Skip to content U.S. Department of Transportation/Federal Highway AdministrationU.S. Department of Transportation/Federal Highway Administration
Office of Planning, Environment, & Realty (HEP)

Traffic Noise Model: Frequently Asked Questions FAQs

Menu: Input


  1. TNM outputs "invalid" results listed in the Sound Levels results table when I model berms. What is causing this error message?

    The toe of the berm (end of slope) can bury receivers and it may not be obvious. This will cause "invalid" results for that receiver. Use the skew view to make sure the toe of the berm is not overlapping a roadway or receiver location.

  2. How do I assign adjustment factors in TNM?

    For instructions on how to assign adjustment factors in TNM 1.0, 1.0b, and 1.1, please refer to Pages 68-70 in the TNM User's Guide.

    For TNM 2.0, you may assign Adjustment Factors in three ways:

    1. Graphically, using the Input, Adjustment Factors menu choice.
    2. Graphically, using the Select button in the Receiver input dialog box.
    3. Using the Edit button in the Receiver input dialog box.

    The following paragraphs discuss problems with these functions in Version 2.0 and previous releases. These problems have been addressed in Version 2.1 and above.

    Users have found that using methods 2 and 3 can occasionally cause crashes in TNM Version 2.0. Use File, Save very often.

    Graphical assignment of Adjustment Factors with the Select button is tricky in TNM 2.0 because it is easy to delete previously assigned factors.

    Clicking on the Show button in the Receiver input dialog box for a given receiver will highlight all roadway segments in the Plan view to which you have assigned Adjustment Factors for that receiver before addressing new ones. However, in TNM 2.0, this button only works properly when assigning factors by the Input, Adjustment Factors menu choice.

    Also, if you do not graphically select all previous roadway segments already assigned for this receiver, their adjustment factors will be deleted when you assign the value to the new segments. When you do select them, their adjustment factors will NOT be changed to the new value; they will retain their old values. Note that you do not have to graphically select roadway segments for which Adjustment Factors were assigned for a different receiver, only for previous segments already assigned for this receiver.

    You must Apply your changes in the Receiver input dialog box, close the table and then re-open it to see the updated factors.

    For the methods of assigning adjustment factors other than the Input, Adjustment Factors menu choice, the Input Adjustment Factors table does not automatically update if it is open as you assign Adjustment Factors. You must Apply your changes in the Receiver input dialog box, close the table and then re-open it to see the updated factors.

    Instructions on how to assign adjustment factors in TNM 2.1 and above are given in the actual adjustment factor input dialog box.

  3. Why are my computed sound level results sometimes lower than my inputted Existing Levels?

    TNM uses your inputted existing levels solely as the baseline for judgment of relative noise impact - that is, impact due to a substantial increase in sound level.

  4. How do you add background noise to TNM results?

    TNM does not currently incorporate the existing background noise levels in its results. The "Existing Level", that can be input for each receiver, is only used to determine the "Increase over Existing" in the results tables. To account for background noise in TNM, it must be added in externally. First, the TNM user must determine/measure the appropriate background noise levels. In many cases, the background level can be measured in the same or similar neighborhood at a distance from the roadway where the highway traffic noise is not heard. Then, TNM's predicted levels can be logarithmically combined with the background noise level (in a simple process in a separate spreadsheet). See the equation below:

    Ladj. = 10 X log10(10(LTNM/10)+10(LB/10))

    LTNM = TNM-generated sound pressure levels [dB(A)], and
    LB = Background noise level [dB(A)].

  5. Why don't my sound levels decrease when I input Adjustment Factors?

    TNM uses adjustment factors to algebraically add, not subtract, from calculated sound levels. Therefore, if you want adjustment factors to reduce calculated sound levels -- say for attenuations that TNM does not calculate -- you must enter them as negative numbers.

    Please refer to the Menu: File, Import DXF/Stamina section of this document for a discussion on importing Stamina shielding factors as adjustment factors.

  6. Should I use adjustment factors for building rows or tree zones?

    In most cases, no. Adjustment factors should not be entered for building rows or tree zones. Adjustment factors can be used in TNM to account for specific pavement type (after your state receives FHWA approval), atmospheric conditions other than absorption, parallel barrier degradations, and other propagation effects not calculated by TNM. In addition, adjustment factors may be used to calibrate the model to measured data (if your state allows such calibration).

  7. Does calculating new active receivers invalidate previous sound level results?

    Calculating for newly "active" receivers invalidates previous active receivers' results. If you run TNM with active receivers and then add more active receivers, TNM invalidates results for previously calculated ones when you try to calculate for newly activated ones. The exception is if you calculate "All" receivers the Calculate, Current Run, All Receivers command after calculating only some receivers using the Calculate, Current Run, Active Receivers command. In this case, you will get a message box asking if you want to recalculate the already-calculated receivers or if you want to skip recalculating those receivers with results.

