The most attractive candidate was UAM-V, a state-of-the-art urban to regional scale model capable of using nested grids with different resolutions (Systems Applications International, SAI). UAM-V also came with the promise of technical support from its developers, SAI, the developers and supporters of the venerable UAM-IV. Thus, UAM-V became the air quality model of choice and the centerpiece of the air quality modeling system.
The controversy associated with this selection is related to the fact that UAM-V was and remains a proprietary model. In other words, SAI owns the computer code that constitutes the model and has restricted access to the code to recoup their investment in model development. At the beginning of OTAG, SAI was able to charge a fee for the use of UAM-V and limit users access to the executable code. When a consortium of industry groups hired another consulting firm to perform a parallel modeling effort, SAI denied the consulting firm a license and access to UAM-V, fearing competition from the other firm. Now after substantial public pressure, the source code is publicly available, but no one is allowed to alter the source code in anyway without SAI's permission.
While these practices may not seem unusual to the average computer software user, this situation was and is very unusual within the practice of air quality modeling. All other air quality models that are used in the development of public policies, including UAM-V's predecessor UAM-IV, are in the public domain. Their source codes are accessible to the public so that they can be scrutinized and examined. Restricting access to UAM-V in general creates a situation in which interested parties cannot confirm the conclusions of other parties by replicating their model runs. Limiting the access to the executable code makes the model a black box: you know what goes in and what comes out, but you don't know what happens inside. Although descriptions of the model's algorithms exist in some documentation, it is only possible to confirm the implementation of these complex algorithms by altering the code to output intermediate answers as the model executes. Because of the scientific and political implications of such restricting access to the code, the decision to use UAM-V has been widely criticized and has led to several calls for a general ban on the use of proprietary models.