The following summarizes changes to the TIEGCM model from v1.8 (released 1/10/2006),
to v1.9 (released 6/20/2008):

Software:

• New CF-compliant netcdf history files have separate vertical coordinate arrays for midpoints (lev), and interfaces (ilev).
• Module addfld replaces addfsech and related routines as a single-interface for saving fields on secondary histories.
• Add gravity-varying geopotential height field (with latitude gradient) to secondary histories (ZG). Also add geopotential on the magnetic grid (ZMAG).
• Default resolution is 5 deg lat x lon, dlev=0.5. The code builds and executes at 2.5 deg and dlev=0.25, but has not been extensively tuned or tested at the double resolution (may require step < 120)
• Add section in Makefile for use of ifort compiler by DASG machines and other intel systems. Also some simplification of scripts/Makefile template.
• Simplification of the electric field calculation (remove redundancy)
• Remove annual tide option (was bndry_annual in bndry.F)
• Remove dynamo=0 option (remove source nodynamo.F)
• Removed read of external file magvol (magfield.F). Now calling apex code unconditionally.
• Bug fix in sub mp_gather2root_sechist (mpi.F)
• Replace hardwired joule heating multiplication factor of 1.5 with named parameter joulefac in duv.F
  (Value of 1.5 was *not* changed)
• Improved filtering mechanism, similiar to timegcm, also some changes to allow proper filtering for the 2.5 deg model.
• Add $lib_netcdf and $inc_netcdf shell vars in job scripts (they are passed to tgcm_config), to facilitate installation and compilation at external non-NCAR sites.
• Flexibility for the user in either providing constants or time dependent values for certain solar-related input parameters via namelist read, OR reading data from external data files (NGDC GPI or NASA/OMNI IMF). These input parameters include ctpoten,hpower,bximf, byimf,bzimf,swden,swvel,f107d, and f107a.
• Add license agreement file tiegcmlicense.txt to the source code directory, and add reference to this file at the top of all source files.

Science:

• Removed minimum electron density restriction in elden.F and increased nighttime flux in qinite.F
• Effective electron-neutral collision frequency rnu_ne was increased by a factor of 4 in lamdas.F
• Updated apex_subs.F to include apex coefficients for 1995, 2000, 2005.
• Corrected Joule heating for ions, includng vertical velocity difference (qjoule.F).
• Significant changes to the auroral parameterization, including empirical relationships between Kp data (GPI run) and hpower and ctpoten. Also positioning of the oval wrt Weimer high latitude convection. Several parameters re the precipitation oval in aurora.F were also changed.
• By default, GSWM migrating diurnal and semi-diurnal tidal perturbations are used. GSWM non-migrating tides are available via namelist read, but are turned off by default.

Known issues:

• Implementation of the Weimer high-latitude potential/convection model is not functioning properly because the precipitation oval does not always correspond to the Weimer pattern. Use of the Wiemer model is not recommended until this major code issue is resolved. Heelis convection remains the default.
• Daytime E-region electron densities are still about 30% too low when the model is run using EUVAC or SEE data solar inputs.
• Auroral generation of nitric oxide during higher levels of activity are probably on the low side.
• The O-O+ collision frequency is a matter of continuing debate. The model default Burnside Factor (COLFAC) remains 1.5. There is some laboratory and empirical evidence that 1.3 may be a better value. This can be set via the namelist input parameter COLFAC.
• Joule heating is increased by a factor of 1.5 by default, to account for sub-grid-scale and related effects. This factor (JOULEFAC, in dt.F) is also a matter of debate.
• Seasonal effects on density and composition (annual and semi-annual variation) are only partially accounted for in the current model configuration, since the lower boundary eddy diffusivity coefficient is constant.
• GSWM non-migrating tides are not yet validated and are off by default. They may be turned on via namelist input, but this is not recommended.
• There are some questions about the neutral dynamics near the lower boundary, so winds at the lowest few levels should be treated with skepticism.
• During significant storms, plasma may be driven to altitudes above the model upper boundary and lost, leading to unrealistic high-altitude ionospheric densities.
• O(1D) excitation by thermal electrons is turned off. Improved photoelectron heating and new cooling rates will be addressed in future versions.
• High-resolution grid (2.5 x 2.5 x H/4) is not verified and should not be used.

New features and options:

• Introduction of the 2005 Dan Weimer high-latitude electric potential model option (see module wei05sc.F)
  

  Please note:

    The Weimer model of high-latitude potential is the intellectual property of 
    Daniel Weimer and may not be extracted, distributed, or used for any purpose 
    other than as implemented in the TIE-GCM.  For further information concerning 
    this model, please contact Dan Weimer (dweimer@vt.edu).
    

• Namelist read parameter POTENTIAL_MODEL can now be one of the following:

      POTENTIAL_MODEL = 'HEELIS'   ; default potential model
      POTENTIAL_MODEL = 'WEIMER01' ; 2001 Weimer model
      POTENTIAL_MODEL = 'WEIMER05' ; 2005 Weimer model (new in tiegcm1.9)
      POTENTIAL_MODEL = 'NONE'     ; no potential model
    

• Import of NASA/OMNI IMF data to optionally drive the Weimer potential model (namelist read parameter IMF_NCFILE). See tiegcm IMF data import
  

  Please note:

  The option of using the Weimer 2005 potential model driven with IMF data is provisionally released in tiegcm1.9. We have made no effort to fill data gaps, and additional work is needed to insure proper positioning of the auroral oval with respect to the Weimer convection, especially during times of intense solar activity. The default potential model is still Heelis.

Documentation:

• Updated Users Guide in [html], or [pdf].
• A draft of a model description document is currently under construction, but is not released with this version.
• An improved website for TGCM models is being designed concurrently with this release of TIEGCM. This will soon be available at the HAO TGCM website