__________________________________________________

               A CENTURY OF SPRAWL IN THE UNITED STATES:
                        GEOGRAPHIC TABULAR DATA

              Chris Barrington-Leigh and Adam Millard-Ball
           __________________________________________________


Table of Contents
_________________

1 FILE FORMAT
2 GEOGRAPHIC UNITS AND TIME SERIES
3 FIELDS/COLUMNS
4 TIPS AND CAVEATS
5 CITATION
6 CONTACT


This file describes the geographic tables published, along with network
graphs described separately, as part of the 2015 PNAS article DOI:
10.1073/pnas.1504033112.  The full citation is below. The dataset
released with this article is archived independently as
[DOI:10.5061/dryad.3k502] ([http://dx.doi.org/10.5061/dryad.3k502]).


1 FILE FORMAT
=============

  Three formats are provided:
  - .pandas files are for use with the Python pandas data analysis
    library ([http://pandas.pydata.org]). They can be loaded with the
    read_pickle() function from within pandas.
  - .tsv files are tab-separated and can be used with any spreadsheet or
    statistical analysis package. Line endings are the simple "\n" POSIX
    type.
  - .shp files are for use with Geographical Information System
    software, e.g. QGIS or ArcGIS. The geographic boundary files are
    from the US Census Bureau
    ([https://www.census.gov/geo/maps-data/data/tiger-line.html]), 2014
    vintage.

  The first two are compressed into one ZIP file. The .shp files are
  compressed into another ZIP file.


2 GEOGRAPHIC UNITS AND TIME SERIES
==================================

  The following files provide the cross-sectional estimates (i.e., the
  stock of nodes):
  - all .shp files
  - Barrington-Leigh-Millard-Ball-PNAS2015-Century-of-sprawl-latest-stock-county.pandas
  - Barrington-Leigh-Millard-Ball-PNAS2015-Century-of-sprawl-latest-stock-county.tsv

  The remaining .pandas and .tsv files provide the full time series (in
  long format) for different geographic aggregations, as follows:
  - US counties
  - US metropolitan regions (CSAs and CBSAs). CSAs are used where they
    are defined; CBSAs are used for metropolitan regions outside a CSA.
  - US states
  - The national level (US)

  The FIPS code lookups are available here (we use the 2013
  delineations): [http://www.census.gov/geo/reference/ansi.html]


3 FIELDS/COLUMNS
================

  Please refer to the Supporting Information for details of sources of
  the underlying data and the process for constructing the dataset.

  The data consist of three measures of street-network sprawl:
  - degree (mean nodal degree)
  - deadend (fraction of deadends)
  - fourway (fraction of 4+ degree nodes)
  - N denotes the number of nodes

  Standard errors of the means are also provided (prefixed se_). They
  reflect an estimate which relies on assuming unbiased measurement and
  sampling, and are simply the sample variance divided by the square
  root of the number of nodes.

  Suffixes denote the series:
  - _parcel denotes the parcel-based series (the suffix is abbreviated
    to Pcl in the .shp files)
  - _TIGERcensus denotes the census-based series (abbreviated to Cen)
  - _TIGER denotes the TIGER/Line series (abbreviated to TGR)

  Notes:
  - The parcel-based series has partial coverage of the US. See the
    Supporting Information.
  - The year-by-year parcel-based and census-based series describe the
    characteristics of newly constructed nodes in a given year.
  - The TIGER/Line series describes the characteristics of the stock of
    nodes in a given year.
  - The allyears and the shape files describes the characteristics of
    the stock of nodes in 2012 (parcel-based) or 2013 (TIGER/Line),
    regardless of the series.
  - The latest stock estimates differ between the parcel-based and
    TIGER/Line series for reasons discussed in Section 1.1 of the SI. In
    brief, the parcel-based series excludes some service or other access
    roads without associated buildings. Moreover, available parcel data
    cover only a subset (in some cases none) of each state or
    metropolitan area.
  - For more information about the differences between the series, refer
    to the Supporting Information.
  - Note that the data are limited to urbanized areas, defined as block
    groups where the majority of blocks were classified as urban in the
    2010 Census.


4 TIPS AND CAVEATS
==================

  - See the Supporting Information for strong caveats on the use of
    TIGER and TIGERcensus data.
  - Ignore low-count cells (see columns N_*)
  - Do not use years before 1920 from any dataset. They are provided for
    completeness, but county parcel records are less reliable prior to
    1920. We recommend collapsing these years to a "pre-1920" value.
  - Use rolling means to smooth years before 1930, and for longer
    periods of time for smaller units of analysis


5 CITATION
==========

  Barrington-Leigh, Christopher and Millard-Ball, Adam (2015), "A
  Century of Sprawl in the United States." Proceedings of the National
  Academy of Sciences, DOI: 10.1073/pnas.1504033112


6 CONTACT
=========

  For further questions, please contact:
  - Chris Barrington-Leigh, McGill University:
    Chris.Barrington-Leigh@McGill.ca
  - Adam Millard-Ball, University of California, Santa Cruz:
    adammb@ucsc.edu