1.1.1. KML Examples

Information

Google Earth is a popular geobrowser that provides an exploratory interface to a rich series of spatial data sets. Some of these may be considered 'formal' - as they are collected centrally and distributed in a top-down manner. Examples include the high resolution air photography and the terrain, road and government layers. Others are less formally produced in a 'bottom up' manner as individuals across the World contribute information to what may be considered the 'Geographic Web' through sites such as Wikipedia and Panoramio . Both the kinds of data and content of individual data sets provided through Google Earth change with some frequency.

We can access these data and compare their distributions and content with our own data through Google Earth . This is achieved through the Keyhole Markup Language. KML is an XML language through which the shapes, locations and visual characteristics of phenomena can be encoded.

If we use KML to describe our geographic information we can load our spatial data sets in Google Earth . We can then visually compare them with other formally and informally generated data sets originating from a range of different sources. Doing so is the essence of a 'mashup'. As more data are made available, more sophisticated mashups are possible. Those that take advantage of the data available through Google Earth and the interactive features of the interface may prove to be particularly useful for visualization.

Example

Let's load some KML files that contain geographic data of different types from different sources. We will be looking at two data sets and comparing them. The first is area-based and consists of resident population estimates from the US Census Bureau (U.S. Census Bureau 2008). The second is point-based, consisting of the locations of georeferenced photographs uploaded to Flickr (Yahoo! Inc. 2008).

The census data are aggregated to state level. Population figures are classified into 9 classes with equal intervals and shaded using a sequential scheme (Brewer 2002), (Harrower et al. 2003).

Population estimates from the Census Bureau (areas) and georeferenced photo locations from Flickr (points) combined in Google Earth.

Flickr is an online photo management and sharing application. The data we are using here consist of a sample comprising the 1,000 'most interesting' photographs according to the Flickr criteria (Yahoo! Inc. 2008) taken in 2008 in the continental US with point georeferences. Here the time at which a photo was taken is symbolised through a 12 hour clock face - with darker symbols for times between 18:00 and 05:59 and lighter yellow symbols for times between 06:00 and 17:59. The data set has been selected as it is rich in spatio-temporal information and typical of the kind of point-based data source that is often used in the geosciences.

We might consider this example a simple geovisualization mashup through which we can view data of different types from different sources. We might use Google Earth to visually explore the data - filtering by space, time and theme. And we might use Google Earth to compare the data we have loaded with the data available through Google Earth as 'layers' - for example, the road networks, air photography and 'Geographic Web'.

Exercise

Let's consider the following questions by visually exploring these data :

Can you suggest geographic relationships / trends in the individual data sets - do these vary geographically?
Can you identify errors or anomalies in the individual data sets - do these vary geographically?
Can you suggest geographic relationships / trends between the data sets - do these vary geographically?
Does viewing the data in this way help you generate ideas?

Download >> Google Earth.
Make sure you have installed the software.
Start Google Earth : the geobrowser should display the Earth.
In the 'Layers' window switch off all layers in the 'Primary Database'
Open the 'Geographic Web' to see the 'Panoramio' and 'Wikipedia' entries.
Download >> http://www.gicentre.org/tutorials/kmlGE/exercise1.kml.
This KML file contains ...
- point data : Flickr - 1000 most interesting photographs
- area data : state level population estimates 2006 (choropleth map)
Experiment with the Google Earth navigation controls and use the TimeLine to select temporal slices of the point data.
Double click in Google Earth or use the 'Search' feature to find an area or location that you know.
Zoom in and out to see whether you can identify geographic patterns or relationships - do these vary as you change the scale at which you are considering the data? Do they vary over time?
Use the list of 'Places' to switch various 'layers' of data on and off. These are organised in a hierarchy in Google Earth which be expanded and contracted by toggling the + / - icons
Click on the point symbols to view details about the images, including thumbnails, titles and interestingness - ordered from 1 to 1000 and emboldened.