Though a large fraction of primary production and organic matter cycling in the oceans occurs on continental shelves dominated by sandy deposits, the microbial communities associated with permeable shelf sediments remain poorly characterized. Therefore, the primary objective of this study was to provide the first detailed characterization of microbial diversity in representative marine sands of the South Atlantic Bight (SAB) and the northeastern Gulf of Mexico (NEGOM) through analyses of SSU rRNA gene (Bacteria), nosZ (denitrifying bacteria), and amoA (ammonia-oxidizing bacteria) sequences. Communities were analyzed by DNA extraction, clone library construction, and terminal restriction fragment length polymorphism (T-RFLP) community fingerprinting. Sediment characteristics, geochemical parameters, and rate measurements were obtained in parallel with microbial community analysis. Microbiological and biogeochemical approaches were coupled, allowing the structure-function relationships of key microbial groups involved in carbon and nitrogen cycling in continental shelf sediments to be examined. In the SAB study (Ch. 1), clone libraries were constructed from both sediment core material and manipulated sediment within column experiments. Rapid organic matter degradation and coupled nitrification-denitrification were observed in column experiments at flow rates and oxygen concentrations resembling in situ conditions. Numerous SSU rRNA gene phylotypes were affiliated with the phyla Proteobacteria (classes Alpha-, Delta-, and Gammaproteobacteria), Planctomycetes, Cyanobacteria, Chloroflexi and Bacteroidetes. Detectable sequence diversity of nosZ and SSU rRNA genes increased in stratified redox-stabilized columns compared to in situ sediments, with the Alphaproteobacteria comprising the most frequently detected group. Alternatively, nitrifier communities showed a relatively low and stable diversity that did not co-vary with the other gene targets. In the NEGOM study (Ch. 2), high throughput techniques were developed and applied to extensively profile overall and denitrifying microbial communities in a large number of sediment samples over various sediment depth intervals, contrasting sites, and sampling periods. Cloning/sequencing and community fingerprinting (T-RFLP) approaches were applied in parallel to characterize microbial diversity and phylogenetic composition. Statistical estimators including species richness, Shannon-Weiner and 1/D indices, nucleotide diversity, gene diversity, evenness, and theta (pi) indicated little difference between four clone libraries constructed from selected depth intervals (0-2 cm, 18-20 cm) at each site in March. In contrast, T-RFLP profiles and robust phylogenetic analysis showed distinct trends in diversity according to site, depth, and time period sampled. The results elucidate predominant phylotypes that are likely to catalyze carbon and nitrogen cycling in marine sands. Several microbial groups (Deltaproteobacteria, Gammaproteobacteria, Planctomycetes) were confirmed as significant contributors to the microbial communities of permeable marine sediments in agreement with previous work. However, the robust sequence database of this study expanded current knowledge to reveal a large overall community diversity including additional groups (Alphaproteobacteria, Bacteriodetes/Chlorobi, and Cyanobacteria) that had not been previously recognized using cultivation-independent methods with inherently lower resolution. The Alphaproteobacteria, in particular, were shown to be relatively abundant in the overall and denitrifying communities at both SAB and NEGOM sites. Although overall diversity increased in response to redox stabilization and stratification in column experiments, the major phylotypes remained the same, indicating that the columns sufficiently mimic in situ conditions. While SSU rRNA gene phylotypes detected by clonal analysis were similar at the phylum level at all sites, the NEGOM site showed much higher species richness in comparison to SAB. At NEGOM, T-RFLP showed distinct differences in community diversity according to site, depth, and time. The sequence database from this thesis will facilitate the development of improved probes and primer sets to be used in quantifying the metabolically active members of permeable sand communities. Rapid community fingerprinting methods developed here should allow for more extensive comparisons across environmental gradients in order to better understand the factors controlling microbial diversity in permeable sediments. / A Thesis submitted to the Department of Oceanography in partial fulfillment of the requirements for the degree of Master of Science. / Spring Semester, 2006. / March 20, 2006. / SSU rRNA, Microbial Community Analysis, Nitrogen Cycle, nosZ, amoA, Sandy Sediments / Includes bibliographical references. / Joel Kostka, Professor Directing Thesis; Heath Mills, Committee Member; Markus Huettel, Committee Member; Lee Kerkhof, Committee Member.