  8. How do I model a reference microphone above a barrier?

    TNM will not allow a microphone to be positioned exactly over a barrier. Offset your receiver position horizontally by 0.1 ft or 0.1 m.

  9. How do I activate/deactivate receivers?

    In the receiver input dialog, go to the Notes tab. In the active column, check/uncheck desired receivers to activate/deactivate them from calculations as desired.

  10. Is there any information regarding what to do to account for the loss of noise indoors?

    Chapter 8 of "Measurement of Highway Related Noise" (Report FHWA-PD-96-046, May , 1996) has measurement procedures for determining building noise reduction in the vicinity of an existing highway (if this is your scenario). If you are trying to forecast interior noise levels for a new highway, HUD guidelines has a method for building IL.


  1. How do I model roadway medians?

    (Applies to TNM version 2.5 only)
    The recommendation on how to model medians has changed from previous versions of TNM. For TNM Version 2.5, the following recommendations apply:

    1. the minimum recommended width is still 3.05 m (10 ft);
    2. if the median is a ground type other than the default, use a ground zone to model the median; and
    3. for a ground zone median, avoid overlapping or matching edges with the adjacent roadways.
  2. How do I model roadway medians?

    (Applies to TNM version 1.0, 1.0b, 1.1, 2.0, and 2.1)
    Users should model roadway medians when the median is greater or equal to 3.05 meters (10 ft.). If you are using a default ground type of field grass, create a ground zone of field grass between two roadways to model a grass median. If you are using any default ground type other than field grass, create a ground zone of lawn between two roadways to model a grass median.

    For all roadways, it is necessary for the median to overlap the roadway edges as defined by their widths because TNM will default to the pavement until it reaches the edges of the roadways. For medians less than 10 ft, use 1 roadway without traffic to model the median or use 2 wide roadways that overlap. Also, do not snap ground zone points to the roadway points when modeling the median. Often check that they are properly overlapping using the Skew Section view.

  3. Do I need to be careful to avoid my roadway widths from overlapping?

    (Applies to TNM version 1.0, 1.0b, 1.1, 2.0, and 2.1)
    No. Instead, try to actually guarantee overlap, so that TNM does not insert an unwanted gap in the roadway pavement, which would in effect introduce impedance discontinuities. The overlap can be as little as 0.03 m (0.1 ft). Avoid exactly matching the edges of the roadways.

  4. Can I model multiple lanes using a single, equally wide lane?

    Yes, but this is not recommended. For more accurate results, users should model each lane separately even if there is no traffic present on the roadway. Also, ensure the roadways overlap so that TNM does not insert an unwanted gap in the roadway pavement, which would in effect introduce impedance discontinuities.

  5. Do I need to be careful to avoid my roadway widths from overlapping?

    (Applies to TNM version 2.5 only)
    No. Instead, try to actually guarantee overlap, so that TNM does not insert an unwanted gap in the roadway pavement, which would in effect introduce impedance discontinuities. The overlap can be as little as 0.03 m (0.1 ft). Avoid exactly matching the edges of the roadways. Also, avoid overlapping or matching roadway edges with the adjacent ground zones (e.g. medians).

  6. What is "average" pavement?

    A pavement type consisting of REMEL data measured on DGAC and PCC pavements combined. Use "average" pavement in nearly all situations. See Appendix A of the TNM User's Guide for FHWA policy related to pavement type.

  7. How should roadway intersections be modeled?

    Two accepted methods exist for modeling intersections in TNM: (A) modeling roadways that stop short of and restart after an intersection, and (B) modeling a complex series of intersecting roadway segments. Both methods produce comparable sound level results at receivers that are not directly adjacent to the intersection being modeled. However, method (A) is much less complex and easier to model, whereas method (B) provides more accurate results at receivers directly adjacent to the intersection. It is up to the individual TNM user to determine which modeling method might be most beneficial to their modeling project.

    Method (A): If you are modeling an intersection and do not have receivers close to the intersection, then follow these steps:

    1. Start two roadways, one for each direction, (north/south) and stop short of the intersection.
    2. Start two roadways right after the intersection (also north/south).
    3. Repeat the same steps for roadways going east/west. Use flow control devices to model acceleration roadways that leave the intersection. Be sure to verify direction of traffic flow when modeling intersections.

    The intersection should look like this:

    When modeling an intersection with no nearby receivers, the roadway segments do not have to touch at the center of the intersection.

