Demographics of Oglala Lakota County

The Keystone Pipeline is an oil pipeline running from the Canadian Tar Sands in Alberta to the Gulf Coast in Texas. The plan for the fourth phase of this project proposes to run a new section of this pipeline under the Missouri River, just upstream of the Oglala Lakota sacred land. The Lakota people rely on this river for their livelihood. In the wake of the Flint, Michigan crisis, the main concern of locals is possible contamination of the water. The implications would be catastrophic leading to the inability to use the river to fish, irrigate crop land or even have clean water to drink.

This map visualizes 2 sets of data obtained from the US Census Bureau on family income and minority populations. The 2 maps show striking similarities. Upon some calculations and research into the maps it was quite apparent that the Oglala Lakota County had the highest percent minority population of any county in the entire United States. Oglala Lakota County also has the 3rd lowest mean family income in the country. It is one of three counties in the United States completely encompassed by a Native American reservation. The Lakota tribe considers the Missouri River sacred since it has been the tribes main source of life since they inhabited the land nearly 1200 years ago.

-Cody J. Crane

UConn MAGIC 2017

American Community Survey Median Household Income Distressed Tracts 2010 to 2013

This visualization uses data from the American Community Survey to display distressed census tracts, which is a tract at 60% or less of the state median household income level. This study ranges from the year 2010 to 2013.

by: Zachary Guarino


“Lost” New England landscape found using LiDAR


At top left, leaf-on 2012 aerial imagery from Connecticut (CTECO); top right is 2010 hillshaded DEM derived from LiDAR data showing stone walls, old road and building foundation; bottom is 1934 aerial photography (available from MAGIC and CT State Library)



New research by Geography graduate student Katharine Johnson and faculty William Ouimet was covered yesterday by National Geographic in their article “Lost” New England Revealed by High-Tech Archaeology. The article features a Q & A with Katharine Johnson, a PhD student in Geography and employee here at MAGIC and the Connecticut State Data Center.

You can read it (and see some cool graphics) here:

The article references a paper by Johnson and Ouimet that was recently accepted and published by the Journal of Archaeological Science about using Light Detection and Ranging (LiDAR) to identify and analyze the historic agricultural landscape of New England that is now hidden by the forest canopy in aerial photography but is visible using LiDAR. Check it out, here:

Historical boundaries

Have you ever wondered what your county or Congressional district boundaries looked like in 1845? Before 1845?  Then this is the blog post for you.

At UCLA, Jeffrey Lewis, Brandon DeVine and Lincoln Pritcher have developed and made freely available Congressional district boundaries for the entire lower 48 states – all the way back to 1789. Their project draws on research previously done by Kenneth Martis, who also provided advice and source materials for the authors. The boundaries are available as an ESRI shapefile and GeoJSON at a slightly lower resolution. Please visit their website for more information and documentation, as well as to download the shapefiles. The three maps below use their data to show the district boundaries at different points in time.











The project used data from the National Historic GIS, as well as the Newberry Library in Chicago. The Newberry Library provides GIS and KML files for historical county boundaries at the state and national levels. Take, for example, this page for Connecticut which explains the data and directs you to an interactive viewer that allows you to compare modern and historical county boundaries from various dates (see below).



Demographic bits and bytes II

The Census Bureau regularly releases various types of information and statistics. This post contains some of the statistics that have been released over the past few months.

“Unmarried moms”

UnmarriedWomenEdThe Bureau reports that about 6 in 10 recent moms in their early 20s are unmarried. The data, from the 2011 ACS, suggests that 62% of women age 20 to 24 who gave birth in the previous 12 months were unmarried. Compare this with 17% for women age 35 to 39. These numbers were also compared to education levels, which show that as education levels rise, the percent of births to unmarried women decline. The states with the highest percentages of unmarried mothers were Washington DC (51%), Louisiana (49%), Mississippi (48%) and New Mexico (48%). The states with the lowest percentages were Utah (15%) and New Hampshire (20%).  The writers of the report for the Census Bureau note that “..the increased share of unmarried recent mothers is one measure of the nation’s changing family structure…Nonmarital fertility has been climbing steadily since the 1940s and has risen even more markedly in recent years.”