The community abundance and diversity of mRNA transcripts for sulfate-reducing prokaryotes were investigated in parallel with rate measurements of sulfate reduction activity in marine subsurface sediments of Aarhus Bay, Denmark. Solid phase and porewater chemistry and sulfate reduction rates were determined on subsamples from sediment cores to 5 m below the sediment surface. Molecular analysis of the dissimilatory (bi) sulfite reductase (dsrAB) mRNA transcripts and 16S rRNA were performed by quantitative reverse transcription real-time PCR and traditional clonal analysis. The distribution of dsrA transcripts was directly linked to both sulfate reduction rates and rRNA content. Quantitative analysis of dsrA gene transcripts indicated the presence of active sulfate reduction at 465 cm below the sediment-water interface, where high methane concentrations persist in regions of near sulfate depletion. These results suggest an abundance of active bacteria in zones of high sulfate reduction and a marked decrease in zones of low sulfate reduction rates. Archaeal communities showed a constant distribution in the sediment profile in zones of intermediate and low sulfate reduction rates, indicating their predominance in sediments where low sulfate and high methane concentrations were observed. Substantiated by biogeochemical and rRNA analysis, the analysis of mRNA gene transcripts serves as a versatile molecular proxy for the study of sulfate reducing communities in marine subsurface sediments. / A Thesis submitted to the Department of Oceanography in partial fulfillment of the requirements for the degree of Master of Science. / Spring Semester, 2009. / September 5, 2008. / PCR, Microbial Ecology, Sufate Reduction / Includes bibliographical references. / Joel E. Kostka, Professor Directing Thesis; Markus Huettel, Committee Member; Kuki Chin, Committee Member.

This study addresses the decomposition of dissolved organic carbon (DOC) in highly permeable coastal sand sediments. DOC fluxes from shelf sediments (~180 Tg C yr-1) are significant, roughly equal to the DOC flux from rivers (~200 Tg C yr-1) and to the rate of carbon burial in marine sediments (~160 Tg C yr-1) (Burdige et al., 1999). DOC thus plays an important role in the global cycles of carbon and nitrogen (Hedges, 1992) and understanding the processes that control DOC dynamics is critical to clarify this role. The small concentrations of lignin in the ocean and the 13C-enriched composition of marine DOC suggest relatively rapid degradation of terrestrial DOC, however, the mechanisms by which the large volumes of DOC released to coastal waters are rapidly degraded following discharge are poorly understood (Hedges, 1992). Photooxidation can decompose DOC to low molecular weight substances (Kieber et al., 1989; Mopper et al., 1992), and this process may account for the removal of 20 to 30% of DOC in coastal waters (Mopper and Kieber, 2002). Microbial activity in the water column and shelf sediments degrades DOC, but the importance of these processes are not well constrained. Approximately 70% of the shelf sediments are relict sands (Riggs et al., 1996), and in the shallow coastal zone a large fraction of these sands are highly permeable and permit circulation of water through the interstitial space (Marinelli et al., 1998; Huettel et al., 1996; Huettel and Gust, 1992). In this shallow environment, strong boundary currents caused by wind, waves, and tides force bottom water loaded with DOC through the sediment ripples and upper surface layers of the sand. Abundance and diversity of microbes in permeable sediments exceed that of the overlying water column (Hunter et al., 2006), and the question arises whether the filtration through the sediment affects the decomposition of DOC and water column DOC dynamics. I tested the working hypotheses that rapid DOC transport along relatively short pathways through the sand significantly enhances the degradation of the DOC and that the sedimentary flushing tightly links sedimentary and water column DOC concentrations. The main objectives of the research were: 1) To measure degradation rates of DOC in percolated permeable sediment and to compare the rates to those in the water column; 2) To assess the magnitude and variability of DOC concentrations in water and pore water at two nearshore environments of the northeastern Gulf of Mexico with permeable sediments, and to determine the links between the sedimentary and water column DOC dynamics; 3) To investigate the processes that control the DOC distribution in the surface layer of the permeable bed. I combined time series and chamber experiments conducted in-situ with laboratory column reactor and flume experiments. While the laboratory experiments were designed to assess DOC degradation rates and processes controlling DOC distribution under well-defined conditions, the field experiments were conducted to gain insight on DOC fluxes and seasonal trends at two coastal sites differing in their DOC input and hydrodynamic characteristics. The flume experiment examined the distributions of DOC resulting from degradation