    Method (B): If you are modeling an intersection with receivers close by, then more detail is needed. Follow the same steps as the previous sample but instead, start modeling the accelerating roadway at the beginning of the roadway entering the intersection, making sure that each roadway contains segments that end at each applicable intersection point (see example below):

    When modeling an intersection with nearby receivers, the roadway segments should touch at the center of the intersection. Use of the snap tool eliminates error.

    It is very important that the coordinates of each roadway segment at a given intersection are identical, otherwise TNM will produce an "Inconsistent Intersection" error. It may not be sufficient to use the TNM Snap Tool when modeling roadway intersections, so make sure to enter in the intersecting roadway coordinates individually in the input dialog boxes.

    It is also important, that each accelerating section of roadway is modeled entirely as one TNM roadway (highlighted in red in the above illustration). The acceleration roadway cannot be split into separate roadways, because TNM cannot continue the vehicle acceleration from one roadway to another.

    Keep in mind that the noise effect due to acceleration varies by vehicle type. For example, the acceleration effect is greater for autos than for the other vehicle types, because autos are quiet for cruise throttle at low speeds. (Compare the various emissions graphs in the TNM Technical Manual.)

  8. Can TNM model vehicle deceleration?

    TNM does not have a function that allows you to model true deceleration. Some users believe that by putting a traffic control device on the other end of a roadway, therefore reversing the direction will decelerate the vehicle instead of accelerating it. This is not true. Modeling a roadway in the reverse direction will cause vehicles to accelerate away FROM the flow traffic device. This would reverse the direction of traffic flow, and would produce emission levels and frequency spectra acoustically different from deceleration. Roadways must always be digitized in the direction of traffic flow.

    Although TNM does not have a function that allows you to model true deceleration, sound users have tried decreasing sound levels in TNM by breaking a roadway into multiple segments with decreasing speeds. Please note that although this does lower sound levels, it will produce emission levels and frequency spectra acoustically different from real life deceleration. Also, keep in mind that a decrease in sound level for decelerating traffic on a roadway may not be discernable at the receiver(s) if the sound levels from nearby roadways are dominant.

  9. How should I model TNM roadways when considering grade adjustments?

    Each mainline should be modeled as a continuous TNM Roadway (not several, strung together end to end) especially when dealing with grade adjustments. When importing STAMINA files that have several roadways and grades in each segment, be sure to combine these roadways into one long TNM roadway, since TNM does not link speeds between roadways.

    You should not start a mainline roadway in the middle of a road with an upgrade. If you have to start a new TNM roadway some distance on an existing roadway with an upward grade, you may want to manually input a lower speed for heavy trucks. You may calculate this speed using Figure 46 on Page 57 of the TNM Technical Manual. Heavy trucks will already be slowed down from the point where the grade meets the new roadway. Be sure to model a separate truck lane, if one exists.

  10. How do I model roadway shoulders?

    Use roadways with no traffic to model shoulders. Using an empty roadway also allows you to drop the elevation of the shoulder since shoulders usually drop 6 inches or so. Apply good engineering judgment when considering modeling such a small change in elevation. When modeling shoulders with overlapping roadways, TNM draws the slope from the center of the roadway to the edge of the shoulder.

  11. How do I create a user-defined vehicle?

    In order to create a user-defined vehicle, you will need sound level emissions data for each vehicle that you are using in the model. To measure the necessary data:

    1. Obtain a copy of the official TNM measurement method from FHWA,
      "Measurement of Highway Related Noise" (Report FHWA-PD-96-046, May , 1996).
    2. In accordance with the official method, measure A-level emissions as a function of speed for your user-defined vehicle type. As required by the official method, measure under the following reference conditions:
      • Cruise throttle
      • Level grade
      • Dense-graded asphalt pavement (DGAC) or Portland cement concrete (PCC)
      These measurements will result in the following 4 parameters for your user-defined vehicle type:
      1. Similar TNM Type:
        The Similar TNM Type represents the vehicle type from the standard list of five TNM vehicle types (i.e., autos, medium trucks, heavy trucks, buses and motorcycles) which is most closely aligned with the user-defined vehicle. The assignment should be based on similarities in subsource heights, acceleration characteristics, and frequency spectrum.
      2. Minimum Level:
        The Minimum Level represents the emission level at idle and at very low speeds as defined by the engine/exhaust noise ("C" coefficient).
      3. Slope:
        The Slope ("A" coefficient) is determined from the analyses of the user's emission level measurements data.
      4. Reference Level:
        There is actually a small mix up in the TNM code, so instead of entering the "Reference Level" in this input-field in order to calculate the "B" coefficient, TNM needs the user to simply enter in the "B" coefficient. The Reference Level represents the emission level for the user-defined vehicle measured at 50 mph. If a preliminary analysis was not performed, the "B" coefficient is easily found from the standard emission level equation since all other variables in the equation are known. Be sure to substitute the emission level at 50 mph into the equation when determining the "B" coefficient.
  12. How do noise analysts determine noise levels when user defined vehicle types or other extenuating factors, such as tunnels or rest stops, are involved?

