As part of the development of the FHWA TNM, FHWA recognized the need for a comprehensive and systematic model validation study to learn more about the sensitivities of the FHWA TNM, evaluate its performance, and identify model issues requiring improvement. This study, being conducted by the Volpe Center, began in July 1999 and is ongoing. Primarily funded by the FHWA, it has also received substantial funding support and manpower from the California Department of Transportation. The large scope of the study required its division into multiple phases. The first phase of the validation study is complete, and the report is available from NTIS report numbers FHWA-EP-02-031.
Phase 1 of the FHWA TNM Validation Study included measurement, modeling with the FHWA TNM Version 2.0, and analysis for 17 sites across the country. The site types were 8 open area sites and 9 noise barrier sites. At each site, acoustical and meteorological data were collected for approximately 6 hours at multiple locations simultaneously, along with traffic and site survey data. At each measurement location, data were collected by manned systems, thus ensuring the highest quality data possible. The large amount of data collected (both overall A-level and 1/3-octave-band) allowed for traffic variation and wind and temperature variation at each site, providing varying conditions for the best possible analysis. Results of the comparison between measured sound levels and FHWA TNM-predicted sound levels are contained in the Validation of FHWA's TNM® (TNM) Phase I Report (HTML) or (PDF - 6.5Mb), August, 2002 The results are presented by site type (open area or barrier), ground type, distance from the roadway, height above the ground, wind speed, and percentage of heavy trucks. The report also contains recommendations on FHWA TNM use, including the importance of "site calibration" when possible. This is a practice performed by many noise consultants and is standard in some states, including California and Florida. The Phase 1 validation study's primary conclusion is that for the sites tested, FHWA TNM propagation algorithms are performing extremely well. However, the study identified two areas of concern: (1) FHWA TNM is over-predicting, when the model's output is not calibrated for site bias based on measurements made close to the roadway; and (2) FHWA TNM is over-predicting for open area, acoustically hard ground sites at far distances (distances generally larger than 500 ft).
With the release of the FHWA TNM Version 2.5, additional investigations were performed in order to update the FHWA TNM Validation Study Phase 1 results. The CD-ROM releasing the FHWA TNM Version 2.5 included the Preview of Validation of FHWA's TNM® (TNM): Phase I (TNM Version 2.5 Addendum" April 2004 (HTML) or (PDF 186Kb). The "TNM Version 2.5 Addendum to Validation of FHWA's TNM® (TNM) Phase 1 report, July 2004" was finalized and released in July 2004 (HTML) or (PDF - 944Kb). Results of the FHWA TNM 2.5 Validation Phase I Addendum show that the FHWA TNM Version 2.5 is performing very well and that the general over-prediction is no longer an issue.
Phase 2 of the FHWA TNM Validation Study is underway. Results will be made public as they become available.
Ground and Pavement Effects using FHWA's Traffic Noise Model 2.5 (HTML) or (PDF - 2.1Mb) - This study investigates the effects of using different ground types based on an improved understanding of the characteristics of the ground types defined in FHWA’s Traffic Noise Model (TNM) version 2.5. This study also investigates the effects of using specific pavement types, as opposed to Average pavement, when the pavement type is known. The results of this study indicate that improvements in the correlation between predicted and measured results can be achieved by selecting the best ground type and pavement type for the model.