CMS: DARTE Annual On-road CO2 Emissions on a 1-km Grid, Conterminous USA, 1980-2012

This data set provides a 33-year, 1-km resolution inventory of annual on-road CO2 emissions for the conterminous United States based on roadway-level vehicle traffic data and state-specific emissions factors for multiple vehicle types on urban and rural roads as compiled in the Database of Road Transportation Emissions (DARTE). CO2 emissions from the on-road transportation sector are provided annually for 1980-2012 as a continuous surface at a spatial resolution of 1 km. This project was funded by NASA's Interdisciplinary Science (IDS) and Carbon Monitoring System (CMS) programs.

DARTE is the first nationally consistent inventory of US on-road CO2 emissions built from bottom-up source activity data and establishes a national benchmark for monitoring, reporting, and verification of emissions that are vital for regulating greenhouse gases, and provided insights into how urban areas contribute to climate change.

DARTE combined the Federal Highway Administration's (FHWA's) Highway Performance Monitoring System (HPMS) roadway-level vehicle miles traveled with year- and state-specific emissions factors for five vehicle types on six classes of urban and rural roads. Vehicle emissions were estimated directly at the scale of individual road segments without the need to downsize emissions using spatial predictors. This approach helped to refine the uncertainty in the spatial distribution of road vehicle emissions and showed the highly nonlinear relationship between population density and emissions.

Inventories are provided in 33 individual GeoTIFF files with 1-km spatial resolution. Each GeoTIFF file is a separate year of CO2 emission estimates. The CO2 emissions data are also proved as an ESRI file geodatabase with a regular 1-km "grid" polygon geometry (7.5 GB uncompressed) and a comma separate value file with identical fields as in the geodatabase.

This data set provides a 33-year, 1-km resolution inventory of annual on-road CO2 emissions for the conterminous United States based on roadway-level vehicle traffic data and state-specific emissions factors for multiple vehicle types on urban and rural roads as compiled in the Database of Road Transportation Emissions (DARTE). CO2 emissions from the on-road transportation sector are provided annually for 1980-2012 as a continuous surface at a spatial resolution of 1 km.

DARTE is the first nationally consistent inventory of US on-road CO2 emissions built from bottom-up source activity data and establishes a national benchmark for monitoring, reporting, and verification of emissions that are vital for regulating greenhouse gases, and provided insights into how urban areas contribute to climate change.

DARTE combined the Federal Highway Administration's (FHWA's) Highway Performance Monitoring System (HPMS) roadway-level vehicle miles traveled with year- and state-specific emissions factors for five vehicle types on six classes of urban and rural roads. Vehicle emissions were estimated directly at the scale of individual road segments without the need to downsize emissions using spatial predictors. This approach helped to refine the uncertainty in the spatial distribution of road vehicle emissions and showed the highly nonlinear relationship between population density and emissions.

Spatial Coverage

Conterminous United States

Spatial Resolution

1-km resolution

Temporal Coverage

The data cover the period 1980-01-01 to 2012-12-31

Temporal Resolution

Annual

Study Area: (All latitudes and longitudes given in decimal degrees)

 

Site	Westernmost Longitude	Easternmost Longitude	Northernmost Latitude	Southernmost Latitude
Conterminous United States	-130.257	-62.0377	50.31	24.0928

 

Data File Information 

Inventories are provided in 33 individual GeoTIFF files with 1-km spatial resolution. Each GeoTIFF file is a separate year of CO2 emission estimates.

The CO2 emissions data are also proved as an ESRI file geodatabase with a regular 1-km "grid" polygon geometry (7.5 GB uncompressed) and a comma separate value file with identical fields as in the geodatabase.

ORNL DAAC Data Processing Note:  The 33 individual GeoTIFF files were derived from the investigator provided ESRI file geodatabase, DARTE.gdb, by rasterizing the Mg_co2_YYYY field.

 

Annual GeoTIFF Files

GeoTIFF files are named DARTE_Mg_CO2_YYYY, where YYYY is the year 1980 - 2012, respectively. CO2 emissions are in units of Megagrams (metric tons) of CO2 emitted per year.

