An improved protocol for the performance evaluation of regulatory photochemical air quality modeling

摘要

Ozone is a secondary air pollutant made by chain reactions of two major precursors, nitrogen oxides (NO_X) and volatile organic compounds (VOCs), under ozone-conducive meteorological conditions. Eulerian photochemical air quality models (PAQMs) are considered the most suitable tool for simulating the effects of multiple precursor emission sources on ozone concentrations under various meteorological conditions. Thus, the 1990 Clean Air Act Amendments (CAAA 1990) highly recommended the use of PAQMs as the primary investigation tool for regulatory ozone problems when a state develops a State Implementation Plan (SIP) and demonstrates that the proposed SIP is effective for controlling future ozone problems. Model performance evaluation (MPE) is the process of gauging the reliability of PAQMs for SIP development and evaluation tools. MPE of PAQMs, however, is a difficult task because of the nonlinearity of the ozone formation mechanism and this makes it hard to judge whether good model performance is for the right reasons. In SIP modeling, it is important to ensure that a model gets its answers for the right reasons because flawed modeling can result in directionally incorrect control recommendations such as irrelevant NO_X controls when VOCs controls are actually necessary.