Landsat at your fingertips

The USGS has put together a slick GUI that let's you browse (as if it were GoogleEarth) and download Landsat data. The interface is described here.

One of my students found this and showed it to me. (Over the next few weeks, hopefully I'll be able to post much of the material and discoveries from my new course "Spatial Data and Analysis".)

Hurricane-induced migration [Plot of the Week]

Impulse:

Hurricane Katrina, as pictured in the Gulf of Mexico at 21:45 UTC on August 28, 2005.

Response:

This map illustrates the national scope of the dispersion of refugees from Hurricane Katrina. It shows the location by zip code of the 800,000 displaced Louisiana residents who requested federal emergency assistance. The evacuees ended up dispersed across the entire nation, illustrating the wide-ranging impacts that can flow from extreme weather events, some of which are projected to increase in frequency and/or intensity as climate continues to change. (Source: Louisiana Geographic Information Center 2005)

1-800-CLOUD-GONE

Ever been sitting by a window in the space station and feel annoyed that clouds are obstructing your view? Charlie Lyod and Chris Herwig of Mapbox (covered before on FE) have a simple but clever solution: sort your data by pixel.

Their explanation is clearer than mine (pun intended). I just wanted to post the pretty pictures.

Before:

After:

I think this idea has several applications beyond clearing the skies.

h/t Young

Unleash your inner cartographer

In my work, I make a lot of maps. But they're usually just a single image and they're mediocre in terms of color-choice and design.  Mapbox is a product from a DC startup that can help us data-jockeys build svelte and scalable maps that integrate data from lots of sources. From their about page:

MapBox is a platform for designing and publishing fast and beautiful maps. We provide MapBox Streets, a complete customizable world base map, develop the powerful open source map design studio TileMill, make it easy to integrate maps into applications and websites, and support all of these tools on top of scalable, high-performance hosting. We've made MapBox developer friendly with an open API.

The development team has worked on all sorts of projects, from tracking elections to helping document hurricane damage. Their blog is also way cool.

h/t Young

Help map agriculture in Africa

Lyndon Estes writes

I have been working with Kelly Caylor developing a crowdsourced crop mapping project that is ready for some user testing.  I have asked my network of friends to test it a bit, in the hopes that we can see how the system performs, and get some initial data to show at my talk on this project next week at AGU.  

His instructions:

Hello Friends, I am kindly requesting your help with a research project. Our goal is to use crowdsourcing + google satellite imagery to map crop fields in Africa. We have just developed our prototype, which connects users on Amazon's Mechanical Turk Service to our field mapping interface.  

So, if any of you had a few minutes to spare over the next few days and an Amazon account (it's very easy to register as a Mechanical Turk user if you have an account, and not much harder to get an Amazon account if you don't have one), I would be very grateful if you could join up and map a few fields. The link below is to our website, which describes the registration and mapping process further.  

This project will be a for-pay endeavor within a few weeks, but for this stage when we are still working out bugs, we are going through Mechanical Turk's testing site (workersandbox.mturk.com), which pays fake money. Once you have registered, please go to sandbox to look for our HITs (Human Intelligence Tasks).  

Your help and feedback will be greatly appreciated, both for development purposes and for providing some data that I can include in my presentation on this project (Tuesday in San Fran).  

Thanks, Lyndon 

http://mappingafrica.princeton.edu/

How much groundwater does Africa have?

Quantitative maps of groundwater resources in Africa

A M MacDonald, H C Bonsor, B É Ó Dochartaigh and R G Taylor

Abstract: In Africa, groundwater is the major source of drinking water and its use for irrigation is forecast to increase substantially to combat growing food insecurity. Despite this, there is little quantitative information on groundwater resources in Africa, and groundwater storage is consequently omitted from assessments of freshwater availability. Here we present the first quantitative continent-wide maps of aquifer storage and potential borehole yields in Africa based on an extensive review of available maps, publications and data. We estimate total groundwater storage in Africa to be 0.66 million km3 (0.36–1.75 million km3). Not all of this groundwater storage is available for abstraction, but the estimated volume is more than 100 times estimates of annual renewable freshwater resources on Africa. Groundwater resources are unevenly distributed: the largest groundwater volumes are found in the large sedimentary aquifers in the North African countries Libya, Algeria, Egypt and Sudan. Nevertheless, for many African countries appropriately sited and constructed boreholes can support handpump abstraction (yields of 0.1–0.3 l s−1), and contain sufficient storage to sustain abstraction through inter-annual variations in recharge. The maps show further that the potential for higher yielding boreholes ( > 5 l s−1) is much more limited. Therefore, strategies for increasing irrigation or supplying water to rapidly urbanizing cities that are predicated on the widespread drilling of high yielding boreholes are likely to be unsuccessful. As groundwater is the largest and most widely distributed store of freshwater in Africa, the quantitative maps are intended to lead to more realistic assessments of water security and water stress, and to promote a more quantitative approach to mapping of groundwater resources at national and regional level.