 

Spatial Data Properties

Spatial Representation Type: Raster
Pixel Depth: 64 bit
Pixel Type: float
Number of Bands: 1

Raster Format: GeoTIFF
Source Type: continuous

Nodata Value:  -1.79769313486e+308
Scale Factor: none
Offset: none
Endian Type: NA

Number Columns: 5,071
Column Resolution: 1,000 meter
Number Rows: 3,560
Row Resolution: 1,000 meter

Extent in the items coordinate system
North: 1952000
South: -1608000
West: -2736000
East: 2335000

xll corner: -36966200
yll corner: -29247500
Cell Geometry: area
Point in Pixel: corner

 

Spatial Reference Properties

Projection:   Lambert_Conformal_Conic
False_Easting:   0.00000000
False_Northing:  0.00000000
Central_Meridian:  -97.00000000
Standard_Parallel_1:  33.00000000
Standard_Parallel_2:  45.00000000
Scale_Factor:   1.00000000
Latitude_Of_Origin:  40.00000000
Linear Unit:   MeterGeographic Coordinate System:  GCS_WGS_1984
Datum:  D_WGS_1984
Prime Meridian:  Greenwich
Angular Unit:  Degree

Extent in the data file coordinate system: 

Top:  1952000 m
Right:  2335000 m
Left:  -2736000 m
Bottom:   -1608000 m

 

ESRI Geodatabase

DARTE.gdb.zip

This is compressed ESRI file geodatabase. The geodatabase contains a single polygon feature class of the DARTE 1-km "grid", with several database fields including emissions for each year 1980 - 2012.

The Spatial Reference properties are the same as those referenced above for the raster GeoTIFF files except that this is a polygon feature class with regular 1-km polygon geometry.  The fields for each polygon are defined in the table below:

The column headers are as follows:

 

Column name	Type	Description
cellid	INTEGER	Unique ID of 1-km grid covering the coterminous U.S. The grid Coordinate system is projected using the projection described below.
ddx	FLOAT	Longitude of SW corner of grid cell in decimal degrees
ddy	FLOAT	Latitude of SW corner of grid cell in decimal degrees
lon_meters	INTEGER	Longitude of SW corner of grid cell in meters
lat_meters	INTEGER	Latitude of SW corner of grid cell in meters
Mg_co2_YYYY	FLOAT	Megagrams (metric tons) of carbon dioxide emitted in year YYYY. The range of YYYY is 1980 through 2012.

 

Comma-Delimited File

DARTE_1km_co2.csv

This is a compressed, comma-delimited file that reports annual carbon dioxide emissions from on-road vehicle activity in the conterminous United States for the years 1980 through 2012.  Data corresponds with the 1-km resolution grid cells of the DARTE.gdb file geodatabase.

 

Although the United States has 5% of the world's population and 30% of the world's automobiles, it emits 45% of global transportation CO2 emissions (1.57 billion metric tons in 2012).  These emissions account for 28% of US fossil fuel CO2 emissions.  Despite the large contributions of on-road emissions, the spatial distribution of emissions are highly uncertain and poorly quantified especially at substate and urban scales. Reducing the uncertainty of on-road CO2 emissions at finer spatial scales is central to understanding the determinants of motor vehicle emissions, constraining carbon budgets, and supporting greenhouse gas emission monitoring and abatement, particularly at the scale of cities.  DARTE can provide valuable information to local and regional climate change mitigation initiatives by providing detailed city-scale greenhouse gas emissions estimates and a tool to assess response to policy.  DARTE's bottom-up approach differs markedly from other emissions estimates that rely on population as a linear predictor of vehicle activity.  DARTE was used to quantify the spatiotemporal varying effects of population density, income, employment and transit use on on-road CO2 emissions across the United States.  DARTE is derived from HPMS roadway-level vehicle miles traveled (VMT) combined with year- and state-specific emissions factors for five vehicle types to calculate CO2 emissions from motor gasoline and diesel fuel consumption on six classes of urban and rural roads.

Within the HPMS database, the annual vehicle miles traveled (VMT) information was inspected by county and functional class to identify potential outliers or structural breaks in the dataset.  A filtering algorithm flagged any observation in an individual county/functional class time series if the magnitude of the year-on-year difference between an observation and adjacent years was greater than two standard deviations from the mean year-on-year difference of that time series.  Of the 761,759 observations in the dataset, roughly 10% were flagged and replaced by the filtering procedure.

