Bourassa, M. A. (2006). Satellite-based observations of surface turbulent stress during severe weather. Atmosphere-Ocean Interactions , 2 , 35–52.
Zavala-Hidalgo, J., Morey, S. L., O'Brien, J. J., & Zamudio, L. (2006). On the Loop Current eddy shedding variability. Atmosfera , 19 (1), 41–48.
Hite, M. M. (2006). Vorticity-Based Detection of Tropical Cyclogenesis . Master's thesis, Florida State University, Tallahassee, FL.
Abstract: Ocean wind vectors from the SeaWinds scatterometer on QuikSCAT and GOES imagery are used to develop an objective technique that can detect and monitor tropical disturbances associated with the early stages of tropical cyclogenesis in the Atlantic basin. The technique is based on identification of surface vorticity and wind speed signatures that exceed certain threshold magnitudes, with vorticity averaged over an appropriate spatial scale. The threshold values applied herein are determined from the precursors of 15 tropical cyclones during the 1999-2004 Atlantic hurricane seasons using research-quality QuikSCAT data. Tropical disturbances are found for these cases within a range of 19 hours to 101 hours before classification as tropical cyclones by the National Hurricane Center (NHC). The 15 cases are further subdivided based upon their origination source (i.e., easterly wave, upper-level cut-off low, stagnant frontal zone, etc). Primary focus centers on the cases associated with tropical waves, since these waves account for approximately 63% of all Atlantic tropical cyclones. The detection technique illustrates the ability to track these tropical disturbances from near the coast of Africa. Analysis of the pre-tropical cyclone (TC) tracks for these cases depict stages, related to wind speed and precipitation, in the evolution of an easterly wave to tropical cyclone.
Hughes, P. J. (2006). North Atlantic Decadal Variability of Ocean Surface Fluxes . Master's thesis, Florida State University, Tallahassee, FL.
Abstract: The spatial and temporal variability of the surface turbulent heat fluxes over the North Atlantic is examined using the new objectively produced FSU3 monthly mean 1°x1° gridded wind and surface flux product for 1978-2003. The FSU3 product is constructed from in situ ship and buoy observations via a variational technique. A cost function based on weighted constraints is minimized in the process of determining the surface fluxes. The analysis focuses on a low frequency (basin wide) mode of variability where the latent and sensible heat flux anomalies transition from mainly positive to negative values around 1998. It is hypothesized that the longer time scale variability is linked to changes in the large scale circulation patterns possibly associated with the Atlantic Multidecadal Oscillation (AMO; Schlesinger and Ramankutty 1994, Kerr 2000). The changes in the surface heat fluxes are forced by fluctuations in the mean wind speed. Zonal averages show a clear dissimilarity between the turbulent heat fluxes and wind speed for 1982-1997 and 1998-2003 over the region extending from the equator to roughly 40°N. Larger values are associated with the earlier time period, coinciding with a cool phase of the AMO. The separation between the two time periods is much less evident for the humidity and air/sea temperature differences. The largest differences in the latent heat fluxes, between the two time periods, occur over the tropical, Gulf Stream, and higher latitude regions of the North Atlantic, with magnitudes exceeding 15 Wm-2. The largest sensible heat flux differences are limited to areas along the New England coast and poleward of 40°N.
Briggs, K. (2006). ENSO Event Reproduction: A Comparison of an EOF vs. A Cyclostationary (CSEOF) Approach . Master's thesis, Florida State University, Tallahassee, FL.
Abstract: In past studies, El Niño-Southern Oscillation (ENSO) events have been linked to devastating weather extremes. Climate modeling of ENSO is often dependent on limited records of the pertinent physical variables, thus longer records of these variables is desirable. Noisy signals, such as monthly sea surface temperature, are good candidates for reproduction by several existing auto-regression techniques. Through auto-regression, influential principal component modes are broken down into a series of time points that are each dependent upon an optimal weighting of the surrounding points. Using these unique numerical relationships, a noisy signal can be reproduced by thus processing the leading modes and adding an artificial record of properly distributed noise. Statistical measures of important ENSO regions suggest that the nature of oceanic and atmospheric anomalous events is cyclic with respect to certain timescales; for example, the monthly timescale. To detect ENSO signals in the presence of a varying background noise field, the detection method should take into account the signal's strong phase-locking with this nested variation. Cyclostationary Emperical Orthogonal Functions (CSEOFs) are built upon the idea of nested cycles, unlike traditional EOFs, which incorporate a design that is better detailed for stationary processes. In this study, both EOF and CSEOF modes of a 50-year Pacific SST record are processed using an auto-regression technique, and several sets of artificial SST records are constructed. Appropriate statistical indices are applied to these artificial time series to ensure an acceptable consistency with the real record, and then artificial data is produced using the artificial time series. In all cases, the cyclostationary approach produces more realistic warm ENSO events with respect to timing, strength, and other traits than does the stationary approach. However, both methods produce only a fair representation of cold events, suggesting that further study is necessary for improvement of La Niña modeling. Shorter records of variables such as sea level height and Pacific wind stress anomalies can hinder the usefulness of auto-regression, owing to time point dependence on surrounding points. Using a regression technique to find an evolutionary consistency (i.e. physically consistent patterns) between one of these variables and a variable with a longer record (such as SST) can eliminate this problem. Once a regression relationship is found between two variables, the variable with the shorter record can be re-written to match the time evolution of the variable with the longer record. Here regression, both EOF and CSEOF, is performed on both sea surface temperature and sea level height (a 20-year record), and sea surface temperature and wind stress (a 39-year record). Once the regression relationships are found, artificial SST time series are incorporated in place of the original time series to produce several artificial 50-year SLH and wind stress data sets. 5 Pacific regions are chosen, and statistics and behavior of the artificial sets within these regions are compared to those of the original data. Once again the cyclostationary approach fares better than the stationary. In particular the EOF assumption of cross correlational symmetry fails to capture the direction-dependence of ENSO evolution, causing inconsistent ENSO behavior. This renders an EOF method insufficient for climate modeling and prediction, and implies that a better aim is to incorporate physical cyclic features via a cyclostationary method.
