Knowing where people are on earth is essential to understanding interaction between humans and their environment. The knowledge is especially meaningful when faced with assessing vulnerability to natural or man-made disasters such as earthquakes, hurricanes, war, or famine. In this exercise we'll perform a spatial query using the Population Density 2007 dataset based on LandScan data from Oak Ridge National Laboratory to estimate how many people were within 35km of the epicenter of the earthquake that hit L'Aquila, Italy on April 6, 2009. L'Aquila is both the capital city of the Abruzzo region and of the province of L'Aquila. It is a bustling university town whose daytime population swells from 73,000 to 100,000 with students, workers, and tourists. Hence, any rigorous population analysis would probably include a range of numbers.

First, we'll open the Population Density map in TerraViva! Global Data Analyst. Then, using the spatial query tool we'll select the Pop Density map as the analysis theme and build a point-based query within a 35km buffer distance of the epicenter. (The spatial query area could have been along a line or within a polygon if our analysis required it.)

Once the query parameters have been set we move our cursor to the map and click on the location of the earthquake epicenter - 42.334/13.334. A mask appears on the map to indicate the query area. The shaded circle in the image at right shows the selected query area, a 35km area around the epicenter. Before reviewing the query results (data table below) we can simply point to any pixel to get thematic information about the pixel. The yellow label shows that the pixel at 42.3533/13.4049 (where the arrow points to the little white dot), within 7km of the epicenter, has a population density of 10,911 people per square kilometer. LandScan pixels are 30 arc-seconds, commonly referred to as 1 km wide at the equator. The high density pixel, however, is not at the equator. The area of this pixel is closer to 0.63 sq km (about 700 meters in width and 900 meters in height). So, the population at this pixel is closer to 6,900 (10,911 x 0.63).

The query results displayed in the data table to the right indicate that 219, 028 people may populate the selected query area. The LandScan Pop Density dataset from ORNL estimates “ambient” population density - where people are likely to be at noon. LandScan uses population reports from official publications, then disaggregates the population according to an algorithm that is based mostly on land use, topography and transportation information. The net effect is that LandScan tends to put people in cities and along roads. For more information about this dataset see TerraViva! GeoServer and TerraViva! Data Hound blog.

*TECH NOTE: The spatial query tool, available in Global Data Analyst, allows the user to choose between the fastest or most accurate results. The quick query tool, available in Global Data Viewer defaults to most accurate results.