be given in the future by the use of Goce geoid data : then the MDT may be computed for the Atlantic Ocean and for the Mediterranean Sea starting from the same first guess (MSS-Goce, filtered at some spatial wavelength).

Category: Data and processing

What are the differences between Jason-1 GDR-b and GDR-c?

 The GDR-c has a different file naming convention and has improved corrections.

• The file extension .NASA and .CNES are not added to the GDR-c file names,
• There are two new parameters, pseudo_datation_bias_corr_ku and mdt,
• Better orbit, sea stat bias, ionospheric range, inverse barometer, and tidal corrections,
• Improved rain and ice flags.

An updated user handbook is provided:

• SMM-MU-M5-OP-13184-CN, Edition 4.1, October 2008 
  Aviso and PoDaac User Handbook - IGDR and GDR Jason-1 Products, pdf version (2 Mo).

Category: Data and processing

What tide models are used in the merged sea level anomaly data?

For Jason-1&2 as well as for Envisat data, the tide models are

• ocean tide is deduced from GOT00.V2 model (R. Ray code) (load tide is included)
• earth tide is deduced from the Cartwright tide potential
• polar tide is deduced  from equilibrium model (IERS conventions)

Each tide data is interpolated on the altimeter data position (time/space interpolation) and subtracted to the altimeter measurement. The same tidal solutions are applied to each mission in order to minimize inter-mission biases.   Independently, multi-mission inter-calibration process (crossover adjustment using Jason-2 as reference) is done considering the data corrected from all geophysical corrections (tides, tropospheric and ionospheric effects, high frequency signal aliasing effect, ...).

Category: Data and processing

How is Mean Dynamic Topography calculated ?

The MDT used for Aviso products is called CNES-CLS09_v1.1 and is computed combining several steps:

• a direct method estimates at large scale a MDT using the MSS and the EIGEN-GRGS.RL02.MEAN-FIELD (this model is based on 4.5 years of GRACE data).
• a synthetic method where sea level anomalies are subtracted from in-situ measurements of the ocean state (dynamic heights over the 1993-2007 period and geostrophic surface velocities over the 1993-2008 period) to compute synthetic estimates of the MDT and the corresponding geostrophic surface velocities. 
• Then, using a multivariate objective analysis, the synthetic estimates of heights and velocities are computed and used to improve the first guess of the MDT at large scale. It's the calculation of the CMDT.
  Finally, a validation is done by comparing the MDT_CNES-CLS09 with others MDT.

The MDT is used to calculate the Absolute Dynamic Topography (ADT) according to ADT = MDT + SLA.

Further information and distribution: Combined mean dynamic topography

Category: Data and processing

Mean Dynamic Topography, Mean Sea Surface, Mean Sea Level or mean profile?

There are several means used in altimetry processing or altimetry-based studies.
The MSS (Mean Sea Surface) is a mean of several years of altimetry data, eventually using several satellites. It is a reference surface (thus you have values on a lon/lat grid).
The MDT (Mean Dynamic Topography) is the previous surface minus the geoid. It is the ocean topography due to the mean currents. It is also a reference surface.
The Mean profile is the mean of an altimeter measurements along the satellite's tracks. It can be used as reference for computing SLA for the same satellite. It is an along-track data (not distributed)
The MSL (Mean Sea Level) is a trend (a number of mm/year), often given with respect to time as a curve.

Category: Data and processing

What are the conditions to get Aviso data?

All Aviso data are free of charge for non-commercial use, whether online (FTP, Opendap) or on media (DVD).

The request must be justified by a short text explaining how the user plan to use the data. This helps Aviso in refining the available data to be better tuned to the users' needs.

For a first level of education, the best is to work with online maps (gif format), e.g. the one available through the Argonautica project, or with the Live Access Server.

Category: Data and processing

How are geostrophic velocities computed at the Equator?

In the geostrophic velocity data distributed by Aviso today, current velocities at the equator are computed using second derivative methode inspired from Lagerloef et al. [1999]. This method is quite usual. In a 5° band around the Equator, a connection is computed to ensure continuity with classical geostrophy.

Lagerloef, G.S.E.; G.T. Mitchum, R.B. Lukas, P.P. Niiler, 1999: Tropical Pacific near-surface currents estimated from altimeter, wind, and drifter data, J. Geophys. Res., 104, 23313-23326

Category: Data and processing

What are the differences between ADT and SLA products?