    The intended purpose of the Federal Highway Administration's Traffic Noise Model (FHWA TNM) is to analyze highway traffic noise. Use of the model to predict noise levels for other sources may not provide accurate results. For example, the FHWA TNM was not designed to be used to analyze racetrack noise; therefore, the accuracy of the model used in this capacity is unknown. To use the FHWA TNM to analyze racetrack noise the following four (4) tasks would need to be completed. It should be noted that additional tasks may be required based on the specific scenario being analyzed.

    1. Measure a statistically sound number of similar types of racecars at a different racetrack using REMEL measurement methodology. This REMEL data is essential since racecars are not included in the FHWA TNM REMEL database.
    2. Determine the parameters necessary to develop a FHWA TNM user-defined vehicle for racecars. This requires knowing which of the 5 spectral shapes matches best with the racecars, as well as to determine the values for the other parameters necessary for input for a user-defined vehicle.
    3. The speeds at racetracks may exceed those measured for the FHWA TNM REMEL database, and if so will therefore be extrapolated.
    4. Examine the best way to model the racetrack, keeping in mind:
      1. That the traffic input will most likely be in hourly volumes,
      2. It may be necessary to model accelerations, and
      3. Classify the ground type in the center of the track correctly since the sound on the far side of a track will propagate over it.

    TNM users may also refer to NCHRP Report 791 Supplemental Guidance on the Application of FHWA's Traffic Noise Model (TNM) for further modeling guidance at locations where vehicles are idling, such as overnight rest stops, or at the end of tunnels or where bridge-deck reflection affects noise levels along with several other scenarios.

  13. How long should my roadways be?

    A rule-of-thumb is that a roadway should be minimally 8 times the distance between the roadway and the most-distant receiver, with the receiver centered along the roadway.

  14. How do I input Ldn and Lden data in TNM?

    (Applies to TNM version 1.0, 1.0b, 1.1, and 2.0)
    In versions previous to 2.1, TNM automatically assumes a 50/50 split of ADT between day and night for Ldn and a 33/33/33 split of ADT between day, evening, and night for Lden. It is recommended that users not use Ldn and Lden but instead modify Equation 15 on Page 66 in the TNM Technical Manual to compute equivalent hourly volumes. The equivalent volumes would give a LAeq1h that is actually a DNL or CNEL. Then input this as LAeq1h traffic type. (Note: When using Equation 15, %day + %evening + %night = 100%; for input to TNM, %day = %evening = %night = 100%).

    Also keep in mind that although lane-by-lane speed and volume is not necessary, HOV and truck restricted lanes should be implemented.

    * Calculations for Ldn and Lden have been modified to incorporate more representative numbers of hours in TNM Version 2.1. Only if your total hourly traffic volumes are the same for each hour of the day, this modification will now produce more representative Ldn and Lden.

  15. How do I input Ldn and Lden data in TNM Version 2.1 and above?

    (Applies to TNM version 2.1 and 2.5)
    Calculations for Ldn and Lden have been modified in TNM Version 2.1 and above to incorporate accurate numbers of hours (Ldn: 15 hours day and 9 hours night; Lden: 12 hours day, 3 hours evening and 9 hours night).

    The calculations for Ldn and Lden in TNM Version 2.0 and all previous versions assumed the entered ADT value was split evenly between day and night for Ldn and day, evening, and night for Lden. For example, for Ldn the previous versions assumed the same total traffic passed by during the 15-hour day as passed by during the 9-hour night. This resulted in higher hourly volumes during the night than during the day, because the same 15 hours of daytime traffic was compressed into only 9 nighttime hours. Because the calculations were not accounting for the different number of hours included in day, evening, and night, erroneously high sound levels were being calculated for Ldn and Lden.

