Sunday, November 20, 2011

11/20/2011 Tornadoes!

I have recently become very interested in “GIS in Disaster Response”, and am investigating ways to serve our country in times of Disasters. Since more and more people are now living on a continually-unstable planet, “Disasters” are certainly not going to “just go away” – they happened in the past, they are happening now, and they will continue to occur in the future.

At a seminar I attended last week, one speaker spoke of his frustration at not having (geographic) information regarding “tornado damage” a few years ago in New Hampshire – the data had not been assembled, in one place, to allow the generation of $ damage estimates across multiple counties. (That situation has improved with the creation of the New Hampshire Mosaic Parcel Map)

This led me to think about a map showing tornadoes – their locations and their paths. Which led me to the Tornado History Project:

I selected Massachusetts (my home, but not a hotbed of tornado activity, or so I thought):

The map shows 152 tornados from 1951 through 2008, with 102 fatalities – nothing to make light of. At the bottom of the page, I click the link for the Storm Prediction Center’s (SPC) historical tornado data file, and get taken to the page where I can download csv files of Tornado/Hail/Damaging Wind:

This data page is a FABULOUS page! Scroll up, and there is a beautiful collection of heat maps, thematic maps, plot charts of tornado reports over time, small-multiple time series maps, monthly Summary Charts, and a map of Actual U.S. tornadoes for January through July, 2011:

Obviously I can not discuss all those visualization tools today – each will be a separate blog this coming week! But for today, I have discovered that, although each event is unique and tragic, tornadoes occur repeatedly over time, the historical data exist, and the visualization tools are already out there. From the data comes knowledge, and with knowledge we can prepare ourselves to respond when disasters (in this case, tornadoes) occur in the future.

Friday, November 18, 2011

11/18/2011 Are Maps Dangerous?

In the November 14-November 20, 2011 issue of Business Week, the article on the drug war in Mexico (“The Drug War in Mexico, Now on the Blogosphere”, page 45) states that “The editors at Nuevo Laredo en Vivo recently compiled [Twitter] reports to create a map of drug sale locations and suspected lookouts.”
article link
The map was published in a blog on September 25, 2011
blog link

It is good that the blog published this map, because it is now unavailable:
link to map page

I do not speak Spanish, and I have not contacted the website to ask why the map was taken down. All I do know is that “The site’s contributors included Maria Elizabeth Macias, the beheaded victim”. It seems that yes, maps are dangerous, and the results can be fatal.

Congressional Redistricting follow-up
Regarding my Nov 8, 2011 blog, there have been some updates/corrections. Unfortunately for Texas, on Sept 19 the U.S. Department of Justice said that their Congressional redistricting map does not comply with Section 5 of the Voting Rights Act, and a court challenge was filed on Sept 23. An excellent discussion of the Texas situation may be found here
blog link

In Arizona, on Tuesday Nov 1, the Chair of the Independent Redistricting Commission was impeached by Governor Jan Brewer and the Arizona Senate, because the Governor did not like the maps that the Commission was generating. Unfortunately for the Governor, on Thursday Nov 17, the Arizona Supreme Court reinstated Colleen Mathis as chair of the Commission, ruling that the impeachment was not lawful. This is fascinating, but a little like watching a train wreck in progress.

Further sources on redistricting:
Michael McDonald, a redistricting expert at George Mason University in Virginia
Jennifer Steen, Assistant Professor of Political Science, Arizona State University
Meredith McGehee, Project Director, Americans for Redistricting Reform
Keesha Gaskins, New York University’s Brennan Center for Justice

Tuesday, November 8, 2011

11/8/2011 new Congressional District maps

Changes following the 2010 census
On December 21, 2010 the U.S. Census Bureau released its official apportionment results for congressional representation. The changes will be in effect for the U.S. elections in 2012, provided no legal challenges to the results significantly change the results.
Gain four seats
Texas
Gain two seats
Florida
Gain one seat
ArizonaGeorgiaNevadaSouth CarolinaUtahWashington

Lose one seat
IllinoisIowaLouisianaMassachusettsMichiganMissouriNew JerseyPennsylvania

Lose two seats
New York
Ohio
link to wikipedia article

The lead story on today’s (Massachusetts) news was 1 – over 50 Massachusetts cities and towns have elections today (out of 351 cities and towns), and 2 – the new Congressional maps have been released, showing how Massachusetts went from ten Congressional Districts down to nine.

My gut sense has been that although the maps for the State House and Senate were released in late October (and Governor Patrick signed them Thursday, November 3, 2011), the new map for the nine Congressional Districts would be released “as late as possible” to avoid rancor/conflict/stalemate. I want to see if Massachusetts is “quick” on this process, “slow” on this process, or “somewhere in between”. It turns out that every state has its own specific timetable – for example, the Florida Legislature is specifically required to redistrict the State House and State Senate districts during the 2012 Legislative Session [Florida Constitution, Art III, s. 16 (a)], and their Congressional Redistricting Plan is scheduled for their Legislative approval between January 10 and March 9, 2012 (there are currently 152 publicly submitted Florida redistricting maps). For Massachusetts, the key restriction is that, according to the Massachusetts Constitution, “Every representative, for one year at least immediately preceding his election, shall have been an inhabitant of the district for which he is chosen” – since the next election for State Representatives is Tuesday, November 6, 2012, the map for the State House must be released before November 6, 2011 in order to allow for the “one year residency” restriction. I believe the sense of that restriction carries over to both the State Senate and Congressional District elections/maps. Glad to see that we are on track!