Absolute Dynamic Topography (ADT) and Sea Level Anomaly (SLA) are defined by:
ADT=MDT+SLA
where MDT is Mean Dynamic Topography.

Category: Data and processing

Could you send me an explanation of the times associated with your gridded data?

The gridded data are computed for a given day. Each map represents the situation on the day indicated in its name. It is not a mean over the period between two files.
To generate these maps, computing methods based on objective analysis allow us to interpolate data in time and in space. In delayed time, the processing window used is centered, e.g. 6 weeks of data before and after the given date are taken into account. In near-real time, only the 6 weeks of data preceding the given day are considered.

Category: Data and processing

When I read two consecutive Ssalto/Duacs NRT SLA files, some fields are exactly the same but the SLA filed is different. Why ?

For example, in the two Ssalto/Duacs NRT Sea Level Anomaly NetCDF files, res_oer_j2_21549_21569.nc contains the data from 21549 julian day to 21569 julian day, while the file res_oer_j2_21550_21570.nc contains the data from 21550 julian day to 21570 julian day. The two numbers in the file name indicate the period covered by the file. That's why you can find the same track in the two files.

The different value you observe on the same point between the two files comes from the time window used for the data processing. The file res_oer_j2_21549_21569.nc was generated on 21570 julian day when only data until 21569 julian day were available. One day later (file res_oer_j2_21550_21570.nc), data for day 21570 were available and the time window was shifted by one day.

This time window mainly impacts the correction of long wavelength signal such as the error on orbit determination and the residual error on high-frequency long-wavelength signal aliasing correction. The estimation of this correction for a given day is more accurate when data in the future (and of course also in the past) are available (i.e. when we are in similar condition to delayed time processing). In this way, we consider that data for a given day should better be read in the file starting from this day when possible.

 

Category: Data and processing

In the gridded data, to which point lat/lon refer to?

Each value has been computed exactly on the Lat/Lon values
defined by LatLonMin and LatLonStep , as specified in the metadata of the NetCDF file.

Explanation is given is section 4.3 of the handbook "Ssalto/Duacs User Handbook: (M)SLA and (M)ADT Near-Real Time and Delayed Time products" that you can find here

Category: Data and processing

Where can I find tables showing the correspondance date / cycle and pass number?

On-line tools on this website make the correspondance between hour and calendar date and cycle and pass number.

Category: Data and processing

How to convert measurement time in local time ?

• For T/P, 86400xTim_Moy_1 + 10^-3xTim_Moy_2 + 10^-6xTim_Moy_3 = time (in seconds)
• For Jason-1&2, 86400xTime_day + 10^-3xTime_sec + 10^-6xTim_microsec = time (in seconds)

Time is given in UTC (Universal Time Coordinate, linked to Greenwich meridian), date in Julian day, which is the elapsed time between a reference epoch (January 1^st 1950, 0h0min0s for Cnes Julian days, January 1^st 1958, 0h0min0s for Nasa Julian days) and measurement date. To convert UTC time in local time, you have to convert it in date, hours, minutes, seconds, then in solar time and finally in your local time. We haven't got any conversion procedure.

Category: Data and processing

What is the impact of missing data in the Duacs products?

When one satellite is missing for a while, due to an incident example, the impact on the products is twofold:

• the along-track files are missing if the incident lasts for more than one day

• the impact on the generation of the map depends on the gap in the data: to generate these maps, computing methods based on objective analysis allow us to interpolate data in time and in space. In delayed time, the processing window used is centered, e.g. 6 weeks of data before and after the given date are taken into account. In near-real time, only the 6 weeks of data preeceding the given day are considered. The maps of formal errors (in the directory "err") will give an overview of the error induced by the missing data in the generation of the maps.

Category: Data and processing

We haven't receive T/P cycles 118/431/432, Jason-1 cycles 178 and 375 to 500

There were no data processed for Topex/Poseidon cycles 118, 431, 432 and for Jason-1 cycle 178, due to incidents.

The cycles 375 to 500 for Jason-1 are missing because the mission ended the repeat orbit phase (until 374) and began geodetic phase from cycle 500 (the cycles have been recomputed).

Category: Data and processing

What are Jason1&2, Topex/Poseidon data resolution?

