# Comparison of 2005 Model-Predicted Concentrations to Monitored Data

A model-to-monitor ratio is simply the modeled concentration divided by the annual average concentration. A model-to-monitor ratio close to 1 for a particular HAP at a monitoring site indicates a high level of confidence in the modeling results for that HAP and monitoring site. Similarly, an average model-to-monitor ratio of several monitoring sites closer to 1 can indicate a high level of confidence in the modeling results.

In addition to the census-tract level ambient concentrations predicted by the NATA 2005, EPA used the model to develop specific receptor-level HAP concentrations for over 1,000 locations which coincided with locations of air toxics monitoring sites. These concentrations were the basis for the 2005 model-to-monitor comparison.

Over 5,400 model-to-monitor ratios were calculated for 69 HAPs. Less than one-tenth of all model-to-monitor ratios were within 10% (i.e., ratios between 0.9 and 1.1), less than one-fifth of all ratios were within 20% (ratios between 0.8 and 1.2), and approximately one-fourth of all ratios were within 30% (ratios between 0.7 and 1.3).

Toluene, benzene, methyl chloride, 1,3-butadiene, formaldehyde, bromomethane, and 2,2,4-trimethylpentane had average model-to-monitor ratios between 0.9 to 1.1. Expanding the range from 0.8 to 1.2 adds xylenes, carbon tetrachloride, methyl-tert-butyl ether, cis-1,3-dichloropropylene, trans-1,3-dichloropropylene, and naphthalene. Gaseous HAPs tended to have average and median ratios closer to 1 than the metal HAPs. Metal PM10 average model-to-monitor ratios ranged from 0.120 (antimony PM10) to 15.305 (beryllium PM10). The chromium PM10 average model-to-monitor ratio is the closest (0.721) of all metals to 1.

For complete details about the model to monitor analysis for the 2005 National-scale Air Toxics Assessment, see file: 2005 NATA Model–to–Monitor Comparison(PDF format, 8.03MB)