of phytoplankton deposited on permeable sediments exposed to unidirectional flow. The results show that DOC from algal cell degradation in the sediment surface layer is rapidly transported deeper into the sediment by advective pore water flows and concentrated in specific zones dictated by the pore water flow field. The DOC profiles showed highest concentrations in the upper 2 cm of the sediment, lower concentrations below that layer and in some cases increasing DOC concentrations in the layers below the flushing zone, resulting in a concave profile shape. The laboratory column reactor experiments demonstrate that fresh DOC originating from phytoplankton, as well as older DOC from terrestrial sources, is rapidly degraded (2.15 to 124.04 µmol l-1 h-1) while passing short distances through permeable sands by the microbial community and that degradation rates in the sediment exceeded those in the water column, approximately 7-fold for my experimental settings. The measurements of water column and pore water DOC concentrations at St. George Island produced the first DOC time series for two shallow Gulf of Mexico coastal environments showing DOC ranges and temporal dynamics. These time series indicate that during the winter season, when hydrodynamic forcing is strongest, water column and sedimentary DOC concentrations are coupled, while no such link could be recorded during the calmer periods of the year. The similarity between the in-situ DOC profiles and those measured in the laboratory flume suggest that plankton deposition, combined with advective pore water transport processes and slow upward diffusion of refractory DOC from deeper layers, shape the concave DOC concentration profiles. The field measurements with advection chambers reveal seasonal variation of the magnitude and direction of DOC fluxes and showed that the permeable coastal sands can act as a sink (winter) or a source (summer) for DOC. / A Thesis submitted to the Department of Oceanography in partial fulfillment of the requirements for the degree of Master of Science. / Summer Semester, 2008. / April 22, 2008. / Sediments, Biogeochemistry, Doc, Dissolved Organic Carbon, Dom, Degradation, Permeable / Includes bibliographical references. / Markus Huettel, Professor Directing Thesis; David Thistle, Committee Member; Thorsten Dittmar, Committee Member.

Sea Surface Temperatures (SSTs) are an important measure of our current weather and climate, as well as an essential variable in both short and long term weather forecasting. Infrared SST retrievals are reliant on passive sensors, and retrieval techniques are influenced by changes in atmospheric composition, including aerosols. Many empirically derived retrieval algorithms are based on matching Top of Atmosphere (TOA) Brightness Temperatures (BTs) from the Advanced Very High Resolution Radiometer (AVHRR) to buoy measurements during clear-sky conditions. Data is cloud-cleared to remove cloud-contaminated data. However, small, but influential, Aerosol Optical Depths (AODs) data may not be flagged as contaminated and the algorithms incorrectly calculate a cold SST due to the radiometer sensing the cooler, elevated aerosol layer temperature. Many studies on aerosol effects on SSTs focus on aerosols due to volcanic eruptions. However, truly operational tropospheric aerosol corrections for daytime and nighttime retrievals have yet to be implemented. This work constitutes a first step to creating an accurate aerosol correction by exploring the sensitivity of aerosols on SSTs. The Santa Barbra DISORT Radiative Transfer model is used to quantify the effects of aerosol contamination on retrieved TOA BTs. The calculated radiances are spectrally averaged over each channel, converted to BTs, and used to calculate an SST using the Naval Oceanographic Office AVHRR algorithms. A radiative transfer model is used to evaluate the SST retrieval error due to varying AOD, height of an aerosol layer, and the satellite zenith angle (or viewing angle). This analysis shows that errors greater than the stated retrieval uncertainty of 0.5 K are observed for AODs greater than 0.25. Two sites with state-of-the-art aerosol measurements are analyzed for AOD variability. The first site, at Anmyon in east Asia, is found to have 14% of the days during the springtime with an AOD greater than 0.25. Based on the AERONET data from a second site in Cape Verde, 65% of the days during the boreal summer are found to have AOD greater than 0.25. Unfortunately, this seasonal peak in dust activity coincides with the active tropical cylogensis season for the region, making accurate SSTs even more vital for prediction purposes. / A Thesis submitted to the Department of Earth, Ocean and Atmospheric Science in partial fulfillment of the requirements for the degree of Master of Science. / Fall Semester, 2010. / October 19, 2010. / Dust, Retrieval Error, Aerosols, Sea Surface Temperature / Includes bibliographical references. / Carol Anne Clayson, Professor Directing Thesis; Mark A. Bourassa, Committee Member; Guosheng Liu, Committee Member; Douglas L. Westphal, Committee Member.