Spatially, along the track, data are available every 7 km (measurements averaged on 1 second, or "1 Hz data"). GDR 10 or 20 Hz data also exists (depending on the satellite), but those are more noisy. Between tracks, for Jason-1&2 and Topex/Poseidon, the distance is up to 315 km at the Equator; for ERS and Envisat, it is about 80 km.
Gridded Aviso data are available at 1/3° on a Mercator grid.

Category: Data and processing

What are the characteristics of the reference ellipsoid you are using?

The Aviso products are referenced to the Topex/Poseidon ellipsoid (see the references below). So the Sea Surface Height of the MSS, CorSSH products are referenced to T/P ellipsoid, even for ERS-1, ERS-2, Envisat and Cryosat products. The SLA and MSLA products are not concerned since they are anomalies.

The reference ellipsoid is the first-order definition of the non-spherical shape of the Earth. For the GDRs, concerning Jason-1&2 GFO and Topex/Poseidon, the reference ellipsoid is:

• radius : 6378136.3
• inverse Earth flattening coefficient : 298.257

For ERS-1, ERS-2, Envisat, Cryosat-2:

• radius : 6378137
• inverse Earth flattening coefficient : 298.257223563

Category: Data and processing

On which platform/Operating system can I use Aviso data?

You can use Aviso data on every platform and Operating system, but the reading software may not be available for your precise configuration (see the available software).
Note that T/P GDR-Ms are coded in little endian format and Jason-1 GDRs in big endian format.

Category: Data and processing

I'd like data over a precise area (data extraction)

- for gridded NetCDF data, an online extraction tool is available
- for Topex/Poseidon and Jason-1 GDR, Aviso can perform extraction for a short period of time on a small area
For the other dataset, users have to perform their own extraction from the global or regional datasets

 

Category: Data and processing

Where can I find NRT products before January 2011?

Actually, there is no Near-Real Time (NRT) product before a certain date (January 20th, 2011 for most of them), in the same way as there is no DT product after this date.

DT gridded products which are more precise than NRT products, substitute for NRT products (more explanations into the Ssalto/Duacs User Handbook (PDF) subsection '4. Delayed Time Component').

Note that the availability date of NRT and DT products changes after each new delivery of Delayed-Time products, as a result, the availability of these dates brings forward (see question '[...] Is there a time delay in data release?')

If you want to work with 'historical' time series, the Delayed-time data are what you should use, not the NRT that are only for operational use.

Category: Data and processing

After conducting MADT studies based on RIO05 MDT product, now new works based on the CNES-CLS09 MDT are different, is it normal?

Yes, this difference can be explained by the MDT reference change that took place in June 2010: new improved version of the Combined Mean Dynamic Topography called MDT_CNES-CLS09 replaces the old RIO05 MDT reference.

The CNES-CLS09 MDT was built using the same method than RIO05 MDT and compared to the previous RIO05 field, its main improvements are :

• a better resolution grid (1/4° instead of 1/2° for RIO05) due to the use of more buoys
• a recent geoid based on 4,5 years of GRACE data, where wavelength shorter than 300 km are filtered out
• an updated dataset of drifting buoy velocities and dynamic heights
• an improved Ekman model to extract the geostrophic component of the buoy velocities
• an improved processing method of the dynamic heights.

You will find more information via links MDT details and MDT description

Category: Data and processing

How get/exploit Aviso products available in NetCDF format (.nc)?

Firstly, to get Aviso data in NetCDF format, download the corresponding files via FTP client (for example FileZilla the Open Source FTP client filezilla-project.org). After downloading, in order to get the corresponding NetCDF files ('.nc'), extract the files with a 'gunzip' command if extension is '.gz' , use the 'tar -xzvf' commande for extensions '.tar.gz'

Aviso proposes two ways to directly work with Aviso NetCDF data :

1. Use Aviso data extraction tool: see Data access services.
2. Use OPeNDAP tool: refer to OPeNDAP details, in particular go to the OPeNDAP website or OPeNDAP support.

Category: Data and processing

Aviso only provides gridded Delayed-Time products (DT MSLA and MADT) until mid January 2011. Is there a time delay in data release?

Nowadays, the Ssalto/Duacs gridded products in Delayed-Time are updated several times a year and this operation produces files from October 14th 1992 to the later processed data (all valid data from all satellites are needed to produce a combined DT product).