    TNM Version 2.1 and above accounts for actual numbers of hours by dividing the traffic equally among all the hours of the day-that is equal hourly traffic in each hour of the day. This still may not be realistic for your project. If not, we suggest the following:

    • Decide volumes and speed for a typical peak-traffic hour (a.m. combined with p.m.), along with its duration in hours.
    • Do the same for a typical daytime off-peak hour, for a typical evening hour (between 1900 and 2200), and for a typical nighttime hour (between 2200 and 0700).
    • Then use TNM to compute four separate LAeq1h values for these four typical hours.
    • Then average these four results by decibel arithmetic (see equation below), where each typical-hour result is first weighted by the number of hours it represents, and where (1) nighttime hour sound levels are increased by 10 dB, and (2) for Lden only, evening hour sound levels are increased by 5 dB. The increases account for the lower tolerance of people to noise during these time periods.
    • The result will be a relatively accurate value of Ldn (or of Lden).

    This value of Ldn or Lden will be accurate only if your projections of nighttime and evening truck traffic are reasonably accurate.

    Equations for calculating Ldn and Lden:

    Ldn = 10 X log10((t1/24) X 10(L1/10) + (t2/24) X 10(L2/10) + (3/24) X 10(L3/10) + (9/24) X 10([L4+10]/10))

    , where
    L1 = LAeq1h for peak daytime traffic,
    t1 = number of hours L1 represents,
    L2= LAeq1h for off-peak daytime traffic,
    t2 = number of hours L2 represents,
    L3= LAeq1h for evening traffic,
    L4= LAeq1h for night traffic, and
    t1 + t2 = 12 hours.

    Lden = 10 X log10((t1/24) X 10(L1/10) + (t2/24) X 10(L2/10) + (3/24) X 10([L3+5]/10) + (9/24) X 10([L4+10]/10))

  16. How does TNM compute acceleration on roadways with traffic-control devices?

    For TNM roadways with a traffic-control device, TNM computes accelerating speeds along the roadway's length as a function of vehicle type and roadway grade until the final speeds are attained or the end of the roadway is reached. For the next roadway, TNM begins anew with that roadway's input speeds. In other words, while TNM tracks speeds from one roadway segment to the next, it does not link speeds from one roadway to the next.

    If you know that vehicles will continue to accelerate past the endpoint of the TNM traffic-device roadway, you must extend this roadway if you wish to allow for continuing acceleration. For example, if heavy truck acceleration will continue past the physical merge point of an on-ramp with the mainline, you may wish to extend the on-ramp past this physical merge point, parallel to the mainline, so that heavy trucks will come close to reaching input speed before the end of the on-ramp roadway. Use Figure 45 on Page 57 of the TNM Technical Manual to compute this length.

    The need to extend this full length depends on the number of trucks on the ramp compared to the mainline and the proximity of receivers. If you have a lower number of heavy trucks on the on ramp compared to the main road, then you don't have to extend it all the way out until it reaches the main road speed. It is very often not necessary to extend the ramp roadway the full distance. In addition, vehicle emissions levels at high speeds (particularly above 40 mph) do not differ much between cruise-throttle and full-throttle vehicles. For this reason, your computations will generally be sufficiently accurate even if you stop traffic-controlled TNM roadways quite short of the full distance implied by Figure 45 in the Technical Manual.

  17. How do I calculate roadway contributions?

    In cases with multiple roadways or roadway segments, some users may want to determine roadway contribution by calculating sound levels for a specific roadway or segment. This can be done by zeroing out the traffic on all of the roadways except the one(s) being investigated, so only its contribution will be included in the receivers' sound levels. Use a zero-traffic roadway, instead of removing the roadway. Do not remove a roadway while checking for other roadway contributions, because roadways help define the ground plane.

  18. When should I use "on struct"?

    Be sure to use "on struct" (located in the General tab of the Roadways Input dialog box) for roadways that are on structure, such as an overpass. It is best to add "on struct" as the final step before calculation after you've made all changes to your run; some users have found that TNM sometimes scrambles the on-structure barriers database when they edit roadways and barriers that are already selected to be "on struct".

    In TNM Version 2.1 and above, you may have to click the "on-struct" box several times before it will respond. Try double-clicking it, then single-clicking it.

  19. How do I model roadways with super elevated roadways (cross-slopes)?

    On roadways that have super-elevation (perpendicular to the direction of traffic flow): If the roadway is relatively flat, you can use 1 wide roadway (limiting this to 3 lanes of traffic) or multiple, slightly overlapping roadways. If the roadway has a large amount of super-elevation, use separate, slightly overlapping roadways with changing elevations to approximate the cross-slope.

  20. Should I model slow and fast lanes separately?

    Users should model slow and fast lanes separately, because the shape of the vehicle's 1/3-octave band frequency spectra is a function of speed and thus propagation, barrier attenuation, ground attenuation, and tree zone attenuation will be sensitive to speed. If you have a multiple lane roadway all with similar lane speeds, users are still recommended to avoid combining more than 3 lanes into a single TNM roadway. Also remember to overlap TNM roadways.

