Data Format and Download Instructions
Below are tract-level noise exposure map data for each state, from Seto and Huang, 2023 paper on the National Transportation Noise Exposure Map for the United States.
Each state file has the following variables:
- GEOID - tract ID
- NAME - tract name from American Community Survey
- variable - ACS table variable name for total population (B01003_001)
- estimate - 5-year (2016-2020) ACS tract estimate for total population
- moe - 5-year (2016-2020) ACS tract margin of error for total population
- coverage_all - sum of the coverage for tract
- coverage_na - sum of the NA (missing) pixels for tract
- coverage_4050 - sum of the pixels in the ≥45 to 50 dB LAeq range for tract (note it's 45 because that's the lower limit for BTS modeling)
- coverage_5060 - sum of the pixels in the ≥50 to 60 dB LAeq range for tract
- coverage_6070 - sum of the pixels in the ≥60 to 70 dB LAeq range for tract
- coverage_7080 - sum of the pixels in the ≥70 to 80 dB LAeq range for tract
- coverage_8090 - sum of the pixels in the ≥80 to 90 dB LAeq range for tract
- coverage_90 - sum of the pixels in the ≥90 dB LAeq range for tract
- noise4050n - estimate of number of persons exposed to noise LAeq ≥45 to 50 dB
- noise5060n - estimate of number of persons exposed to noise LAeq ≥50 to 60 dB
- noise6070n - estimate of number of persons exposed to noise LAeq ≥60 to 70 dB
- noise7080n - estimate of number of persons exposed to noise LAeq ≥70 to 80 dB
- noise8090n - estimate of number of persons exposed to noise LAeq ≥80 to 90 dB
- noise90n - estimate of number of persons exposed to noise LAeq ≥90 dB
- noise4050p - estimate of proportion of persons exposed to noise LAeq ≥45 to 50 dB
- noise5060p - estimate of proportion of persons exposed to noise LAeq ≥50 to 60 dB
- noise6070p - estimate of proportion of persons exposed to noise LAeq ≥60 to 70 dB
- noise7080p - estimate of proportion of persons exposed to noise LAeq ≥70 to 80 dB
- noise8090p - estimate of proportion of persons exposed to noise LAeq ≥80 to 90 dB
- noise90p - estimate of proportion of persons exposed to noise LAeq ≥90 dB
- geometry information - either as sf geometry in the R rda files, or as shapefile in the shapefiles. All geometry includes coordinate system info in the files themselves.
R instructions
To read the rda files in R, first install and load the sf package. The read in the state file into a R variable:
library(sf)
WA_map <- readRDS("tractresultWA.rds")
Shapefile instructions
Shapefiles can be simply added into ArcGIS or other GIS software. Note because of the limitation in field name length, the above variable names are truncated in the shapefiles. It should be obvious which ones are which. But, the field names in the shapefiles are shortened to (e.g., "variabl", "estimat", "cvrg_ll" (for coverage_all), "covrg_n" (for coverage_na), "cv_4050" (for coverage_4050), "ns4050n" (for noise4050n), "ns4050p" (for noise4050p).
Estimating Population-Weighted Exposures
In some cases, it may be useful to compute a single noise exposure metric for the census tract instead of having separate estimates of population exposed to various noise level categories. In some analyses, others have used the tract-level median exposure. If needed, this can be easily computed from the original BTS noise raster files. However, the median noise level of the raster pixels within a tract may not represent population exposure very well if the population is not evenly distributed in the tract. Instead, we recommend using a population-weighted noise exposure metric for the tract.
The population-weighted noise exposure metric for noise levels LAeq ≥45 dB can be readily calculated in either R or ArcGIS using the data provided in the files:
Popn-weighted Noise Exposure = (noise4050n * 47.5 dBA) + (noise5060n * 55 dBA) + .... + (noise90n * 90 dBA) / (noise4050n +...+ noise90n)
Save the variable naming information and instructions
Download README Documentation with the above instructions
Download R sf data objects
Download shapefiles
Questions
Robots don't index this page