The gridded DT data are dated until January 19th, 2011. Since April 2011, these DT maps have a daily temporal resolution while present ones were released with a weekly temporal resolution (gridded NRT products are already released with an equivalent daily resolution).

 

Category: Data and processing

Show all / Hide all

How could I log on with my account?

Symptoms:

• I could not download the data from the authenticated ftp folders.
• I don't see any place to enter my username/password.
• I see only data from 1992 in the folders/I don't see the data I'm interested in the folders, how can I get more recent data?

To access to data from authenticated ftp site, firstly you must have a personal account provided by the Aviso Users Services after you have properly filled the registration form available online.

Secondly, you have to connect with your personal account information by proceeding as follows:

With a web browser:

• Using Firefox: WRITE ftp://yourLOGIN:yourPASSWORD@ftp.aviso.oceanobs.com into the URL bar.
• OR using Internet Explorer: WRITE ftp://ftp.aviso.oceanobs.com, click on Page, choose 'Open FTP Site in Windows Explorer' then enter your login and password.

Then browse 'global' directory ... (DO NOT browse into the 'pub' folder, it leads you to the anonymous account where only data from October, 14th 1992 to December, 30th 1992 are available).

Otherwise, see below the other ways to log in:

• by using a FTP client (FileZilla for example): enter 'ftp.aviso.oceanobs.com' in the hostname, then your login and password in the appropriate boxes.
• By using Unix commands: write 'ftp ftp.aviso.oceanobs.com' (and then enter your login/password). You can also use the command line "wget --no-proxy" with the suitable options (--user=YOUR_LOGIN  --password=YOUR_PASSWORD).

Some parameters settings have maybe to be checked: passive mode, proxy settings...

Category: Data and processing

What is the difference between the two file formats available in the FTP folders for the DUACS Along-Track data?

Since the end of September 2011, the DUACS Along-track products in Delayed-Time are available in a new format, making the data easier to use. This new format follows the standard NetCDF CF-1.4 (Climate and Forecast).
Each variable in the file has an associated description of what it represents, including physical units if appropriate, and that each value can be located in space and time: a variable refers to data stored in the file as a vector.

Filename nomenclature, paths and access rights are the same: the new format CF files are stored in folders suffixed with "_cf" at the same levels as folders containing older format files. Example: in the FTP folder ‘global/dt/upd/sla/’, folders j2/ and j2_cf/ contain respectively data in the old format and new format CF for the Jason-2 mission.

Note that the DUACS Along-Track products with the old format will still be distributed until the end of January 2012, leaving the users time to adapt to the new format CF. Beyond that date, the new format CF will be the only available format.

Category: Data and processing

Why do I observe a bias between Rio05 and RioMed around each side of Gibraltar Strait ?

The difference in mean dynamic topograpy between the Mediterranean Sea and the Atlantic Ocean in the two different products of MDT that we distribute (RioMed and Rio05) is due to the fact that both surfaces have been computed starting from different large scale first guess :

• In the Rio05 MDT, the first guess is a merging between the Levitus climatology at low and mid-latitudes and, at higher latitude, a large scale MDT based on a Mean Sea Surface (CLS01) minus a geoid model (EIGEN-GRACE3S) filtered at 400km resolution. The average of the RIO05 MDT over the latitude range between 40°N and 40°S is equal to the average over the same area of the Levitus climatology.
• On the other hand, the RioMed MDT was constructed starting from the mean of the MFSTEP Mediterranean forecasting model . The average over the Mediterranean Sea of the RioMed MDT is thus equal to the average of the MFSTEP MDT.

There is no reason why these two constants should be the same, and that is why you observe a bias between the Rio05 MDT and the RioMed MDT.

To merge the two fields, you might consider that the two constants just east of Gibraltar are about the same that the seal level just west of Gibraltar, but I have no clue of how best to deal with this.

A solution will be given in the future by the use of Goce geoid data : then the MDT may be computed for the Atlantic Ocean and for the Mediterranean Sea starting from the same first guess (MSS-Goce, filtered at some spatial wavelength).

Category: Data and processing

What are the differences between Jason-1 GDR-b and GDR-c?

 The GDR-c has a different file naming convention and has improved corrections.