 Per student education spending decreases

For the first time in nearly four decades, the amount of money spent per student in the US has decreased. The Census Bureau reports that this is the first time spending has decreased since it began collecting annual data in 1977. The 50 states and Washington DC spent $10,560 per student in 2011, which was down 0.4% from 2010.

Top state spenders: Spending

  • New York ($19,076)
  • District of Columbia ($18,475)
  • Alaska ($16,674)
  • New Jersey ($15,968)
  • Vermont ($15,925)

Lowest state spenders:

  • Mississippi ($7,928)
  • Arizona ($7,666)
  • Oklahoma ($7,587)
  • Idaho ($6,824)
  • Utah ($6,212)

Connecticut public schools received $796,156 in revenue from Federal sources during the 2011 Fiscal year – however the state with the highest revenue from federal sources was California at $9,990,221 with Texas in second at $7,818,075. Revenue from state sources for Connecticut in 2011 was $3,171,891 in total, while for California it was $37,690,834 with New York in second at $23,188,002. See this news release for more information about expenditures and revenue for public schools.

arab households in the us 2006-2010

Another news brief issued by the Census Bureau gives a “national-level portrait” of US household, and specifically those with Arab ancestries. The brief uses American Community Survey data to determine Arab ancestry. For example, the survey asks for “ancestry or ethnic origin.” From responses collected, the Census Bureau considers anyone who reported being “Algerian, Bahraini, Egyptian, Emirati, Iraqi, Jordanian, Kuwaiti, Lebanese, Libyan, Moroccan, Omani, Palestinian, Qatari, Saudi Arabian, Syrian, Tunisian and Yemeni to be of Arab ancestry.”

The data itself shows that in the US, the population with Arab ancestry increased from 850,000 in 1990 to 1.2 million in 2000. The ACS data from 2006-2010 shows that an estimated 1.5 million people (0.5 percent of total population) with Arab ancestry now live in the US – a 76% increase since 1990. The number of households has also increased – from 268,000 in 1990 to 511,000 in 2010.


 economic characteristics of households in the US – 2011 (4th quarter)

And for all households in the US, the Census Bureau has released a set of tables that detail their economic characteristics for the 4th quarter of 2011. The available data ranges from median monthly household cash income to labor-force status, and receipt of benefits from selected means-tested noncash benefit programs.

Quarterly summary of state & local government tax revenue – 2013 (1st quarter)

The Census Bureau has released statistics for the first quarter of 2013 that detail the quarterly tax revenue statistics on property, sales, license, income and other taxes. Statistics are available for individual state governments as well as at the national level. Some of the categories broken down by state include various licenses, such as alcoholic beverages, hunting and fishing, and motor vehicles. Property, sales and gross receipts and income taxes are also included in the table.

Click here for an interactive visualizations of taxes collected by state

Click here for an interactive visualization of taxes collected by state

Computer and internet use in the US

The Census Bureau has also released information about internet use within households, as well as the impact of smart phones. The report was written using data collected in a supplement to the Current Population Survey (CPS) in July 2011. The supplement includes questions “about computer ownership, internet use both inside and outside the home, and the additional devices that people use to go online.” According to the report, the Census Bureau has asked questions related to computer use since 1984, and internet ComputerUseuse since 1997.

As you may have guessed, computer use has changed drastically in Americans’ homes since 1984. In 2011, 75.6% of households reported having a computer – compared to 8.2% in 1984 and even 61.8% in 2003 (this surprised me). In 2011, 71.7% of households reported accessing the internet, while only 18% did so in 1997 – via that now nostalgic dial-up modem. Only 54.7% had access in 2003.