By sampling individual rain events over a 1-year period at three sites situated around a known point source of atmospheric mercury, we have attempted to quantify its influence on local mercury deposition. A suite of trace elements was also analyzed on these rain event samples. A multi-element analytical program was set up using a Thermo-Finnigan "Element" ICP-MS. We identified 46 elements that are significantly enriched in rain samples relative to the method blank, including the alkali metals and alkaline earth elements, all three rows of the transition metals, and the rare earth elements. The total mercury concentrations in the rainwater samples ranged from 2-40 ng/L. The volume weighted mean rainfall mercury concentrations ranged from 9.2-9.8 ng/L, and there were no significant differences in the rainfall Hg deposition between the three sites. Principal component factor analysis (PCFA) was used to evaluate co-variance between mercury and trace element deposition. PCFA showed a strong crustal factor, a strong sea-salt factor, a mysterious "P" factor and a strong mercury factor. The mercury factor linked mercury with Bi, Ga, Pb, Sb and V. Plume dispersion modeling and air-mass back trajectory analysis have been conducted for each rain event. The goal of this research was to use mercury and trace element relationships in an effort to identify, and quantify, the impacts from various emission sources in the region on rainfall chemistry. We calculated the annual integrated percent of mercury associated with coal burning ranging between 15-47%. Using models to estimate the impact of local deposition we found that less than 1% of the mercury we measured is from the known point source. / A Thesis submitted to the Department of Oceanography in partial fulfillment of the requirements for the degree of Master of Science. / Fall Semester, 2006. / July 14, 2006. / Pensacola, Trace Metal, Florida, Hg, Coal Fired Power Plant / Includes bibliographical references. / William M. Landing, Professor Directing Thesis; Philip Froelich, Committee Member; Markus Huettel, Committee Member.

This dissertation is a collection of five calcareous nannofossil and one stable isotope studies on materials from Ocean Drilling Program (ODP) Legs 183 (Site 1135), 206 (Site 1256), and 207 (Site 1259) that target two important paleoceanographic events: 1) the middle/late Miocene carbonate crash, and 2) the Paleocene/Eocene Thermal Maximum (PETM). Site 1256 nannofossil biostratigraphy in Chapter 1 refined the author's shore-based shipboard Quaternary-middle-Miocene nannofossil biostratigraphy with 16 zones/combined zones recognized based on 28 nannofossil datums. This chapter provides a chronologic framework for the age calibration of the first occurrence (7.18 Ma) and last occurrence (6.32 Ma) of Reticulofenestra rotaria, calculation of linear sedimentation rates, age determination of basalt basement (~14.5 Ma), and the recognition of the "carbonate crash" paleoceanographic event at the middle/late Miocene boundary. Reworked nannofossils and lithologic changes also allow a reading of a three-episode redepositional history (4.7, 8.3, and 10.7 Ma, respectively) in the eastern Pacific. The detailed examination of the Site 1256 material also yielded well-preserved Discoaster stellulus, for which only the distal view had been depicted in the original description. In Chapter 2, a redescription and re-illustration of both sides of this asterolith is provided. This should prevent misidentification of specimens in proximal view, thereby raising its potential application for middle-late Miocene biostratigraphy. Based on the above age model, in Chapter 3 stable oxygen and carbon isotopes were used for the first time to explore the late/middle Miocene "carbonate crash". This carbonate transition is a widespread (eastern and central equatorial Pacific, Indian, South Atlantic, and the Caribbean), sharp decrease in carbonate mass-accumulation rates, which has previously been considered only a dissolution event. The positive correlation (R2 = 0.75) between d13C and CaCO3 mass accumulation rates during 5-14 Ma at ODP Site 1256 clearly demonstrates that carbonate accumulation is mainly biologically controlled. The coincidence of the carbonate crash with negative excursions in d13C and d18O values suggests a causative mechanism related to surface-water productivity, as a result of surface-water warming and reduced upwelling. Based on these observations, one could speculate that the major middle/late Miocene sea-level drop may have caused the complete closure of the Indonesian Seaway, resulting in a piling-up of surface warm water in the west Pacific. The eastward spread of this nutrient-poor water then would have warmed sea-surface temperatures and reduced upwelling in the central and eastern Pacific, thereby creating a prolonged "El Nino" scenario and reducing biological productivity of phytoplankton. The reduction in carbonate supply to the deep waters consequently caused a rapid shoaling of the carbonate compensation depth, thereby triggering