• The file extension .NASA and .CNES are not added to the GDR-c file names,
• There are two new parameters, pseudo_datation_bias_corr_ku and mdt,
• Better orbit, sea stat bias, ionospheric range, inverse barometer, and tidal corrections,
• Improved rain and ice flags.

An updated user handbook is provided:

• SMM-MU-M5-OP-13184-CN, Edition 4.1, October 2008 
  Aviso and PoDaac User Handbook - IGDR and GDR Jason-1 Products, pdf version (2 Mo).

Category: Data and processing

What tide models are used in the merged sea level anomaly data?

For Jason-1&2 as well as for Envisat data, the tide models are

• ocean tide is deduced from GOT00.V2 model (R. Ray code) (load tide is included)
• earth tide is deduced from the Cartwright tide potential
• polar tide is deduced  from equilibrium model (IERS conventions)

Each tide data is interpolated on the altimeter data position (time/space interpolation) and subtracted to the altimeter measurement. The same tidal solutions are applied to each mission in order to minimize inter-mission biases.   Independently, multi-mission inter-calibration process (crossover adjustment using Jason-2 as reference) is done considering the data corrected from all geophysical corrections (tides, tropospheric and ionospheric effects, high frequency signal aliasing effect, ...).

Category: Data and processing

How is Mean Dynamic Topography calculated ?

The MDT used for Aviso products is called CNES-CLS09_v1.1 and is computed combining several steps:

• a direct method estimates at large scale a MDT using the MSS and the EIGEN-GRGS.RL02.MEAN-FIELD (this model is based on 4.5 years of GRACE data).
• a synthetic method where sea level anomalies are subtracted from in-situ measurements of the ocean state (dynamic heights over the 1993-2007 period and geostrophic surface velocities over the 1993-2008 period) to compute synthetic estimates of the MDT and the corresponding geostrophic surface velocities. 
• Then, using a multivariate objective analysis, the synthetic estimates of heights and velocities are computed and used to improve the first guess of the MDT at large scale. It's the calculation of the CMDT.
  Finally, a validation is done by comparing the MDT_CNES-CLS09 with others MDT.

The MDT is used to calculate the Absolute Dynamic Topography (ADT) according to ADT = MDT + SLA.

Further information and distribution: Combined mean dynamic topography

Category: Data and processing

Mean Dynamic Topography, Mean Sea Surface, Mean Sea Level or mean profile?

There are several means used in altimetry processing or altimetry-based studies.
The MSS (Mean Sea Surface) is a mean of several years of altimetry data, eventually using several satellites. It is a reference surface (thus you have values on a lon/lat grid).
The MDT (Mean Dynamic Topography) is the previous surface minus the geoid. It is the ocean topography due to the mean currents. It is also a reference surface.
The Mean profile is the mean of an altimeter measurements along the satellite's tracks. It can be used as reference for computing SLA for the same satellite. It is an along-track data (not distributed)
The MSL (Mean Sea Level) is a trend (a number of mm/year), often given with respect to time as a curve.

Category: Data and processing

What are the conditions to get Aviso data?

All Aviso data are free of charge for non-commercial use, whether online (FTP, Opendap) or on media (DVD).

The request must be justified by a short text explaining how the user plan to use the data. This helps Aviso in refining the available data to be better tuned to the users' needs.

For a first level of education, the best is to work with online maps (gif format), e.g. the one available through the Argonautica project, or with the Live Access Server.

Category: Data and processing

How are geostrophic velocities computed at the Equator?

In the geostrophic velocity data distributed by Aviso today, current velocities at the equator are computed using second derivative methode inspired from Lagerloef et al. [1999]. This method is quite usual. In a 5° band around the Equator, a connection is computed to ensure continuity with classical geostrophy.

Lagerloef, G.S.E.; G.T. Mitchum, R.B. Lukas, P.P. Niiler, 1999: Tropical Pacific near-surface currents estimated from altimeter, wind, and drifter data, J. Geophys. Res., 104, 23313-23326

Category: Data and processing

What are the differences between ADT and SLA products?

Absolute Dynamic Topography (ADT) and Sea Level Anomaly (SLA) are defined by:
ADT=MDT+SLA
where MDT is Mean Dynamic Topography.

Category: Data and processing

Could you send me an explanation of the times associated with your gridded data?

