Or at least those we know about. And where are the known meteorite landing places on earth? These impact zones show where scientists have found meteorites, or the impact craters of meteorites, some dating back as far as the year 2,300BC. The data is from the Meteorological Society and doesn't show those places where meteorites may have fallen but not been discovered

• Map made with CartoDB. See how they did it

• Download the data behind this map

Click image above to view the map.

It’s been six months since my most popular post How the Public Actually Uses Local Government Web Maps: Metrics from Denver; I’ve been gratified by the feedback.  But despite laying out detailed metrics showing that single-topic maps garner 3x the traffic of traditional portals, that user-friendly text search is critical to the map experience, and that users don’t spend time fiddling with default viewer parameters, I’ve found two particular reactions troubling:

• No one has widely circulated their own web metrics that shine a more positive light on map portals.

• People still keep building public-facing map portals. And writing press releasesabout them.

            The road to mediocre web experiences is paved with good intentions

For the next few posts, then,  we’ll lay out some major drawbacks of standard web portals as well as suggest a few alternatives along the way.  While the baseline scenario I have in mind are public-facing government mapping portals, those rolling corporate intranet solutions would do well to take heed.

Problem #1: Map Portals Over-Focus on the Map, Under-focus on Text-Based Search and Discovery

The dominant finding from Denver’s metrics is that the public approaches maps to retrieve particular bits of information, and then leave.  And how does everyone do Search & Discovery?  Think of your favorite search engine: you start typing into an auto-suggest box, then you get a text list of possibilities.  Both Google Maps and Bing Maps use a similar visual grammar–search text-box across the top, listings on the left, map on the right.

            Users are familiar with a map that contextualizes text-based search

GIS people all-too-easily lose sight how unfamiliar map navigation is for the general user, especially when confronted with an “immersive” experience.  The text box, then, featured prominently and with auto-complete, is a life-raft of familiarity.  Contrast that to this typical setup that we find in the Lancaster County GIS Property Searchsite:

          Ye shall be known by your Property Account number 

The search box is on the left side of the map, with Account search separate from Address search.  OK, what’s an Account number? Address is familiar, but why is House Number separate from Street Name?  Because, as we can all guess, that’s how it’s stored in the database.  And none of this is auto-complete, so my address-lookup is really an address-guess.  Being a clever sort, I enter “100″ “Main” and get 15 possibilities popping up in a Results window on the other side of the map. And only two fields are clearly visible, the Property Account and the Address (which I already entered).  Now this business with the Account number being featured so prominently is starting to make me wonder if this site is really meant for the public or the clerks in the Assessor’s Office who work with Account numbers all day long.  Finally, instead of plotting the properties in the map as obvious, clickable placemarks, they’re rendered as polygons, which, at the zoom level necessary to encompass all 15 possibilities, makes them barely visible.  Which I figured out when I went to remove a smudge from my monitor.

We’re not pointing fingers at Lancaster County here, because these are the same choices GIS people inflict on the public every day: unfamiliar interface layouts, text fields that don’t auto-complete, and results windows that pop up in other parts of the screen.

Takeaway #1: Users Want and Need a Large, Obvious, Auto-Complete Text Box to Drive Search & Discovery

Coming Soon: All Those Layers are Getting in the User’s Way

Unemployment is down again in the UK, with the number of jobless reaching a 15-month low. The unemployment rate for June to August 2012 was 7.9 per cent of the economically active population, down 0.2 from March to May 2012. There were 2.53 million unemployed people, down 50,000 from March to May 2012. 

The map shows the rate of youth between ages 16-24, claiming unemployment benefits. Birmingham Ladywood remains the constituency with the highest claimant rate at 11.4 per cent with Hodge Hill following closely behind at 10.1 per cent.

 
Travelex has created a Google Map to help film lovers and avid travelers find where in the world the best and most prominent blockbusters in history were filmed.

