A detailed noise prediction model was used to compare 11 highway noise barrier locations in Florida. Insertion losses, ground effects, shadow zones, and overall trends were determined or analyzed, or both. Each location was modeled using STAMINA2.0 (current Federal Highway Administration regulatory model), STAMINA2.1 (Florida's version of STAMINA2.0 with state-specific emission levels), the Traffic Noise Model (often referred to as TNM; this model will replace STAMINA2.0 in the year 2002), and the University of Central Florida Community Noise Model (CNM5.0). The modeled results were then statistically compared with the measured results. Statistical evaluation results were similar for all models for overall, absolute prediction compared with the measured value, with STAMINA 2.1 being slightly better. All models provided adequate results, but ranges of error were significant. When the propagation components were explored, by comparing reference levels with those behind the barrier, the TNM was significantly better. The results also provided further insight into the benefited regions behind the barrier, a more detailed understanding of how the models perform for this complex interaction with the ground and sound wave, and how background levels change the actual size and shape of the benefited region.
Abstract