  21. Should I model HOV and truck lanes separately?

    Pay close attention to HOV lanes, truck lanes, etc. HOV lanes and truck lanes have different speeds and traffic volume by vehicle type that will affect sound level results. Model these lanes as separate TNM roadways.

  22. When does TNM stop acceleration for traffic on a roadway?

    TNM will stop acceleration when it reaches its user input target speed (speed inputted in the traffic input tab), but will maintain full throttle throughout the entire segment. If the roadway is shorter than the length specified in Figures 43-45 in the TNM Technical Manual, then the traffic will not reach the target speed(s) by the end of the roadway. This is not a problem if the traffic has reached a high-enough speed (particularly above 40 mph) so that the cruise-throttle and full-throttle emissions are nearly the same.

  23. When digitizing roadways, how close should I match the curves in the Plan View?

    When digitizing roadways, try to match curves in the Plan View within plus or minus 2 m (6 ft). This will almost always ensure the sound level results are precise to within 1 decibel. Keep in mind that cases with elevation changes may be more affected from a small change in object location. With no nearby receivers, you may relax this precision. In addition, always try to maintain approximately the correct roadway length on tight curves, by "straddling" the actual roadway centerline with your straight-line-segment approximation within TNM. Concerning Z coordinates, Section 8.3.3 of the TNM User's Guide discusses the need for additional roadway points to adequately match the vertical roadway coordinates.

  24. Are roadways allowed to intersect with other TNM objects?

    In general, do not allow roadways to intersect any other type of TNM input. In addition, do not digitize roadways inside of tree zones. TNM will also detect and report all illegal intersections during Input, Input Check (see Section 8.15 in the TNM Users Guide). Roadways can intersect other roadways, but only at a common roadway point (same X, Y and Z coordinates). Sometimes it is necessary to physically type in these identical coordinates, rather than relying upon TNM's "snap."

  25. What is the maximum speed I can model heavy trucks in TNM?

    When modeling heavy trucks, do not use speeds in excess of 120 km/hr (75 mph) on roadways with grades greater than 1.5 percent. Doing so will cause TNM to output an error message and stop sound level calculations.

  26. I am interested in modeling vehicles idling. Would modeling their speed limit at 1 mph accomplish this and be an accurate representation?

    This could work in some cases, but you need to think about why /under what condition you are modeling. Remember that TNM is set to model a constant flow line source. Idling vehicles are more of a stationary point/area source if they are stationary for a long period. If you want to include the effects of stopped vehicles at a signal or some similar source, they would produce an insignificant portion of the total noise, which would be dominated by acceleration away from the light. It is recommended that users work this type of problem out with paper and pen instead.

  27. How can the newer pavement types be credited in the TNM?

    TNM currently allows the user to select from 4 pavement types: DGAC, OGAC, PCC, and an average pavement type (combined DGAC and PCC). These data are based on measurements performed in seven States around the country. Incorporating new advanced pavement is possible, however future implementation will depend on FHWA policy, and a statistically sufficient amount of in-situ data measured using a consistent methodology. Current FHWA policy states, "Unless definite knowledge is available on the pavement type and condition and its noise generating characteristics, no adjustments should be made for pavement type in the prediction of highway traffic noise levels." The use of a pavement type other than "Average" must be substantiated and approved by the FHWA, in advance of its use. "Future case" runs should also use average pavement unless otherwise justified.

  28. If I have a 45 mph secondary road with a stop sign, it appears to me that taking flow control into account is not really necessary. Are people finding meaningful differences by including flow control?

    We would definitely recommend modeling the flow control device. Although you will not see a large difference between the A-weighted Sound Levels for Autos at 45 mph Cruise and the A-weighted Sound Levels for Autos accelerating in speeds of 0 to 45 mph Full Throttle (as seen in Figures 7 & 8 of the TNM Tech Manual), the vehicle emission spectra vary greatly over that range (as seen in Figures 17 - 20), and should be taken into account.

  29. We are provided ADT traffic volume for a given roadway. Page 63, Item 4 at the bottom of the page (TNM User's Guide) asks if you wish to apply traffic to the entire roadway. Does the software divide the ADT between northbound and southbound or does the user have to input half? e.g. if the combined traffic in both directions is 22,000, does the user enter 11,000 or 22,000?

    If you are given an overall ADT, which encompasses lanes in both directions, you will need to split that up amongst the lanes you modeled in TNM. At a minimum, that will be 50/50 for northbound/southbound, but it could be more complex if you modeled more lanes (i.e.: 3 separate northbound lanes, a separate HOV lane, etc.). In your example, if you had only two roadways (one for each direction), you would split the 22,000 and enter 11,000 as your ADT for your northbound roadway and 11,000 for your southbound roadway. Inputting ADT assumes you are calculating Ldn or Lden. If you wish to calculate LAeq1h, you must switch the Traffic Entry Type to LAeq1h and input hourly traffic volume.