Click to enlarge. Copyright ERL

See related field experiment on valuing ground water protection here.

h/t Kyle

Trophic level as a measure of economic development?

I will return this this new Nature article in another post, since it brings up a lot of interesting points, but I think this ancillary point was worth highlighting (even though the authors don't make this case): Looking at where humans sit in the food chain might be a useful (or at least interesting) measure of economic development.  In rich countries, land is used to feed animals that we eat.  In poor countries, people eat a higher fraction of crops themselves.  We've known about this phenomenon for a while now, but this map from the paper is striking because of its fully global perspective:

Here we show the fraction of the world’s total cropland that is dedicated to growing food crops (crops that are directly consumed by people) versus all other crop uses, including animal feed, fibre, bioenergy crops and other products. Averaged across the globe, 62% of total crop production (on a mass basis) is allocated to human food, 35% for animal feed (which produces human food indirectly, and less efficiently, as meat and dairy products) and 3% for bioenergy crops, seed, and other industrial products. There are striking disparities between regions that primarily grow crops for human consumption (such as Africa, South Asia, East Asia), and those that mainly produce crops for other uses (such as North America, Europe, Australia).
Copyright: Nature

AidData

Our colleague Johannes points us to AidData.org:

AidData is a collaborative initiative to provide products and services that promote the dissemination, analysis, and understanding of development finance information. At the core of the AidData program is the AidData web portal, which is a gateway to nearly 1 million records of development finance activities from donors around the world. Complementing the work of the OECD, whose Creditor Reporting System (CRS) is the official source of statistics for all OECD member countries, the AidData portal aims to provide access to development finance activities from a wide range of donors in an accessible format. In addition to providing access to these data, the team works on other projects that make it easier to access and analyze aid information, such as geocoding.

These are the guys behind the World Bank's Mapping for Results platform, previously covered here. Their blog ("First Tranche") is particularly worthwhile.

Stata's new geocode function

Our friend and colleague Reed Walker points out that there's a new Stata program called geocode that allows an internet-connected copy of Stata to query Google Maps for latitude and longitude. From Adam Ozimek at the excellent econ and general social science blog Modeled Behavior :

In the upcoming Stata Journal I have a paper with a coauthor that lets Stata query Google Maps in order to find latitude and longitude for addresses or other locations, also known as geocoding. What makes this useful is that you can have weird formatting, spelling errors, or missing information in your address or location variable and the program can still geocode it as well as Google Maps can place it on a map.

There's another program that calculates the "Google maps distance" (i.e., actual travel time distance as opposed to as-the-crow-flies geometric distance) between any two addresses. More details can be found here. This is very exciting.

Maps + Data Presentation

Since Sol recently posted on both NOAA's lovely data visualization labs and an assortment of map resources, I figured it might be worth pointing our readers to Cartastrophe. From the site's about page:

There are a lot of bad maps out there. They lurk in brochures, on company websites, and in magazines. They confuse, they miscommunicate, and they make it hard to learn anything about the world. Sometimes they leave off Sicily. They’re made by people who have to rush against tight deadlines, by people who are pressured by their bosses or clients to make bad design choices because it “looks cool,” and by people who were thrust into map-making jobs without any training.

[...]

We learn a lot from seeing what went wrong in someone else’s experience. I hope to amuse, but also to educate — to help people (myself included) understand what the elements of a good map are. And maybe, just maybe, if people are better able to understand what makes up a bad map, they’ll start demanding better ones.

The site is effectively the cartographic counterpart to Andrew Gelman's consistently superb data presentation critiques, and offers a huge amount of useful and thoughtful advice. Highlights:

• Texas grows 70% each year
• A village floating off the coast
• The vanishing kingdom

The last link is particularly nice for its reminder that as much as many of us love Google maps (I have difficulty imagining how I went through life before it...) it can still contain major errors.

Map resources

I spoke with some librarians today who pointed me to two excellent online resources for historical maps, the University of Texas Map Library and the David Rumsey Map Collection.  There is so much untapped information on these sites its overwhelming.  Below were two favorites that I found on Rumsey's site.

The Distribution of Wealth, 1870

Ranking of States by Income, Debt, Literacy, etc, 1880