Well, we already know that Florida is nowhere near close to getting their “official” Congressional District map. What about some of the other states listed above? Note: any state, regardless of losing or gaining Congressional seats, can redraw Congressional District Boundaries (except the following states which only have one Congressional Representative: Alaska, Delaware, Montana, North Dakota, South Dakota, Vermont, and Wyoming).

The new map for Texas Congressional Districts (PLANC185) is currently available as a shapefile (Texas PLANC185). The creation data within the shapefile is June 15, 2011, although it looks like the actual legislative dates are:
June 22, 2011 signed in the State Senate
June 24, 2011 signed in the State House, and sent to the Governor
July 18, 2011 signed by the Governor

Although Arizona’s Independent Redistricting Commission has approved a Draft Congressional District map (Arizona Redistricting Map), the Commission only recently (Saturday, November 5, 2011) completed the 2nd round of public hearings. They still need to issue the “Final Map”. Note: after issuing their “Final Map”, all states must engage in the Department of Justice preclearance process before the maps are truly finalized. They seem to have no “official timetable”.

Saturday, November 5, 2011

11/5/2011 The Display of Information – Print and Online

There were news articles this week about the increase of poverty in the United States (The New York Times article link). The primary source is the report released by the Brookings Institute (“The Re-Emergence of Concentrated Poverty: Metropolitan Trends in the 2000s”, released Thursday, November 3, 2011). I am interested in these types of reports/articles for three reasons:
- as an American, I want to be aware of the ever-changing economic and demographic profiles of our society,
- as a person, I want to know how our society will deal with its problems, including poverty, and
- as someone who deals with data, I want to understand data, and understand and implement the most effective methods to transform that data into information.

The New York Times article

The online New York Times article consists only of text, plus four hyperlinks (“the report”, “The Truly Disadvantaged”, “Center on Poverty, Inequality and Public Policy”, and “earned income tax credit”) [the print version, of course, does not have even the hyperlinks]. Although there are mentions of geographic places (“Midwestern: Toledo and Youngstown in Ohio, and Detroit”, and “Sun Belt areas like Cape Coral, Fla., and Fresno, Calif.”), there are no accompanying maps or visualizations.

Just as it is important to discuss regional differences (Europe, Japan, North America, etc), and not just the entire “global economy”, an important part of the story of Concentrated Poverty in America is regional differences within the United States.

I went to the Brookings Institute website to both view their presentation of the Report, and to download the Report itself. Brookings link

The Brookings Institute Report

Before discussing the website, let’s look at the pdf Report. The 35-page report includes bar charts, number tables, a nationwide map, detail maps for Detroit, Dallas, Chicago, and Atlanta, appendices with the raw data, and two pages of endnotes.

The nationwide map shows colored dots (“circles that are all the same size”) over a map of the states. The dots are colored into four ranges, depending upon their “Concentrated Poverty Rate, 2005-09”. The colors are a good intuitive blue-color ramp (darker = more), but I would like to see the “count of areas” next to each colored dot in the legend:

15%< [24 areas] means there are 24 metropolitan areas with Concentrated Poverty Rates Greater Than 15%

As for the “circles that are all the same size”, welcome to “Displaying Data 101” – sometimes there is too much data to display on a single (static) map. In this case, because of the great number of metro areas in the Northeast, you can only see them all if they are relatively-small symbols (circles, squares, diamonds). Using different symbols for size variables could be used, but it might just make the map too busy.

Examining the map makes me wonder about applying Spatial Statistics to the data – examining each set of colored dots just gives me a “random” feeling, but that can certainly be validated through examining the Spatial Distribution of the data (and each data set). Another blog for another time.

The Brookings Institute online article

The Brookings online article is longer than the New York Times article (606 words versus 488 words), and includes the (linked) ability to “find concentrated poverty statistics for your metropolitan area”, plus a video. There are also two very nice maps of the 100 largest “metropolitan areas” in the United States. Although the maps are static (there is no slider to show changing data-over-time), the white non-outline of the light gray states provides the perfect background for the multivariate data.

Map 1 shows
- latitude/longitude location,
- size of circles equals current counts of people living in extreme-poverty census tracts, and
- color of circles indicates increase/decrease change from Census2000 data.

Additionally, the webpage takes advantage of the user interface, and offers the user a “hover over” ability – when you hover over a metro area, a detail popup appears with statistics relating to that metro area:


Map 2 shows
- latitude/longitude location,
- size of circles equals current level of Concentrated Poverty Rate, and
- color of circles indicates increase/decrease change from Census2000 data.

My next blog will discuss taking the data from the pdf Report into an Excel spreadsheet, then into Tableau Software for visualization over time.