The gridded data are computed for a given day. Each map represents the situation on the day indicated in its name. It is not a mean over the period between two files.
To generate these maps, computing methods based on objective analysis allow us to interpolate data in time and in space. In delayed time, the processing window used is centered, e.g. 6 weeks of data before and after the given date are taken into account. In near-real time, only the 6 weeks of data preceding the given day are considered.

Category: Data and processing

When I read two consecutive Ssalto/Duacs NRT SLA files, some fields are exactly the same but the SLA filed is different. Why ?

For example, in the two Ssalto/Duacs NRT Sea Level Anomaly NetCDF files, res_oer_j2_21549_21569.nc contains the data from 21549 julian day to 21569 julian day, while the file res_oer_j2_21550_21570.nc contains the data from 21550 julian day to 21570 julian day. The two numbers in the file name indicate the period covered by the file. That's why you can find the same track in the two files.

The different value you observe on the same point between the two files comes from the time window used for the data processing. The file res_oer_j2_21549_21569.nc was generated on 21570 julian day when only data until 21569 julian day were available. One day later (file res_oer_j2_21550_21570.nc), data for day 21570 were available and the time window was shifted by one day.

This time window mainly impacts the correction of long wavelength signal such as the error on orbit determination and the residual error on high-frequency long-wavelength signal aliasing correction. The estimation of this correction for a given day is more accurate when data in the future (and of course also in the past) are available (i.e. when we are in similar condition to delayed time processing). In this way, we consider that data for a given day should better be read in the file starting from this day when possible.

 

Category: Data and processing

In the gridded data, to which point lat/lon refer to?

Each value has been computed exactly on the Lat/Lon values
defined by LatLonMin and LatLonStep , as specified in the metadata of the NetCDF file.

Explanation is given is section 4.3 of the handbook "Ssalto/Duacs User Handbook: (M)SLA and (M)ADT Near-Real Time and Delayed Time products" that you can find here

Category: Data and processing

Where can I find tables showing the correspondance date / cycle and pass number?

On-line tools on this website make the correspondance between hour and calendar date and cycle and pass number.

Category: Data and processing

How to convert measurement time in local time ?

• For T/P, 86400xTim_Moy_1 + 10^-3xTim_Moy_2 + 10^-6xTim_Moy_3 = time (in seconds)
• For Jason-1&2, 86400xTime_day + 10^-3xTime_sec + 10^-6xTim_microsec = time (in seconds)

Time is given in UTC (Universal Time Coordinate, linked to Greenwich meridian), date in Julian day, which is the elapsed time between a reference epoch (January 1^st 1950, 0h0min0s for Cnes Julian days, January 1^st 1958, 0h0min0s for Nasa Julian days) and measurement date. To convert UTC time in local time, you have to convert it in date, hours, minutes, seconds, then in solar time and finally in your local time. We haven't got any conversion procedure.

Category: Data and processing

What is the impact of missing data in the Duacs products?

When one satellite is missing for a while, due to an incident example, the impact on the products is twofold:

• the along-track files are missing if the incident lasts for more than one day

• the impact on the generation of the map depends on the gap in the data: to generate these maps, computing methods based on objective analysis allow us to interpolate data in time and in space. In delayed time, the processing window used is centered, e.g. 6 weeks of data before and after the given date are taken into account. In near-real time, only the 6 weeks of data preeceding the given day are considered. The maps of formal errors (in the directory "err") will give an overview of the error induced by the missing data in the generation of the maps.

Category: Data and processing

We haven't receive T/P cycles 118/431/432, Jason-1 cycles 178 and 375 to 500

There were no data processed for Topex/Poseidon cycles 118, 431, 432 and for Jason-1 cycle 178, due to incidents.

The cycles 375 to 500 for Jason-1 are missing because the mission ended the repeat orbit phase (until 374) and began geodetic phase from cycle 500 (the cycles have been recomputed).

Category: Data and processing

What are Jason1&2, Topex/Poseidon data resolution?

Spatially, along the track, data are available every 7 km (measurements averaged on 1 second, or "1 Hz data"). GDR 10 or 20 Hz data also exists (depending on the satellite), but those are more noisy. Between tracks, for Jason-1&2 and Topex/Poseidon, the distance is up to 315 km at the Equator; for ERS and Envisat, it is about 80 km.
Gridded Aviso data are available at 1/3° on a Mercator grid.

