Noise & Population Impact Data Frequently Asked Questions

• Where did the data used in the study come from?
• The methodology employed in the study is documented here.  The input data consisted of: 1) the FAA’s published noise maps for BSR, SERFR, and DAVYJ; 2) Population data (2000, 2010) from the U.S. Census Bureau, and estimates (2016, 2021) from ESRI, Inc., a commercial GIS and data provider.
• How accurate is the population data used?

• The US Census Bureau’s decennial population counts have been carried out every 10 years since 1790.  The results are used to determine representation in the House of Representatives and the Electoral College, and to determine how more than $400 billion dollars of federal funding each year is spent on infrastructure and services.  In other words, the federal government is highly-motivated to produce an accurate Census count.  Indeed, the last (2010) Census was found to have been the most accurate in decades, with an error rate of 0.01 percent.
• ESRI’s population estimates (2016, 2021) were found to be the most accurate amongst 5 leading data vendors in a study conducted by researchers from UC Riverside and UC San Diego.
• Rather than simply assume uniform population density within a Census Block Group (a collection of multiple Census Blocks), we chose to use a more accurate method of determining population counts within the FAA’s noise contours using Block-level (the most specific) Census data in combination with a method known as Weighted Block Centroid apportionment to provide a more accurate (though somewhat computationally-intensive) population estimate for an arbitrarily-shaped geographic area.  Each of the routes studied involved an analysis of thousands of separate Census Blocks.

• How accurate/valid are the FAA Noise Estimates?

• These noise estimates were generated by the FAA for the Select Committee, for the express purpose of comparing the noise impact of the three routes.  In fact, the FAA itself generated comparisons of BSR vs. DAVYJ and SERFR vs. DAVYJ using these noise estimates.  Therefore, we believe our use of the FAA’s DNL values from their noise maps is consistent with their intended usage.
• Note the FAA’s published DNL values do not take population density into account. Thus the key contribution of TPHA is to combine noise modeling and population data, a combination that should play a role in the Select Committee’s evaluation.

• Were the FAA’s DNL noise estimates based upon a single atypical flight?

• By definition, DNL is a 24-hour A-weighted noise dose, so it doesn’t make sense to talk about DNL values in the context of a single flight.
• The FAA described their methodology when they presented their noise estimates to the Select Committee: their noise simulation used 60 days of actual flight tracks, randomly chosen from a 1 year interval, for SERFR and (2014) BSR.  Using the SERFR flight tracks as a starting point, the FAA simulated the application of Class B and OPD “fixes”, and a ground track shift, in order to arrive at their noise results for DAVYJ.  The FAA recently affirmed in the Select Committee meeting on October 3, 2016 that their noise simulation for DAVYJ was “as accurate as they could make it”.

• That said, the FAA has only provided a limited explanation of the data used and assumptions made in their noise estimates.  If the FAA should revise their noise predictions for the three routes, that would of course change the TPHA and population impact results.

TPHA Frequently Asked Questions

• Why did you create the TPHA metric?

• “Total Persons Aggravated (TPA)” is a metric proposed by members of the Select Committee to the FAA, to help evaluate the impact of various routes on the communities below them.  The FAA said they were unable to provide this metric--so we did it for them (as “TPHA”) by taking as inputs the FAA’s noise estimates for the BSR, SERFR, and DAVYJ routes, US Census Bureau block-level population estimates, and well-established federal research into the relationship between high DNL and community annoyance.

• What is TPHA good for?

• The purpose of TPHA is to enable a comparative evaluation of the impact of various routes (historical, current, and proposed).  In any comparison of routes with different noise impacts, it should be possible to generate a count of population affected within each noise contour.  But how can we compare two routes with populations across multiple noise contours?  TPHA provides a way to compare the impact of these two routes using an objective standard.

• Why “Highly Annoyed” instead of “Aggravated”?

• While “aggravation” lacks a precise technical definition, and its relationship to DNL is unknown, the relationship between DNL and “community annoyance” has been studied extensively by the federal government and there is an accepted standard.  The federal government has adopted the so-called “updated Schultz Curve” as the basis of federal government policy-making (by the FAA, DoD, NASA, HUD, EPA, and others) with regard to transportation noise for decades.  The U.S. Federal Interagency Committee on Noise (FICON) declared in its 1992 report that annoyance was its preferred ‘‘summary measure of the general adverse reaction of people to noise,’’ and that ‘‘the percentage of the area population characterized as ‘highly annoyed’ by long-term exposure to noise’’ was its preferred measure of annoyance.

• What is the Schultz Curve, and why is it relevant?

• T.J. Schultz's original 1978 study found a correlation between the percentages of groups of people highly annoyed and the level of average noise exposure measured in DNL.  This is commonly referred to as the Schultz curve. It represents the synthesis of a large number of social surveys that relates the long-term community response to various types of noise sources (161 data points in all), measured using the Day-Night Average Sound Level (DNL) metric. An updated study of the original Schultz study that was based on the analysis of over 450 data points collected through 1989 essentially reaffirmed this relationship.
• In 1992, the Federal Interagency Committee on Noise (FICON) found that the updated “Schultz Curve” remains the best available source of empirical dosage effect information to predict community response to transportation noise.  The shape of the updated Schultz curve is shown below:  

• Is TPHA an arbitrary, fabricated metric?

• The TPHA metric relies on well-established science and transportation policy which has been in place for decades (see FICON 1992, and FAA Aviation Noise Abatement Policy 2000 Section 4.2 “Assessing Aviation Noise: Assessment Methodologies”).  See the explanation of the federal-government’s “Schultz Curve” above for more details.
• The TPHA metric was originally created to address a need expressed by members of the Select Committee, and to demonstrate that objective comparisons of the noise and population impacts of different routes are possible.
• TPHA is not a new concept; we recently discovered a normalized version of this calculation is also known as Level-Weighted Population (LWP).  LWP is, as it turns out, a widely-used metric to compare and evaluate population impacts across different noise levels.  Once we learned the official name for this metric, we found many instances of its use for this purpose (Examples: Phoenix Sky Harbor, Portland, Reno/Tahoe, Phoenix-Mesa, Scottsdale, Santa Barbara, Hawthorne).

• Are the results yielded by the TPHA calculation inconsistent with the FAA noise maps?

• No, the TPHA calculations simply quantify what the FAA Noise Maps are showing, taking population density into account.
• For instance, the FAA’s DAVYJ noise map shows a significantly increased >45 dBA DNL noise contour that extends south of the MENLO waypoint (whereas the >45 dBA DNL noise contour for both SERFR and BSR are confined to the Bay, per the FAA’s noise maps).  Our analysis provides additional supplemental information which aids in the interpretation of the FAA’s noise maps, specifically that the population within this DAVYJ noise contour now impacts nearly 25,000 people across East Palo Alto, Menlo Park, and Palo Alto, whereas there were no people living within the >45 dBA DNL noise contours for BSR or SERFR.