Culin, J. C. (2006). Wintertime ENSO Variability in Wind Direction Across the Southeast United States . Master's thesis, Florida State University, Tallahassee, FL.
Abstract: Changes in wind direction in association with the phases of the El Niño-Southern Oscillation (ENSO) are identified over the Southeast region of the United States during the winter season (December-February). Wind roses, which depict the percentage of time the wind comes from each direction and can graphically identify the prevailing wind, are computed according to a 12-point compass for 24 stations in the region. Unfolding the wind rose into a 12-bin histogram visually demonstrates the peak frequencies in wind direction during each of the three (warm, cold and neutral) phases of ENSO. Normalized values represent the number of occurrences (counts) per month per ENSO phase, and comparison using percent changes illustrates the differences between phases. Based on similarities in wind direction characteristics, regional topography and results from a formal statistical test, stations are grouped into five geographic regions, with a representative station used to describe conditions in that region. Locations in South Florida show significant differences in the frequencies in wind direction from easterly directions during the cold phase and northerly directions during the warm phase. North Florida stations display cold phase southerly directions, and westerly and northerly directions during the warm phase, both of which are significant for much of the winter. Coastal Atlantic stations reveal winds from westerly directions for both phases. The Piedmont region demonstrates large variability in wind direction due to the influence from the Appalachian Mountains, but generally identifies warm phase and cold phase winds with more zonal influences rather than just from south or north. The Mountainous region also indicates southerly cold phase winds and northerly warm phase winds, but also reveals less of an influence from ENSO or significantly different distributions. Comparisons between observed patterns and those obtained using the NCEP/NCAR Reanalysis data reveal how the model-derived observations resolve the ENSO influence on surface wind direction at selected locations. Overall, resolution of the strength of the signals is not achieved, though the depiction of the general pattern is fair at two of the three locations. Connections between the synoptic flow and surface wind direction are examined via relationships to the storm track associated with the 250 hPa jet stream and sea level pressure patterns during each extreme ENSO phase. Discussion of reasons the NCEP reanalysis illustrates surface wind direction patterns different from those derived from observations is included.
Petraitis, D. C. (2006). Long-Term ENSO-Related Winter Rainfall Predictions over the Southeast U.S. Using the FSU Global Spectral Model . Master's thesis, Florida State University, Tallahassee, FL.
Abstract: Rainfall patterns over the Southeast U.S. have been found to be connected to the El Niño-Southern Oscillation (ENSO). Warm ENSO events cause positive precipitation anomalies and cold ENSO events cause negative precipitation anomalies. With this level of connection, models can be used to test the predictability of ENSO events. Using the Florida State University Global Spectral Model (FSUGSM), model data over a 50-year period will be evaluated for its similarity to observations. The FSUGSM is a global spectral model with a T63 horizontal resolution (approximately 1.875°) and 17 unevenly spaced vertical levels. Details of this model can be found in Cocke and LaRow (2000). The experiment utilizes two runs using the Naval Research Laboratory (NRL) RAS convection scheme and two runs using the National Centers for Environmental Prediction (NCEP) SAS convection scheme to comprise the ensemble. The simulation was done for 50 years, from 1950 to 1999. Reynolds and Smith monthly mean sea surface temperatures (SSTs) from 1950-1999 provide the lower boundary condition. Atmospheric and land conditions from January 1, 1987 and January 1, 1995 were used as the initial starting conditions. The observational precipitation data being used as the basis for comparison is a gridded global dataset from Willmott and Matsuura (2005). Phase precipitation differences show higher precipitation amounts for El Niño than La Niña in all model runs. Temporal correlations between model runs and the observations show southern and eastern areas with the highest correlation values during an ENSO event. Skill scores validate the findings of the model/observation correlations, with southern and eastern areas showing scores close to zero. Temporal correlations between tropical Pacific SSTs and Southeast precipitation further confirm the model's ability to predict ENSO precipitation patterns over the Southeast U.S. The inconsistency in the SST/precipitation correlations between the models can be attributed to differences in the 200-mb jet stream and 500-mb height anomalies. Slight differences in position and strength for both variables affect the teleconnection between tropical Pacific SSTs and Southeast.