The report also discusses the disparities in internet use amongst different racial, ethnic, and even age groups in the United States. For example, in 2011, internet use was most common in households where the householder was between 35-44 years old (81.9%). Households that had householders over the age of 55 had much lower rates. Additionally, the report found that householders with higher levels of education also report higher rates of internet usage. For more detailed information, click the link above to access a PDF of the report.


Summer 2013 Geofocus Newsletter

GF_SummerThe Summer 2013 issue of Geofocus is out!  Please visit the CT User to User website to access the newsletter:

In this issue, we here at MAGIC have announced that there are now 2 new aerial photography centerpoint indexes available in KML format, and .shp format coming soon! The indexes are for the 1934 and 1965 aerial photography collections, which encompass the entire state.

The indexes can be used to preview photos and to download a TIFF or PDF version of the photograph. The indexes serve as a reference point for the photographs, and none of the photographs have been georeferenced or orthorectified.

To access the indexes, please visit our website:

Are you a pixel?

Many social media sites allow the option for “geotagging” – letting others know where you did something. Where you took a photo, where you made a comment or status update, etc. Obviously the most common two are Twitter and Facebook, but more and more applications ask for your location when you share a bit of information such as a status update, a photograph, or a 140-character bit of wisdom. Though this might seem relatively mundane, the location where you did or said something combined with what that thing was is one of the most powerful datasets available to not only geographers, but a slew of other disciplines including psychology, sociology, marketing and anthropology, to name a few. It is now apparent that a person, or the location where a person does something, is a data point or pixel in the context of the image that is the wider data world. So make your data count!


People as pixels: Image by artist Craig Alan and from TechEBlog

Though this is not a new concept by any means, every day more and more individuals are using this information or finding new ways to create interesting and informative maps or visualizations that provide commentary on location-specific attributes of not only the United States, but the world. In addition to maps that contain spatial data, there are visualizations of textual data from social media sources, and social media and crowd-sourced mapping have also been used for crisis-mapping situations. For example, using the Twitter Search API, ESRI created maps that include a spatial representation of tweets overlaying other information for both the 2011 Japan earthquake, and the recent tornadoes in Oklahoma. Facebook even has its own page called Facebook Data Science that gleans information and creates visualizations that it posts.

Though this post is only skimming the surface of much broader academic and theoretical issues, here are some examples of maps and visualizations that we appreciated, and even some ways that you can make your own. If you know of other interesting examples that we missed, let us know!

The Geography of Hate

  • Using Twitter, all tweets containing each ‘hate word’ were aggregated to the county level and normalized by the total Twitter traffic in each county.

Geography and Football

  • Based on user’s status messages and their relationship to others, Facebook created a map showing locations of where people lived who supported specific teams.
  • This blog post went even further with that map, to try to explain some specific unexpected distributions (i.e. Minnesota Vikings fans living in North Carolina).

Mapping the world by “check-ins”

  • This map, posted on Facebook Data Science and created by Paul Butler in 2010, shows an image of the world created by visualizing where people “checked-in” on Facebook – each pixel represents 3 check-ins. The visualizations are stunning.

Image courtesy of Joel Salisbury

Foursquare, in collaboration with Samsung, has recently also created the “Time Machine,” which allows users to create infographics of all of their check-ins, “ever”! Foursquare users can access this feature by going to In addition to the creation of a beautiful map (see below), foursquare will crunch the data to create graphs and other comparisons between types of places visited, recent check-ins, frequent connections, most visits to specific locations, and all kinds of mind-boggling statistics you might not have known about yourself until you see them summarized in a pretty image.



Other maps that have recently been in the news include 122 that depict dialect differences itsgrinderfor different regions in the United States. Joshua Katz, a PhD student in Statistics at UNC created a model to define dialect regions based on The Cambridge Online Survey of World Englishes, by Bert Vaux and Marius L. Jøhndal. The”Results” tab on their website also shows maps of various responses as points if you log in and take the survey yourself. Though only 22 of Joshua’s maps were published in this Business Insider article, all 122 can be seen here. Being from Rhode Island, I was a little surprised that “grinder” didn’t make it on to the subs vs hoagies map and similarly shocked that over 80% of respondents in Providence, Hartford and Boston said it was called a “sub”!)

