USING GIS TO TEACH NATURAL DISASTERS (AND VICE VERSA) TO NON-GEOLOGY STUDENTS

Geographic Information Systems provide a great opportunity to learn Earth science concepts in an interactive and engaging way. Students learn skills that allow them to analyze primary data and to use the scientific method, while exploring basic relationships between Earth systems. At Drew University, NJ, this is particularly important because there are a limited number of geology courses and no geoscience major or minor. Data relevant to natural disasters are particularly engaging to non-geology students, because they are closely tied to current events and because basic interpretations are straightforward. More in-depth analyses can be added on, as needed. One exercise that has proven particularly engaging focuses on addressing the question of whether the U.S. has areas with high occurrence of tornadoes (i.e., a “Tornado Alley”). Students download state boundaries and historic tornado locations, including an attribute for the number of fatalities from each storm, from the National Atlas website. They use a spatial join to determine the number of storms and number of tornado-related fatalities in each state. The concept of normalizing data is introduced to calculate the tornado occurrence per state area (determined by calculating the states’ geometries), thereby eliminating any large-state bias in the spatial join. A second issue addressed through normalization is a large-population bias: states with higher populations will tend to have more fatalities from storms. Students download population tables from the US Census website, joining by attribute to the states layer. They can then calculate the tornado-related fatalities per 1000 residents for each state. The result is that the Midwest has had a high occurrence of tornadoes and tornado-related fatalities, but some states are emphasized or de-emphasized by normalization (e.g., North Dakota has had a moderate number of fatalities overall, but a high number of fatalities per capita). In Spring 2011, students extended these concepts in independent projects to look at risk to nuclear facilities from tornadoes and earthquakes, employing spatial joins, joining by attributes, and buffering.