About the Author Introduction Methods and Procedures Analysis and Results Discussion Conclusion Bibliography
Methods and Procedures
Data layers were acquired from several sources. Some data was downloaded for free from the internet, while other data was available in class. Three data layers were from a project conducted by the Center For Hazards & Risk Research (CHRR) in 2005: Global Cyclone Mortality Risks and Distribution, Global Cyclone Proportional Economic Loss Risk Deciles, and Global Cyclone Hazard Frequency and Distribution.
These three data layers are raster files which show the global coverage of human mortality, economic loss as a proportion of Gross Domestic Product (GDP) per analytical unit, and hurricane frequency based on hurricane events in Florida, between 1980 and 2000. Each of these data layers is divided into deciles, 10 classes consisting of approximately equal numbers of grid cells. Therefore, class 1 denotes low mortality risk, proportional economic loss risk, or frequency risk of hurricanes relative to surrounding cells, while 10 denotes high mortality, proportional economic loss, or frequency of hurricanes relative to surrounding cells.
Another data layer, Tropical Storm Intensity Zone Global Coverage, was derived from the Munich Reinsurance Company's World Map of Natural Hazards. This shape file shows zones based on the five different wind speeds of the Saffir-Simpson Hurricane Scale. Also, Land Surface Elevation 24 is a vector data layer depicting the elevation of the entire state of Florida as contour lines, measured in feet. Although parts of the elevation data are more than 20 years old (1980s), this is not necessarily a problem, since the data layers on mortality, proportional economic loss and frequency span an equally long time range. These five data layers were obtained from Geodata.gov.
A base map of the United States of America was obtained in class. I edited the layer, removing Alaska, Hawaii, and other offshore U.S. territories. By selecting by attribute, I created a new layer from the edited base map which was the state boundary of Florida. A data layer depicting the boundaries of urban areas in the whole of the United States, including Alaska and Hawaii for the year 2000 was obtained from the U.S. Census Bureau website. I then selected by attribute and selected all urban areas from the state of Florida, and saved the selection as a new layer. A data layer of Florida’s coastline was obtained from the Florida Coastal Everglades Long Term Research Institute website.
I converted the three CHRR layers on mortality risk, proportional economic loss risk and frequency risk from raster files into shape files, and then performed intersects between each of the three and the Florida state data layer. Thus, I obtained hurricane mortality risk, proportional economic loss risk and frequency risk just for the state of Florida.
Hurricane Hazard Frequency and Distribution Deciles in Florida (1980 to 2000)
Mortality Risk Distribution in Florida as a Result of Hurricanes (1980 to 2000)
Proportional Economic Loss Risk Deciles in Florida as a Result of Hurricanes (1980-2000)
For the tropical storm intensity zone data layer, I created individual data layers for each tropical storm in the vicinity of Florida. Then, I performed intersects between each storm zone layer and the proportional economic loss risk layer for Florida. Therefore, I was able to isolate the proportional economic risks for the state which fall under the influence of different storm zone levels.
Tropical Storm Intensity Zones along the Southern Extent of the Eastern Seaboard
I performed intersects between the Florida mortality risk, proportional economic loss risk and frequency risk data layers and the Florida urban areas layer I had already created. As a result, three data layers were created which showed the relative levels of risk for the three variables in only urban areas. Similarly, I performed erases between the urban Florida layer and the three state-wide risk layers in order to produce layers of mortality risk, proportional economic loss risk, and frequency risk for non urban areas in Florida.
Mortality Distribution Deciles in Urban Areas of Florida Compared to Non-Urban Areas
Hazard Frequency of Hurricanes in Urban Areas of Florida Compared to Non-Urban Areas
Proportional Economic Loss Risk Deciles in Urban Areas of Florida Compared to Non-Urban Areas
The vector file on elevation for the whole state of Florida was classified according to natural breaks among five class intervals of elevation in feet. Next, an intersect was performed between it and the layer of urban areas in Florida. This created a layer of urban Florida elevation. While analyzing this data, an outlier was identified, an elevation of over 700 feet, when all other records where under 300 feet. This outlier was removed for the purpose of generating the natural break class ranges, so that the classes would not be overly skewed.
Visual Comparison of All Urban Elevation and Proportional Economic Loss Risk from Hurricanes
Visual Comparison of All Urban Elevation and Hazard Frequency of Hurricanes
Visual Comparison of All Urban Elevation and Mortality Risk Distribution from Hurricanes
I then performed spatial joins between the urban mortality risk layer and the ‘contour’ attribute in the urban Florida elevation layer. The product of this operation is that I was able to calculate the average elevation for each polygon in the urban mortality risk layer. I undertook a similar procedure involving the urban proportional economic loss risk layer as well.
A 5 km buffer around the Florida coastline vector file was created. With this buffer, I could select urban areas within 2.5 km of the coast of Florida, or within 2.5 km of either side of a river or canal, for the purposes of singling them out from urban areas further inland. However, insufficient time was available to incormporate this aspect into my analysis.