Edit: Because I was so surprised by the lack of “grinder,” I went back to look at the original survey point data from the Cambridge Online Survey. Sure enough, the responses in the Northeast are much more nuanced than the above map suggests. Not only is “grinder” a fairly common response, but so is “hero” in the NYC and Long Island area. Neither response is really represented in the map above. Below is the map of the original point data. Another copy can also be accessed here.

















Finally, if you’d like to do some textual mapping, this article explains how to download your Twitter archive and create a word or tag cloud that visualizes your interests or most commonly-tweeted themes (a word cloud for this post is below!). Similarly, if there is a topic you find interesting, will allow you to search for every few days’ worth of hashtag topics on Twitter.








Another edit – just discovered a map, via GISLounge, where Twitter users are mapped globally by the type of smart phone that they use. You can pan and zoom to look at specific areas of the world. Pretty amazing!

Hurricane season by the numbers

As residents of Connecticut, Hurricane Sandy still remains fresh in our recent memory. As one of the most costly hurricanes in New England and even US history, the storm reminds us how important it is to be prepared. Hurricane season in the US begins on June 1st and lasts through November 30th. Every year, the Census Bureau produces statistics to aid emergency planning, preparedness and recovery efforts.

Historical Storms


Track of Hurricane Bob (1991)

We all have our memories of hurricanes-past as well, especially those that have made landfall in New England. The Great Hurricane of 1938, the 1944 hurricane, Carol (1954), Gloria (1985), Bob (1991), Irene (2011) – we certainly won’t forget those names any time soon. All of these storms were so destructive that each one of those names has been retired from the Atlantic tropical cyclone naming convention. I remember, having grown up in coastal Rhode Island, when Bob made landfall and the eye eerily and silently passed over our house to give us a brief respite from the furious wind that sounded like a freight train.

The National Hurricane Center maintains all of the historical data about these storms – their tracks, various statistics, and GIS data as well. So now, 22 years later, I can look up Bob’s track, intensity, and dates. If you’re interested, click here for more information from the National Hurricane Center data archive.


1938: Flooding from the storm surge in Bushnell Park, Hartford CT

The storm surge of the 1938 hurricane was between 14 and 18 feet in coastal Connecticut, which caused the Connecticut River in Hartford to reach a level of 35.4 feet – 19.4 feet above flood stage. There was widespread destruction in southern New England, and upwards of 500 deaths (Source).

For pre- and post-storm aerial imagery for Hurricane Sandy, check out the USGS Coastal Change Hazards site.





Census Bureau – Facts for Features: Hurricane Season

In the Hurricane’s Path


The number of hurricanes during the 2012 Atlantic hurricane season, with only two of them as major hurricanes (Category 3-strength or higher). However, one of the major hurricanes was Hurricane Sandy. It struck southeastern Cuba at Category 3 strength, then made landfall in New Jersey as a post-tropical cyclone. It was the second costliest cyclone on record (not adjusted for inflation) at $50 billion, ranking only behind Hurricane Katrina from 2005. The only other hurricane to make landfall in the U.S. in 2012 was Hurricane Isaac, which hit Louisiana.
Source: National Hurricane Center

82.2 million

Population as of July 1, 2012, of coastal states stretching from North Carolina to Texas — the areas most threatened by Atlantic hurricanes. An estimated 26.2 percent of the nation’s population live in these states.
Source: 2012 Population Estimates

34.1 million

Population in 1960 of the states stretching from North Carolina to Texas. Approximately 19 percent of the nation’s population lived in these areas at that time.
Source: 1960 Census <>


Percentage growth of the population of the states stretching from North Carolina to Texas between 1960 and 2012.
Source: 2012 Population Estimates and 1960 Census


Collective land area in square miles of the states stretching from North Carolina to Texas.
Source: 2010 Census <>

