Using the FAA’s published DNL noise contours for BSR (circa 2014), SERFR, and DAVYJ arrivals, we generated georeferenced polygons outlining each of the primary DNL noise contours (35-40 dBA, 40-45 dBA, and >45 dBA) in the immediate vicinity of each of the flight paths. While this approach ignores the extent of the noise impact due to delay vectoring, the FAA’s data does not provide reliable guidance as to the noise impact away from the flight path due to limitations in their noise simulation.
In addition, note that the FAA’s noise modeling only takes into account the noise generated by flights on these specific routes; many of the affected regions are also affected by other airplane traffic (e.g. BDEGA and Oceanic arrivals into SFO, SFO southbound departures, SJC traffic, etc), and so this analysis will likely understate the true impact of all airplane noise in our region. Finally, be aware that the impact of low-frequency noise is not adequately reflected in A-weighted decibel units, and DNL itself does not account for the increased impact of repetitive aircraft noise. That said, A-weighted dB and DNL measures are widely used in impact analysis and policy-making, and the availability and support for such metrics cannot be ignored.
Using commercial Geographic Information System (GIS) software, we were able to calculate the intersection between noise contour polygons and underlying Census Block-level population data to count the number of affected people living in each of the noise contours.
Bar charts tabulating population counts within the studied noise contours vs. distance from MENLO were calculated by slicing the studied noise contours into ~45 equal width segments of 0.77 nautical miles, slicing roughly perpendicularly to the flight path, and performing a population count within each of the resultant polygons. The graph renders the population counts as a stacked bar chart for each noise contour, and orders them by their distance from MENLO. We have included annotations for the waypoint HEMAN (because the noise contours extend north of MENLO) and relevant waypoints along each of the routes for ease of reference and comparison with other data visualizations.
In response to a request from the Select Committee on South Bay Arrivals, the FAA modeled the noise impact of the BSR, SERFR, and DAVYJ arrival routes on underlying communities. Consistent with FAA policy, the noise modeling was done based upon 60 randomly-chosen days in a 1 year period. For SERFR and BSR, actual flight tracks were used (for BSR, the 1 year period was calendar year 2014). For DAVYJ, the FAA modified SERFR flight tracks to reflect procedural changes anticipated for the notional route. These noise simulation results were presented by the FAA to the Select Committee at their Working Meetings on August 18, 2016 (SERFR, DAVYJ) and September 1, 2016 (BSR).
The FAA presented DNL noise contour maps for the three routes, as well as contours showing the change in DNL levels for DAVYJ vs. SERFR, and DAVYJ vs. BSR. The noise contour maps used ranges of values (e.g. 35-40 dBA DNL, 40-45 dBA DNL, etc.) and the size of these ranges was chosen by the FAA. Because the FAA’s noise modeling could not adequately simulate the extent of delay vectoring from these routes, we limited our studied noise contours to just those portions in the immediate vicinity of the flight path. As a result, the TPHA figures calculate here will underestimate the true impact of these routes, particularly on neighboring communities heavily impacted by delay vectoring.
To generate our population impacts, it was necessary to create georeferenced versions of the FAA noise maps to accurately capture the outlines of the relevant noise contours, and to determine their precise location. Fortunately, the FAA noise contour maps include the location of several waypoints, whose locations are precisely defined using GPS coordinates, which made georeferencing straightforward. Once polygons outlining each of the noise contours were generated, our GIS software calculated the area of intersection between these noise contours and underlying Census Block Groups. The population count for each area of intersections was generated by the GIS system’s Weighted Block Centroid apportionment method which, though computationally expensive, uses underlying Census Block data to more accurately estimate population data in smaller geographic areas.
The U.S. Census Bureau conducts its count every 10 years. In intervening years, the Census Bureau publishes estimates of population growth at the Census Block Group level. Our GIS provider, ESRI, provides an automated system for estimating population of within smaller, user-defined geographic areas, computed using Centroid Block Weighting of Block-level Census data. The ESRI 2016 estimate used in calculating TPHA is estimated population for July 1, 2016, and uses Census data supplemented by additional data from Experian, United States Postal Service, Internal Revenue Service, building permits and housing starts, and other data sources. In addition to total population count, in some renderings we have also included a count of the population aged 18+. Initially, we had intended to include a count of registered voters within the studied noise contours, but this data was not easily available. While the Age 18+ dataset may not be a direct substitute for a count of eligible voters (citizenship is not considered, for instance), we believe this count may still have some utility for the Select Committee and other elected officials.
To determine the ground elevation along the three routes, we used the Google Maps Elevation API, with a sampling rate of every 152 (SERFR) to 170 (BSR, DAVYJ) feet. The API returns elevations < 0 for underwater areas, so we set the floor to 0 (sea level).
For the altitudes of the flight routes, we used the FAA’s published average altitudes for BSR and SERFR, and the mid-point of the FAA’s published min and max altitudes for the notional DAVYJ route. Where average altitudes were not provided by the FAA (e.g. in the vicinity of ANJEE and WWAVS waypoints in the Monterey Bay), we extrapolated using the FAA’s preferred 2.85º glide slope from the nearest FAA-provided data point.
These altitudes assume the application of any required Class B fix to these routes and, thus, do not attempt to show any current level-offs being flown. The AGL distance along the route is calculated as the difference between the ground elevation and the average or midpoint altitudes of the studied flight route.
 Geographical Information Systems simplify the process of visualizing, analyzing, and interpreting geographical data by providing built-in support for common operations and techniques.
 A Census Block is the smallest geographic unit used by the United States Census Bureau. Census blocks are generally small in area. In a city, a census block looks like a city block bounded on all sides by streets. Census blocks in suburban and rural areas may be large, irregular, and bounded by a variety of features, such as roads, streams, and transmission lines. Each of the routes studied involved an analysis of thousands of Census Blocks.
 The FAA presented two noise simulation results for a notional DAVYJ route, one using lower altitudes (similar to the current SERFR route) and one using higher altitudes (similar to the historical BSR route). At this meeting, Steve May of the FAA indicated that he expected DAVYJ would use a descent profile similar to SERFR, and proceeded to compare the noise impacts of SERFR and DAVYJ using the lower altitude version of DAVYJ. As a result, we performed our analysis using the FAA-provided DAVYJ noise contours which reflect the lower (similar to SERFR) altitudes.
 We limited our analysis to three DNL ranges, 35-40, 40-45, and >45 dBA, both because these represent the areas of highest impact, and because the FAA’s forced modeling of delay vectoring using 4 discrete routes makes it impossible to accurately separate the noise contours < 35 dBA caused by the main flight path vs. those of the vectored routes.
 A Census Block Group is a geographical unit used by the United States Census Bureau which is a collection of Census Blocks. It is the lowest level at which the Census Bureau publishes population estimates in-between decennial counts.
 ESRI is an international supplier of geographic information system (GIS) software and applications. In 2014, Esri had a 43% share of the GIS software market worldwide, more than any other vendor.