The Travelex Movie Lover's Guide to the World app allows users to explore the filming locations of the top 250 movies of all time (according to IMDB). The map features red markers pinned in locations all over the world.

Users can click a location marker to find out what was filmed there or they can click one of the movies listed to see all of the various locations where it was filmed.

Now that the election has been definitively declared for President Barack Obama (although the electoral votes for Florida are still being decided), it's time to look at post-election maps and analysis.  The presidential election was a close race this year with Obama leading with 60,794,790 popular votes (50.4%) to Romney's 57,901,194 votes (48%).  As the outstanding absentee and provisional ballots are counted and verified, those numbers will change.  The electoral breakdown curently stands at 303 for Obama and 206 for Romney with Florida's 29 electoral votes not yet allocated due to the close race in that state.

Maps that analyze elections results at the county level provide better insight into where the most intense support for each candidate lies.  Most states have a "winner takes all" when it comes to electoral votes where the candidate with the highest popular vote gets all the electoral votes.  With over 3,000 counties in the United States (known as parishes in Louisiana and boroughs in Alaska), mapping at this level produces a better picture of the geopolitical landscape of the U.S.

Max Rush at the Chicago Sun Times broke down the margins by county for the 48 lower states (Alaska and Hawaii are not included).  His map took a look at the margin of votes for each candidate (that is, how many more votes the leading candidate had than his opponent within each county).  The map helps depict the geopolitical landscape of the United States and shows where support for President Obama was the strongest.  The map uses absolute numbers, so of course, areas with higher populations will stand out more than rural populations.  Looking at the election margins map, it becomes immediately obvious that Obama has the most support in urban areas as compared to the more rural counties.  Major cities along the West Coast, Chicago, and the metropolitan areas of the East coast showed the highest support for Obama.  Maricopa County was the county with the largest margin of victory for Mitt Romney.  Not surprisingly, Utah seems to be the only state where all counties went to Romney. Wyoming follows as a close second in widespread support for Romney with only Teton County leaning Democratic.

Election map showing margin of votes by county for the 2012 Presidential Election. Map by Max Rust for the Chicago Sun Times. Click on map for larger image or download a PDF of the map.

The NY Times also produced a top down map that looked at the size of the leads.  Click on the "Size of Lead" map on the NY Times Presidential Elections map page to access the map.  This map used graduated symbols to denote the size of the lead.  Data for Hawaii is included but data for Alaska is missing.  Hover over each circle for a popup for the breakdown in number of votes and percentage for the candidates.  Click on a state to zoom to that level.

Size of lead post-election map for the 2012 Presidential Election from the NY Times.

The map from the NY Times that I found particularly interesting is the Shift from 2008 map (also available from the same page).  The map symbolizes the shifting trends between Democratic and Republican voters since the previous presidential election.  The map symbolizes with arrows (red for Republican and blue for Democratic) the percentage shift in votes between 2008 and 2012.  The rotation (blue arrows point left and red arrows point right) and length of each arrow symbolizes the proportional shift in votes since 2008. For example, St. Bernard Parish in Louisiana experienced a 10% gain in Democratic votes (and a 10% loss in Republican votes since 2008) and is symbolized by a long blue arrow.

The shift map is particularly interesting in that it shows shifts counter to the overall win for each state.  For example, almost the entire West Coast shows a shift towards an increase in Republican votes even though Washington, Oregon, and California all went for Obama.  Conversely, the states of Mississippi, Alabama, and Georgia show a pattern of increasingly Democratic votes despite those three states voting for Romney.

For a dynamic shift map, see the NY Times page on "How Obama Won Re-election" which has a 2008 and 2012 dynamic shift map which evokes the popular Wind Map by Fernanda Viégas and Martin Wattenberg.  Toggle back and forth between 2008 and 2012 to see the shifting tide from Democratic to Republican.

Change in share of votes between the Democratic and Republican candidates from 2008 to 2012. Source: NY Times.