  30. Why do my Stamina roadways only show 3 vehicle types?

    Because Stamina only allowed 3 vehicle types, TNM imported the roadway traffic similarly. To populate Stamina roadways with TNM's five vehicle types, create a new road to replace the Stamina road by using the Snap tool and the roadway toolbar icon to trace over the existing Stamina road. Copy the traffic from the Stamina road and enter the additional traffic for the buses and/or motorcycles. Then delete the original Stamina road.

  31. What is the point of taking noise field measurements - just to see if your data is close?

    Field noise measurements can be done for comparison. If you are modeling traffic noise, you're likely writing up a report or section of a report with your findings. If you have a field monitoring point, you can enter that in as your ambient noise level, and then when you run your analysis, TNM will automatically identify which receptors exceed the ambient level.

    99 times out of a hundred you will be designing noise barriers for some future condition. In the real world of highway design, that is typically twenty years from the anticipated first year of operation. That future condition will consist of traffic volumes, traffic mix, and operating speed based on assumed parameters. The future roadway/receiver geometry for improvement projects will certainly change, sometimes dramatically between the roadway and the modeled receivers. Other times there will be only minor modifications on the roadway cross section between the roadway and receivers when the improvements are made primarily toward the median area.

    You either have a Type 1 (substantial alteration to existing highway or highway on new alignment) or a Type 2 project (retrofit barrier on existing highway). Existing measurements are essential for both scenarios given adequate funding. For Type 1 projects, one of the measures of an impact is a substantial increase over your existing conditions (see your States guidance on defining "substantial"). Taking measurements in an area where a new highway is proposed (few noise sources) is essential as this area has the highest potential for a "substantial increase over existing noise", and truly cant be modeled since there are no traffic sources. For other Type 1 projects where existing noise sources are present, modeling would work if you're pressed and you have a simple area. It is recommended to monitor peak noise conditions, calibrate to existing conditions (using traffic you counted during monitoring), and then re-model for existing conditions using more statistically documented traffic for the base year (monitoring data is snap shot in time). Then modify this file for future conditions.

  32. How do I use my traffic data to "calibrate" my TNM run to existing conditions?

    The following excerpt is taken from the report, Validation of FHWA Traffic Noise Model (TNM) Phase I (Report number FHWA-EP-02-031, August, 2002) and provides guidelines on calibrating your model:

    The approach for calibrating TNM to measured data depends on the state noise policy for the state of the highway project. Although each noise analyst should refer to their state noise policy, the following guidance gives examples of how to best use TNM for highway traffic sound level predictions.

    In California, the Department of Transportation (Caltrans) gives specific guidance on calibrating a noise prediction model [Hendriks 1998]. Their calibration process is defined as follows: an adjustment is made to the calculated future noise levels by algebraically adding a calibration constant derived from the difference between measured and calculated noise levels at representative sites. The types of sites to which the calibration process is applied include highway widening projects, design of retrofit noise barriers, or other improvements that do not significantly change highway alignment or profile. Sound levels are measured at representative locations at a site during peak noise hour, in accordance with FHWA's measurement procedures [Lee 1996], and the site is modeled using exact site geometry and traffic input. The difference between measured and predicted sound levels is then calculated and applied to future sound levels, unless it is 1 dB or less; if the difference is 5 dB or greater, the measurements should be investigated. Please refer to the Caltrans document [Hendriks 1998] for further details.

    In Florida, the Department of Transportation (FDOT) also gives specific guidance on calibrating a noise prediction model [Lindeman 2001]. FDOT takes a different approach than Caltrans. Field measurements are conducted along all existing or proposed roadway segments or links that may be affected by the resulting highway traffic noise. Sound levels are measured at a representative site during peak noise hour, in accordance with FHWA's measurement procedures [Lee 1996], and the site is modeled using exact site geometry and traffic input. A comparison is made between the predicted and measured sound levels; if the levels are within ± 3 dB of one another, this is considered an indication that the model is within an accepted level of accuracy. If the difference is greater than ± 3 dB, further investigation into the problem is required; this may require adjusting the model (improving modeling techniques) and/or repeated field measurements for verification, investigating until an acceptable difference is reached. Please refer to the FDOT document [Lindeman 2001] for further details.