Direct quantification of the uncertainty in US on-road emissions is made impossible by the absence of independent data sources against which to compare government estimates.  Further external validation of emission inventories is hampered by the lack of independent measurements of source activity, but it is informative to compare the effects of different model methodologies and proxy performance on the consistency of emissions estimates.  Because DARTE estimates were not calculated using population or road density, we were able to evaluate the performance of these variables as spatial predictors by comparing DARTE with other well-known inventories. We aggregated DARTE’s roadway-scale emissions so that we could compare directly to the native resolutions of EDGAR and the Vulcan Project.  Comparing DARTE with these existing downscaled inventories, biases of 100% or more were found in the spatial distribution of urban and rural emissions.  These biases were mainly the result of mismatches between inventory downscaling proxies and the actual spatial patterns of vehicle activity at urban scales.  The DARTE 1-km grid cells nest exactly within the Vulcan Product’s 10-km grid cells for ease of comparison and combination.

Following is a synopsis of the DARTE development steps. A full description of the procedure and analyses of results are provided in Gately et al. (2015).

The Highway Performance Monitoring System (HPMS) is a traffic monitoring program overseen by the Federal Highway Administration. All U.S. states and the District of Columbia are required to submit annual estimates for a wide variety of traffic and highway condition data to HPMS in compliance with federal statute. This data has been archived in a digital tabular format since 1980. The HPMS dataset used to generate the 1980-2009 emissions estimates in DARTE is well described in the archived versions of the HPMS Field Manual. The core data features of the HPMS are individual road segments that are defined by the state transportation departments. For each road segment HPMS reports the center-line length in miles, annual average daily traffic (AADT), functional class, urban/rural context, and county. AADT is a measure of the number of vehicles that traverse the road segment on an average day, adjusted for seasonal and day-of-week variation.

Annual vehicle miles travelled (VMT) on each available road segment in HPMS were calculated by multiplying AADT by the length of the road segment and then by 365 days. Road-segment VMT were aggregated to the lowest common level of geography including 12 roadway functional classes included in HPMS, six for urban roads and six for rural roads. HPMS data were aggregated to counties by using the reported functional classes and a geocoded road network for the U.S. that is also partitioned by these functional classes. State-level VMT data for rural local roads were integrated into the county-level database by allocating state totals by functional class to each county. 

For each year, county, and functional class, annual VMT from HPMS were partitioned into the five vehicle types used by the Highway Statistics Series Table VM-4: Passenger cars, passenger trucks (SUV, pickups, minivans), buses, single-unit trucks, and combination trucks.  VMT was then used to calculate fuel use and associated CO2 emissions. A calibrated fuel economy and VMT by vehicle type was used to calculate gallons of motor gasoline and diesel fuel consumption, which was then converted to CO2 emissions for each year, county, and road functional class.  Emissions were assigned to a geographic information system (GIS) layer of the US road network obtained from the 2012 Census TIGER/Line geodatabase.

Andres R.J., et al. (2012) A synthesis of carbon dioxide emissions from fossil-fuel combustion. Biogeosciences 9:1845–1871. doi:10.5194/bg-9-1845-2012 .

Gately, C.K. et al., PNAS, 2015.  Cities, traffic, and CO2: A multidecadal assessment of trends, drivers and scaling relationships.   vol. 112 no. 16:4999-5004.  doi:10.1073/pnas.1421723112 . 

Gurney K.R, et al. (2009) High resolution fossil fuel combustion CO2 emission fluxes for the United States. Environ Sci Technol 43:5535–5541. doi: 10.1021/es900806c

Highway Performance Monitoring System Field Manual. Office of Highway Policy Information, Federal Highway Administration, Washington DC, 2010. URL:  http://www.fhwa.dot.gov/ohim/hpmsmanl/hpms.htm

Highway Statistics Series, Table VM-4; 1980−2012; Federal Highway Administration: Washington, DC. http://www.fhwa.dot.gov/policyinformation/statistics Accessed July 1, 2013.

JRC/PBL (2011) Emission Database for Global Atmospheric Research (EDGAR) (European Comission, Joint Research Centre/Netherlands Environmental Assessment Agency, Ispra, Italy), Version 4.2 Available at edgar.jrc.ec.europa.eu. Accessed May 1, 2014.

US Census Bureau (2012) TIGER/Line Shapefiles (US Census Bureau, Washington DC).  Available at www.census.gov/geo/maps-data/data/tiger-line.html. Accessed July 1, 2013.