Thermodynamical Initial Guess Retrieval (TIGR)

What is the TIGR dataset ?

The Thermodynamic Initial Guess Retrieval (TIGR) dataset, in its latest version, is a climatological library of 2311 atmospheric situations representative of the Earth’s atmosphere, selected by statistical methods from 80 000 radiosonde reports (Chédin et al., 1985; Achard, 1991; Chevallier et al., 1998).

Each situation is described, from the surface to the top of the atmosphere (0.0026hPa), by the values of temperature, water vapor and ozone concentrations on a given pressure grid.

There are several data sets associated with the above-mentioned atmospheric profiles:
            – the radiances (or brightness temperature equivalents) that a satellite-borne instrument (e.g. TOVS, ATOVS, AIRS, IASI, Modis, Seviri, IIR, etc…) would have observed under several observing conditions (satellite zenith angle, surface characteristics, etc…).
            – the transmittances and Jacobians (partial derivatives of the brightness temperature with respect to the temperature, the gas concentration for H2O, O3, CO2, N2O CO, CH4, etc., the temperature and emissivity of the surface).
            – the radiances, transmittances, and Jacobian profiles are generated using the advanced radiative transfer model 4A, 4A/OP – Fast simulation of the atmospheric radiative transfer (aeris-data.fr)

TIGR: a bit of history

Since its creation (early 80’s) until today, TIGR, associated with direct or inverse radiative transfer models, has opened up and broadened many subjects related to the spatial observation of the Earth’s surface and atmosphere. This is true both from the point of view of physical, statistical and numerical approaches and from that of the atmospheric or climatological variables retrieved. This is evidenced by the numerous articles on subjects as varied as: Bayesian inversion of the radiative transfer equation, neural networks, copula classification, independent component analysis, temperature profiles, water vapour, GHGs, surface emissivities, surface temperature, forest fires, CO2 limb observation (ACE-FTS), aerosols, etc., without forgetting the assistance in the definition, monitoring and quality control of sounders or imagers in earth observation (https://ara.lmd.polytechnique.fr/).
To date, TIGR has been distributed to some 500 research, educational, private, operational and space meteorological institutes around the world.

Since 1985, the TIGR dataset knew several versions :

• TIGR : 398 atmospheres (1985)
• TIGR1 : 1207 atmospheres
• TIGR2 : 1761 atmospheres
• TIGR3 : 2311 atmospheres
  • new tropical atmospheres (profiles 1 to 872)
  • mid-lat and polar atmospheres unchanged (profiles 873 to 2311)
• TIGR2000 : 2311 atmospheres
  • TIGR3 +
  • The ozone profile is deduced from the Ugamp climatology (Li and Shine, 1995)
  • change in the longitude convention : -180:+180
• TIGR2000_6CORPS : 2311 atmospheres
  • ATMOS measurements have been used to improve the extrapolation of the water vapour towards the upper pressure levels.

The current distributed version of TIGR

A new version (dated May 2010) and referenced as ‘atm4atigr2000_v1.2_43lev‘ is now distributed whose main characteristics are given below.

• The air mass classification of the TIGR atmospheres

As in the previous version, the TIGR dataset includes 2311 atmospheric profiles. These atmospheric situations are now stratified by a hierarchical ascending classification into five types of air masses (Tropical, temperate -Midlat1-, cold temperate and summer polar -Midlat2-, very cold polar of the Northern Hemisphere -Polar1-, winter polar -Polar2-), according to their virtual temperature profiles (Achard, 1991; Chédin et al., 1994).

• Other features of the current version

• Pressure range: to take into account the ability of high spectral resolution instruments or limb viewing instruments to probe high atmospheric levels, the pressure range has been extended to 0.0026 hPa (instead of 0.05 hPa in the previous version). As a result, the number of levels is equal to 43 (instead of 40 in the previous versions).
• Temperature profiles: between 37 hPa and 0.05 hPa (levels 18 and 4 respectively), temperature profiles have been replaced by a monthly climatology (*) from the ERA_Interim reanalyses and then extrapolated up to 0.0026 hPa (level 1) based on the Ace_Scisat Instrument level 2 outputs.
• Water vapour profiles: between 380 hPa and 0.05 hPa (levels 32 and 4 respectively), water vapour profiles have been replaced by a monthly climatology (*) from the ERA-Interim reanalyses, and then extrapolated up to 0.0026 hPa (level 1) based on the Ace_Scisat Instrument level 2 outputs.
• Ozone profiles: originally based on an UGAMP climatology, the ozone profiles have been replaced between the surface and 0.05 hPa (levels 43 and 4, respectively) by a monthly climatology (*) from the ERA-Interim reanalyses, and then extrapolated up to 0.0026 hPa based on the Ace_Scisat Instrument level 2 outputs.
• Collocations with ERA-Interim climatology:  As mentioned above, collocations with ERA_Interim are necessary at low pressure levels where radiosondes do not provide the required values (e.g. ozone at lower pressures). The ERA-Interim reanalyses have a spatial resolution of 0.75° and outputs corresponding to 00h, 06h, 12h and 18h. We created a 12-year (1998 to 2009) monthly climatology of temperature, water vapor, and ozone profiles at these spatial and temporal resolutions.

Each initial TIGR profile is spatially and temporally collocated with the corresponding ERA-Interim climatology according to the following rules: i) spatial collocation is based on a simple nearest neighbor method, ii) the year is chosen randomly among the twelve corresponding years of the climatology.

What does the distributed version of TIGR contain?

The TIGR package contains the following elements:

• The TIGR database (atmospheric description): atm4aTIGR2000_v1.2_43lev.dsf .
• The corresponding readme: readme.TIGR2000_v1.2_43_lev.doc.
• A Fortran 77 source program to read the TIGR database: lecTIGR2000.f .
• atm4aTIGR2000_v1.2_43lev.ref: As an example, the output of the Fortran program lecTIGR2000.f when it is executed at the LMD. This output is in ASCII format, rearranged by columns (Pressure, Temperature, Water Vapor, Ozone) and by rows, each row describing a pressure level. This output is an example from only one of the TIGR atmospheres
• The format of the TIGR database is described in the readme file.

Data Download

Dr. Noelle.Scott
Laboratoire de Météorologie Dynamique
Analyse du Rayonnement Atmosphérique
Ecole Polytechnique
Route départementale 36
91128 PALAISEAU CEDEX
FRANCE