the carbonate crash. The PETM was a catastrophic, rapid greenhouse-forced global warming event ~55 m.y. ago that triggered an abrupt turnover in ocean chemistry and circulation as well as biota. Chapter 4 represents a quantitative study of the response of nannoplankton to the PETM at Demerara Rise, equatorial Atlantic (Site 1259). Toweius, Fasciculithus, and Chiasmolithus sharply decrease at the onset of the PETM, whereas Chiasmolithus, Markalius cf. M. apertus, and Neochiasmolithus thrive immediately after the event, which also signals the successive first appearances of Discoaster araneus, Rhomboaster, and Tribrachiatus. Two main environmental factors were extracted by correspondence analysis of relative abundance data. The time series of the two factors shows that during the PETM, 1) environmental stress (most likely from changes in seawater pH) increased and may well have also induced the evolution of ephemeral nannofossil "excursion taxa"; and 2) surface-water productivity increased at this site presumably due to higher runoff from continental areas. The local phytoplankton opportunist, Markalius cf. M. apertus, is described as a new species in Chapter 5, which will be published under the name Coccolithus bownii. Results presented in Chapter 6 from Site 1135 on the Kerguelen Plateau, Southern Ocean suggest that nannoplankton responded differently to the PETM at southern high latitudes. The onset of the carbon isotope excursion occurs within an 18-cm interval (instead of 1-2 cm as observed in most deep-sea sections) before the peak is reached, displaying a linear mixing curve. This indicates that the release of light carbon was a gradual, single injection, instead of multiple pulses as suggested in previous work, and that this sequence is highly expanded as a result of high sedimentation rates at this relatively shallow oceanic site. This is evidenced by the high numbers of dissolution-susceptible holococcoliths (Zygrhablithus bijugatus) preserved throughout the sequence. Although r- and K-selected specialists exponentially increase in abundance at the onset, Chiasmolithus abruptly drops but then rapidly recovers, whereas Discoaster and Fasciculithus show opposite trends, indicating that in high latitudes, surface-water oligotrophy prevailed at the carbon isotope excursion (CIE) onset but mesotrophic conditions dominated the CIE recovery. These observations confirm previous results from ODP Site 690 on Maud Rise. The intensive dissolution of susceptible holococcoliths and the poor preservation of the assemblages are believed to have been caused by the effects of corrosion caused by the methane release. The different responses of nannoplankton to the PETM and the contrasts evident in previous work from the open ocean vs the continental margins further demonstrate that the response to the PETM can be influenced by local differences in geologic setting and oceanographic conditions. / A Dissertation submitted to the Department of Geological Sciences in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Fall Semester, 2007. / July 13, 2007. / Paleoceanography, Paleocene/Eocene Thermal Maximum, Middle/Late Miocene Carbonate Crash, Productivity, Stable Isotopes, Calcareous Nannofossils / Includes bibliographical references. / Sherwood W. Wise, Jr., Professor Directing Dissertation; Richard L. Iverson, Outside Committee Member; Anthony J. Arnold, Committee Member; Joseph F. Donoghue, Committee Member; Yang Wang, Committee Member.

This research involves the creation of a new storm surge index that incorporates many variables important in storm surge generation like maximum winds, radius of maximum winds, pressure, translation speed, and bathymetry. Using a two-dimensional, barotropic ocean model, power laws have been developed that describe the relationship between storm surge and changes in maximum wind, radius of maximum winds, pressure, and bathymetry. Direct curve fitting is used to describe the relationship between storm surge and changes in translation speed since a power-law relationship does not exist in that case. A database of 39 landfalling, United States hurricanes between 1986 and 2007 is used to evaluate the quality of the index. Storm parameters for all database storms are compiled using the extended best track dataset, and an index value is calculated for each storm in the database. Correlation analysis is then performed using the index values and observed maximum storm surge heights. Finally, an extensive error analysis is presented to demonstrate uncertainties in the index in both forecast and post-analysis situations. / A Dissertation submitted to the Department of Meteorology in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Summer Semester, 2008. / April 23, 2008. / Hurricane, Flood, Storm Surge, Forecasting / Includes bibliographical references. / Carol Anne Clayson, Professor Directing Dissertation; Jennifer Georgen, Outside Committee Member; Robert Hart, Committee Member; Paul Ruscher, Committee Member; Peter Ray, Committee Member; Steven Cocke, Committee Member.

