About the Author Introduction Methods and Procedures Analysis and Results Discussion Conclusion Bibliography
Conclusion
The first hypothesis could only be partially proven with the current data. The mode of proportional economic loss risk was highest in storm zone 5, and decreased as one moved to zones 4 and 3, but then increased between zones 3 and 1. Thus, the area of Florida with the lowest mode of proportional economic loss risk was within storm zone 3, where the mode was gridcode 5, and indicating average proportional economic loss risk. The higher modes of proportional economic loss risk in the less intense storm zones may indicate that more people are living there because it recieves less violent storms imapcts. However, this may result in some home owners taking fewer precautions in fortifying their properties in the north, since they know storm winds there are less intense. This sort of attitude could be one way of trying to account for why economic loss risk could be higher in the north rather than in mid-Florida, since residents there are more experienced and better prepared. Many other issues may exist surrounding this topic, and further research is recommended. At presant, my current hypothesis fails to fully corrispond to the data, and may need to be revised.
The second hypothesis stipulated that both mortality risk and proportional economic loss risk would be greater in urban areas compared to non-urban areas. While such a relation was observed in relation of mortality risk, a similar relation was not observed with the proportional economic loss risk. As already mentioned, additional variables such as land uses, may help to explain why the mode for proportional economic loss risks is higher in non-urban areas. Some problems may involve the definition of urban and non-urban. For instance, the urban data obtained from the U.S. Census Bureau classified ‘urban’ as comprising both urban areas and urban clusters, although no mention is made regarding suburban areas, or whether they are included as part of urban areas or clusters. I recommend that this hypothesis be modified to only consider mortality risk and exclude proportional economic loss risk from consideration. Clearly, the portion of the current hypothesis regarding proportional economic loss risk is false, and therefore a new, separate hypothesis is needed to investigate this variable separately.
The third hypothesis called for higher human mortality and property damage in urban areas at lower elevations than higher elevations. The reasoning for this was that these areas would be affected both by strong winds and storm surges. The results of the histograms did not seem to reflect this. Instead, urban areas below a certain height were found to experience the full range of mortality risk. A similar pattern is observed regarding proportional economic loss risk. Box plots did show a concentration of 100% of areas being below 160 feet for the highest gridcodes (8, 9, 10), compared to low and moderate risks areas which where found to extend to higher elevations. Despite these results, the general trend which the third hypothesis calls for is not broadly reflected, and thus the validity of this hypothesis is at best inconclusive.
Although my three hypotheses were logical, and commonsensical, people and the decisions they make on where to invest in infrastructure and construction, are not always logical. Confounding variables may exist, and their discovery and exploration may help to account for some of the surprising and unexpected results, based on the hypotheses I proposed. Also, I believe that my hypotheses should not yet be ruled out completely due to errors which may have resulted in certain gridcodes being under-represented during the analyses of risk. Also, I recommend that future studies include absolute levels of mortality and proportional economic loss risk, rather than working with relative risk, since working with these abstractions can be difficult and confusing at times.