10 Year Anniversary of Hurricane Isabel


The costliest and deadliest hurricane of 2003, Hurricane Isabel made landfall in the U.S. on the Outer Banks of North Carolina on Sept. 18, destroying many homes on the barrier island. Isabel later moved north through Virginia and Washington, D.C., and ended up causing about $3 billion in damage to the mid-Atlantic region.
Sources: National Hurricane Center

Category 2

The strength of Hurricane Isabel at landfall based on the Saffir-Simpson scale, with maximum sustained winds measured at 105 mph. Isabel reached a peak as a Category 5 storm on Sept. 11 south of Bermuda, but gradually weakened as it approached landfall.
Source: National Oceanic and Atmospheric Administration <>


Counties that encompass the land area of the Outer Banks of North Carolina. The area includes parts of Currituck, Dare and Hyde counties.
Source: Census Quickfacts <>


Population of Currituck County, Dare County and Hyde County in North Carolina in 2012.
Source: 2012 Census Population Estimates


The number of occupied housing units in Currituck, Dare and Hyde counties combined.
Source: 2007-2011 American Community Survey Estimates

$236,500; 321,200; and 93,600

Median home value of owner-occupied units in Currituck, Dare and Hyde counties, respectively.
Source: 2007-2011 American Community Survey Estimates

32.0, 19.5 and 25.1 minutes

Mean commuting time to work for residents in Currituck, Dare and Hyde counties, respectively.
Source: 2007-2011 American Community Survey Estimates

7.8%, 11.1% and 25.1%

The percent of people who live below poverty level in in Currituck, Dare and Hyde counties, respectively.
Source: 2007-2011 American Community Survey Estimates

History of Hurricane Naming Conventions


The name of the first Atlantic storm of 2013. Hurricane names rotate in a six-year cycle with the 2013 list being a repeat of the 2007 names.
Source: National Hurricane Center <>


The number of hurricane names officially retired by the World Meteorological Organization. Although hurricane names are recycled every six years, for reasons of sensitivity, hurricane names that were so deadly and costly that re-use of the name would be considered inappropriate are retired.
Source: World Meteorological Organization <>


The year the Weather Bureau officially began naming hurricanes.
Source: Atlantic Oceanography and Meteorological Laboratory <>


In one of the busiest Atlantic hurricane seasons on record, 28 named storms formed, forcing use of the alternate Greek alphabet scheme for the first time. When the National Hurricane Center’s list of 21 approved names runs out for the year, hurricanes are named after Greek letters. Of the 28 named storms in 2005, 15 were hurricanes, with four storms reaching Category 5 status (Dennis, Katrina, Rita and Wilma) and three more being considered major.
Source: Atlantic Oceanography and Meteorological Laboratory <>

Preparing for Emergencies Using Census Bureau Statistics

The growth in population of coastal areas illustrates the importance of emergency planning and preparedness for areas that are more susceptible to inclement weather conditions. The U.S. Census Bureau’s official decennial census and population estimates, along with annually updated socio-economic data from the American Community Survey, provide a detailed look at the nation’s growing coastal population. Emergency planners and community leaders can better assess the needs of coastal populations using Census Bureau statistics.

State government tax collections

The Census Bureau reports that state government tax collections reached a record high of nearly $800 billion in fiscal year 2012.

The information comes from the 2012 Annual Survey of State Government Tax Collections, a report containing statistics on the fiscal year tax collections of all 50 state governments. Tax categories include property taxes, license taxes, and income taxes but are broken down further to also include motor fuel taxes, severance taxes, and hunting license taxes as well.

Data is available in American FactFinder as a detailed table that indicates the amount and type of taxes each state government collected. For example, Connecticut collected $15,419,556 in taxes for FY2012, and of those the category that comprised the most was Income Taxes ($7,996,509) and Sales Tax ($6,677,074)

Click image to interact with data

Click image to interact with data

California ranked first on the Total Tax list, while Massachusetts ranked 10th, and Connecticut ranked 18th. New Hampshire was 49th, and South Dakota 50th.