The Florida manatee (Trichechus manatus latirostris) is an obligate aquatic mammal that inhabits shallow coastal waterways. Previous research has demonstrated that manatees vocalize using frequencies that range from < 0.5 kHz to 12 kHz (Frisch and Frisch, 2003; Nowacek et al., 2003) and hear frequencies between 0.4 kHz and 45 kHz (Bullock et al., 1982; Gerstein et al., 1999). Although the auditory anatomy has been described previously (Fischer, 1988; Ketten et al., 1992; Chapla and Rommel, 2003), potential sound pathways and middle ear mechanisms have not been analyzed. In the current study, Computerized Tomography (CT) is utilized to generate three-dimensional reconstructions of manatee heads and isolated earbone (tympanoperiotic) complexes in order to visualize the in situ arrangement of soft tissue and bone. Density data attached to the CT data are used to make calculations regarding the transmission and reflection of sound waves, with varying angles of incidence, as they encounter the boundaries of different tissue layers. Sound waves with 90° angles of incidence (relative to the tissue/water interface) will transmit 94% to 99% of their total energy intensity to the skull. Sound waves approaching at oblique angles to the surface of the head will transmit less of their energy intensity. The fraction of energy intensity in the transmitted wave will decrease with a decrease in the angle of incidence. Additional calculations of critical angles and an isospeed channel suggest that the inner fatty tissue layer (bounded by muscle and bone) does not provide an efficient channel for sound waves. Other potential sound pathways are discussed. A model of inertial bone conduction predicts that the peak displacement amplitude of the stapes, relative to that of the head, will be 8.4 dB at 200 Hz, and that inertial bone conduction may be possible between 125 Hz and 1000 Hz. Comparisons are made with data from humans and golden moles (Mason, 2003). / A Thesis submitted to the Department of Oceanography in partial fulfillment of the requirements for the degree of Master of Science. / Spring Semester, 2006. / April 7, 2006. / Middle Ear Anatomy, Three-Dimensional Reconstruction, Manatee Hearing, Inertial Bone Conduction / Includes bibliographical references. / Douglas P. Nowacek, Professor Directing Thesis; Sentiel A. Rommel, Outside Committee Member; Markus Huettel, Committee Member; Louis St. Laurent, Committee Member.

The purpose of this study is to test a perfect prognosis scheme used to develop lightning guidance for the warm season (May-September) for three regions of the country (Colorado, Washington, D.C., and Oregon). This guidance product is developed on a 10 × 10 km grid at 3-hourly intervals. Four seasons of RUC20 analyses and NLDN data are used to develop BLR equations for predicting one or more flashes (PROB ≥ 1), as well as predicting the amount of lightning (PROB ≥ T) for each 3-h period. RUC analyses of differing geopotential heights for each region and a map typing procedure to develop lightning frequencies for five dominant flow regimes are implemented to capture small-scale enhancements due to local forcing that are not well resolved by NWP models. Binary logistic regression (BLR) is used to develop equations for forecasting one or more flashes, while a negative binomial (NB) statistical model is used to predict the amount of lightning, conditional on one or more flashes occurring. The map-type frequencies are used as candidate predictors for both the BLR and NB models. RUC-analyzed parameters describing moisture, temperature, wind and stability also are used as candidate predictors in the PP equations. The perfect prognosis scheme is evaluated using independent data from mesoscale models during the 2009 warm season. Output from the National Centers for Environmental Prediction (NCEP) 13-km RUC (RUC13) and the NCEP 12-km North American Mesoscale Model is used to evaluate the performance of the new lightning guidance products. The goal is to beat lightning forecasts that are attained from climatology and persistence alone for each region. For most forecast periods, there is good agreement between the model forecasts and observed lightning verification. Brier score analysis indicates that our models beat climatology in forecasting one or more flashes, as well as the amount of lightning. Reliability also indicates that our models perform well for forecasting one or more flashes, but not for forecasting the amount of lightning. The goal is to make the technique a fully operational guidance product that can be used for NWS offices in these three regions. / A Thesis submitted to the Department of Earth Ocean and Atmospheric Science in partial fulfillment of the requirements for the degree of Master of Science. / Summer Semester, 2010. / May 18, 2010. / Statistical Studies, Lightning Forecasts / Includes bibliographical references. / Henry E. Fuelberg, Professor Directing Thesis; Philip Sura, Committee Member; Guosheng Liu, Committee Member; Andrew I. Watson, Committee Member.