    Whatever the calibration process, it is important to apply good engineering judgment to the modeling and field measurements. Among other items, the placement of the microphone(s), site geometry, surrounding objects, extraneous noise, and highway traffic noise fluctuations must all be considered. Referring specifically to the sound level measurements, it is important to capture multiple samples of an appropriate length when measuring highway traffic sound levels. For highway noise measurements, guidance on the sample period to use and the number of samples to obtain can be found in the noise barrier standard [ANSI 1998] and FHWA's highway noise measurements report [Lee 1996]. Briefly summarizing, it is recommended to capture at least three acoustically clean samples (with six being preferred), where the sample length depends on the traffic flow; very steady traffic flow requires 5-minute acoustical averages and less steady traffic flow (but not sparse traffic) requires 15-minute acoustical averages./

    Hendriks 1998Hendriks, Rudolf W., Technical Noise Supplement: A Technical Supplement to the Traffic Noise Analysis Protocol (California Department of Transportation, Division of Environmental Analysis, Sacramento, CA, 1998) - Website URL for document:

    Lindeman 2001Lindeman, Winfield M., Project Development and Environmental Manual (PD & E), Part 2, Chapter 17 (Florida Department of Transportation, Tallahassee, FL, 2001)

  33. When should I use average pavement?

    FHWA policy states, "Unless definite knowledge is available on the pavement type and condition and its noise generating characteristics, no adjustments should be made for pavement type in the prediction of highway traffic noise levels." The use of a pavement type other than "Average" must be substantiated and approved by the FHWA, in advance of its use. "Future case" runs should also use average pavement unless otherwise justified.

    In measured vs. predicted studies, users should pay close attention to pavement type. Many users compare predicted levels and measured levels, find a difference and apply it directly to the results. They should be cautious when doing this with future cases. Keep in mind terrain line changes that may occur.

    FHWA is currently approving use of open-graded asphalt concrete (OGAC) in some states. To gain approval, the states must demonstrate with measurements the noise reduction obtained by the pavements within their state.

  34. Can TNM calculate two traffic types simultaneously (i.e., LAeqh1h and Ldn)?

    Use only one Traffic Entry Type per TNM run. TNM cannot combine different types in the same run, including use of Hourly volumes for some roadways and hourly percentages for others. Please refer to the Entering Traffic Volume section on Pages 61-63 of the TNM Users Guide for instructions on how to enter traffic data.

  35. When inputting traffic speeds, should I use the posted speed limits?

    No. When modeling roadway traffic speeds, use actual vehicle speeds. Do not use posted speed limits. Obtain a traffic engineer to obtain actual traffic counts, speeds, volume, and input into TNM. Although lane-by-lane speed and volume is not necessary, HOV and truck restricted lanes should be implemented.

  36. What happens when I switch the direction of a TNM roadway?

    TNM uses the starting point of a roadway segment to define the roadway. When you switch the direction of a roadway, the information for a roadway segment will now shift to a neighboring roadway segment. This is because the roadway now takes the original starting point going one direction and uses it as a starting point for the other direction. This can present a problem, if the input data of your roadway is significantly different from segment to segment, because now the information might conceivably be on the incorrect segment.

    Roadway before directions are reversed:

    Roadway with three segments, with arrows (representing direction of traffic flow) pointing from left to right. The starting point of the roadway has no hash-mark, indicating the roadway start point has no flow control device (e.g., traffic light).

    Roadway after direction is reversed. Notice the highlighted segment has shifted because the starting pointing of that segment is now going the opposite direction:

    Roadway with three segments, with arrows (representing direction of traffic flow) pointing from right to left. The starting point of the roadway has no hash-mark, indicating the roadway start point has no flow control device (e.g., traffic light).

    A good practice is to digitize or graphically input each roadway in the direction of flow. When you import a DXF file that has roadways in opposite directions, change the direction of the roadway before applying traffic data.

    Please note, that when you reverse the direction of a roadway with a flow control device, the flow control device gets moved to the other end of the roadway-that is, to the roadway's new "start point." In most cases, this is not desired. It is important to digitize a roadway in the direction of traffic, and when you are satisfied with the roadway's coordinates, then apply the traffic and flow control device.

    Roadway with a flow control device on one end:

    Roadway with three segments, with arrows (representing direction of traffic flow) pointing from left to right. The starting point of the roadway has a hash-mark, indicating the roadway start point has a flow control device (e.g., traffic light).

    Roadway after direction is reversed. The flow control device has moved to the other end of the roadway as a result.

    Roadway with three segments, with arrows (representing direction of traffic flow) pointing from right to left. The starting point of the roadway has a hash-mark, indicating the roadway start point has a flow control device (e.g., traffic light).
Updated: 10/19/2015
HEP Home Planning Environment Real Estate
Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000