Access to master’s theses is restricted to current ERAU students, faculty, and staff.

Date of Award

2014

Document Type

Thesis - Open Access

Degree Name

Master of Science in Aeronautics

Department

Graduate Studies

Committee Chair

John M. Lanicci, Ph.D.

Committee Member

Christopher G. Herbster, Ph.D.

Committee Member

Jennifer E. Thropp, Ph.D.

Abstract

The use of observational datasets to determine the occurrence frequencies of extreme weather events has gained a lot of recent interest due to concerns about the potential regional impacts from global climate change. Extreme-value theory can quantify the return frequency of the most extreme events, using climatologically short data sets and the assumption that such short climatological periods are stationary. However, the resulting analyses must be used with caution since they may not accurately reflect the potential of extreme events in the future due to climate change and variability. Accurately predicting extreme-event likelihood is important for building realistic long-range planning scenarios for a number of weather- and climate-sensitive interests.

This study used extreme-value theory to analyze a short period (15-year), high-density rainfall dataset from NASA Kennedy Space Center’s observational network. This data was acquired through the Tropical Rainfall Measurement Mission archive website. The researcher employed the National Center for Atmospheric Research’s Extremes statistical software package for the analysis of 24-hour rainfall at the locations of the 32 tipping-bucket gauges in the network. This type of analysis is highly sensitive to data that may have been misreported, invalid, or missing, therefore, additional quality control was required. The quality-controlled rainfall gauge data was subsequently gridded using a Barnes-style objective analysis with minimal smoothing, in order to estimate missing values while preserving maxima in the initial data. The high-resolution gridded rainfall data was used by the Extremes program to estimate a series of event return levels over the studied region.

Analyses of the gridded data show that that the 100-year events are around 315 mm and 433 mm for 24-hour and 72-hour durations, respectively. The wet-season analysis 100-year event estimation was around 426 mm and is similar to the yearly analysis, indicating that the majority of the annual extremes are from wet-season events. The yearly and wet-season 100-year return levels appear to be realistic and consistent with previous literature and estimates from the longer period of record at Titusville; however, some results from the dry-season analysis do not appear to be realistic, as they indicate the rainfall frequency distribution has an abnormally bounded tail shape. The dry-season 100-year return-levels are likely greater than the 170 mm model consensus produced from the analysis of the gridded data. The better-behaved Titusville analysis suggests the dry-season 100 year return level is around 250 mm. Findings indicate large uncertainty associated with long-period estimates and high spatial variations in rainfall extremes across the Kennedy Space Center region.

Included in

Climate Commons

Share

COinS