There are no IAEA-GNIP stations in Florida. In order to improve our understanding of the proxy climate records preserved in speleothems, tree rings and lake sediments and to explore the utility of stable oxygen and hydrogen isotopes in the investigation of the water cycle in North Florida, there was a need to establish rainwater sampling stations to fill a gap in the oxygen and hydrogen records of rainfall in North Florida. Rain samples collected from Tallahassee and Pensacola in North Florida were analyzed for stable oxygen and hydrogen isotopic ratios. The Tallahassee samples span a time period from May 2006 to January 2011 and the Pensacola samples were collected from November 2004 to January 2010. Water samples were also collected in late 2010 and early 2011 from two coastal lakes (Eastern Lake and Western Lake in Walton County) and two ponds (Pond 1 and Bluesink in Apalachicola National Forest) in North Florida for stable isotope analyses. The δ18O and δD values of rain samples all plot either on or close to the Global Meteoric Water Line (GMWL) on a δ18O vs δD plot while most of the lake/pond water samples plot below the GMWL indicating that the lake waters in those area have been significantly affected by evaporation. Stable isotopic ratios of the Tallahassee samples ranged from -13.6 to 0.4‰ for δ18O and -99.4 to -2.2‰ for δD values (n=43) while those of the Pensacola samples ranged from -16.0 to 1.5‰ for δ18O and -115.8 to 118.9‰ for δD (n=260). Lake waters yielded δ18O values ranging from -4.7 to 1.9‰ and δD from -31.8 to 14.2‰. Pensacola samples have weighted monthly mean δ18O and δD values that ranged from -7.3 to -1.5‰ and from -4.3 to -18.7‰, respectively. The annual weighted mean δ18O and δD values for Tallahassee for study period are -4.5 and -23.3‰ respectively and -4.3 and -18.7‰ for Pensacola. The average d-excess value is 14‰ for Tallahassee and 9‰ for Pensacola. There is a negative correlation between the precipitation amount and the stable isotopic ratios of preciptation during the study period in North Florida. However, the "amount effect" can only explain less than 20% of the variance in the δ18O values of precipitation likely due to variations in the source and history of atmosphric moisture. There is a weak (but not significant) positive correlation between the average air temperature and the stable isotopic compositions of precipitation, suggesting that temperature is not a major factor in controlling the stable isotopic composition of precipitaion in this area. In Talahassee, rain samples for winter months had δ18O values ranging from -5.3 to -2.6‰, δD values from -30 to -6.2‰, and d-excess from 9.7 to 18.8. During the summer season, rain in Tallahassee displayed a larger range of isotopic variation, with δ18O and δD values ranging from -13.6 to 0.4 and from -99.4 to -7.7‰, respectively, and d-excess ranging from -11.2 to 14.7. The larger isotopic variability observed in the summer is primarily due to tropical storms. Rain samples from Pensacola had slightly more enriched monthly mean δ18O values than Tallahassee samples probably because Pensacola is located closer to the main moisture source ( the Gulf of Mexico). Tropical storms (such as Alberto and Fay) not only produced large amounts of precipitation but also are characterized by very negative δ18O and δD values. The much lower than normal δ18O and δD values of tropical storms are likely recorded in tree rings and speleothems, and thus identification of storm signals in tree rings or speleothems would allow reconstruction of paleostorm history in the area. The unique isotopic signatures of tropical storms can also serve as a tracer to study the source and flow path of groundwater. / A Thesis submitted to the Department of Earth, Ocean and Atmospheric Sciences in partial fulfillment of the requirements for the degree of Master of Science. / Spring Semester, 2011. / April 1, 2011. / Isotope, Precipitation, Temperature, Amount Effect, Continental Effect / Includes bibliographical references. / Yang Wang, Professor Directing Thesis; James Tull, Committee Member; Lynn Dudley, Committee Member; Yuch-Ping Hsieh, Committee Member.