Category: Data and processing

What are the characteristics of the reference ellipsoid you are using?

The Aviso products are referenced to the Topex/Poseidon ellipsoid (see the references below). So the Sea Surface Height of the MSS, CorSSH products are referenced to T/P ellipsoid, even for ERS-1, ERS-2, Envisat and Cryosat products. The SLA and MSLA products are not concerned since they are anomalies.

The reference ellipsoid is the first-order definition of the non-spherical shape of the Earth. For the GDRs, concerning Jason-1&2 GFO and Topex/Poseidon, the reference ellipsoid is:

• radius : 6378136.3
• inverse Earth flattening coefficient : 298.257

For ERS-1, ERS-2, Envisat, Cryosat-2:

• radius : 6378137
• inverse Earth flattening coefficient : 298.257223563

Category: Data and processing

On which platform/Operating system can I use Aviso data?

You can use Aviso data on every platform and Operating system, but the reading software may not be available for your precise configuration (see the available software).
Note that T/P GDR-Ms are coded in little endian format and Jason-1 GDRs in big endian format.

Category: Data and processing

I'd like data over a precise area (data extraction)

- for gridded NetCDF data, an online extraction tool is available
- for Topex/Poseidon and Jason-1 GDR, Aviso can perform extraction for a short period of time on a small area
For the other dataset, users have to perform their own extraction from the global or regional datasets

 

Category: Data and processing

Where can I find NRT products before January 2011?

Actually, there is no Near-Real Time (NRT) product before a certain date (January 20th, 2011 for most of them), in the same way as there is no DT product after this date.

DT gridded products which are more precise than NRT products, substitute for NRT products (more explanations into the Ssalto/Duacs User Handbook (PDF) subsection '4. Delayed Time Component').

Note that the availability date of NRT and DT products changes after each new delivery of Delayed-Time products, as a result, the availability of these dates brings forward (see question '[...] Is there a time delay in data release?')

If you want to work with 'historical' time series, the Delayed-time data are what you should use, not the NRT that are only for operational use.

Category: Data and processing

After conducting MADT studies based on RIO05 MDT product, now new works based on the CNES-CLS09 MDT are different, is it normal?

Yes, this difference can be explained by the MDT reference change that took place in June 2010: new improved version of the Combined Mean Dynamic Topography called MDT_CNES-CLS09 replaces the old RIO05 MDT reference.

The CNES-CLS09 MDT was built using the same method than RIO05 MDT and compared to the previous RIO05 field, its main improvements are :

• a better resolution grid (1/4° instead of 1/2° for RIO05) due to the use of more buoys
• a recent geoid based on 4,5 years of GRACE data, where wavelength shorter than 300 km are filtered out
• an updated dataset of drifting buoy velocities and dynamic heights
• an improved Ekman model to extract the geostrophic component of the buoy velocities
• an improved processing method of the dynamic heights.

You will find more information via links MDT details and MDT description

Category: Data and processing

How get/exploit Aviso products available in NetCDF format (.nc)?

Firstly, to get Aviso data in NetCDF format, download the corresponding files via FTP client (for example FileZilla the Open Source FTP client filezilla-project.org). After downloading, in order to get the corresponding NetCDF files ('.nc'), extract the files with a 'gunzip' command if extension is '.gz' , use the 'tar -xzvf' commande for extensions '.tar.gz'

Aviso proposes two ways to directly work with Aviso NetCDF data :

1. Use Aviso data extraction tool: see Data access services.
2. Use OPeNDAP tool: refer to OPeNDAP details, in particular go to the OPeNDAP website or OPeNDAP support.

Category: Data and processing

Aviso only provides gridded Delayed-Time products (DT MSLA and MADT) until mid January 2011. Is there a time delay in data release?

Nowadays, the Ssalto/Duacs gridded products in Delayed-Time are updated several times a year and this operation produces files from October 14th 1992 to the later processed data (all valid data from all satellites are needed to produce a combined DT product).

The gridded DT data are dated until January 19th, 2011. Since April 2011, these DT maps have a daily temporal resolution while present ones were released with a weekly temporal resolution (gridded NRT products are already released with an equivalent daily resolution).

 

Category: Data and processing