Taylor, J. P. (2006). Comparison of ECMWF and Quikscat-Derived Surface Pressure Gradients . Master's thesis, Florida State University, Tallahassee, FL.
Abstract: A technique based solely on QuikSCAT data is developed for determining suspect differences between QSCAT and ECMWF pressure gradients. Pressure fields are computed from scatterometer winds using a variational method that applies a gradient wind conversion. Kinematic analysis of the satellite wind field is performed in order to determine which parameters are physically related to the suspect pressure gradients. It is discovered that the likelihood of these suspect occurrences has the greatest dependence on relative vorticity, total deformation, and the curvature Rossby number. A broad range of these values is tested and a single assessment criterion is derived based upon the value of several skill scores. Overall, the assessment criterion is able to correctly identify the majority of suspect pressure gradients; yet considerable over-flagging does occur in many instances. However, the over-flagging is not random: the false alarms are tightly clustered around the suspect areas, resulting in flagged regions that are too large. Identification of the location of suspect areas in pressure products should be useful to forecasters.
Banks, R. (2006). Variability of Indian Ocean Surface Fluxes Using a New Objective Method . Master's thesis, Florida State University, Tallahassee, FL.
Abstract: A new objective technique is used to analyze monthly mean gridded fields of air and sea temperature, scalar and vector wind, specific humidity, sensible and latent heat flux, and wind stress over the Indian Ocean. A variational method produces a 1°x1° gridded product of surface turbulent fluxes and the variables needed to calculate these fluxes. The surface turbulent fluxes are forced to be physically consistent with the other variables. The variational method incorporates a state of the art flux model, which should reduce regional biases in heat and moisture fluxes. The time period is January 1982 to December 2003. The wind vectors are validated through comparison to monthly scatterometer winds. Empirical orthogonal function (EOF) analyses of the annual cycle emphasize significant modes of variability in the Indian Ocean. The dominant monsoon reversal and its connection with the southeast trades are linked in eigenmodes one and two of the surface fluxes. The third eigenmode of latent and sensible heat flux reveal a structure similar to the Indian Ocean Dipole (IOD) mode. The variability in surface fluxes associated with the monsoons and IOD are discussed. September-October-November composites of the surface fluxes during the 1997 positive IOD event and the 1983 negative IOD event are examined. The composites illustrate characteristics of fluxes during different IOD phases.
Yu, P. (2006). Development of New Techniques for Assimilating Satellite Altimetry Data into Ocean Models . Ph.D. thesis, Florida State University, Tallahassee, FL.
Abstract: State of the art fully three-dimensional ocean models are very computationally expensive and their adjoints are even more resource intensive. However, many features of interest are approximated by the first baroclinic mode over much of the ocean, especially in the lower and mid latitude regions. Based on this dynamical feature, a new type of data assimilation scheme to assimilate sea surface height (SSH) data, a reduced-space adjoint technique, is developed and implemented with a three-dimensional model using vertical normal mode decomposition. The technique is tested with the Navy Coastal Ocean Model (NCOM) configured to simulate the Gulf of Mexico. The assimilation procedure works by minimizing the cost function, which generalizes the misfit between the observations and their counterpart model variables. The “forward” model is integrated for the period during which the data are assimilated. Vertical normal mode decomposition retrieves the first baroclinic mode, and the data misfit between the model outputs and observations is calculated. Adjoint equations based on a one-active-layer reduced gravity model, which approximates the first baroclinic mode, are integrated backward in time to get the gradient of the cost function with respect to the control variables (velocity and SSH of the first baroclinic mode). The gradient is input to an optimization algorithm (the limited memory Broyden-Fletcher-Goldfarb-Shanno (BFGS) method is used for the cases presented here) to determine the new first baroclinic mode velocity and SSH fields, which are used to update the forward model variables at the initial time. Two main issues in the area of ocean data assimilation are addressed: 1. How can information provided only at the sea surface be transferred dynamically into deep layers? 2. How can information provided only locally, in limited oceanic regions, be horizontally transferred to ocean areas far away from the data-dense regions, but dynamically connected to it? The first problem is solved by the use of vertical normal mode decomposition, through which the vertical dependence of model variables is obtained. Analyses show that the first baroclinic mode SSH represents the full SSH field very closely in the model test domain, with a correlation of 93% in one of the experiments. One common way to solve the second issue is to lengthen the assimilation window in order to allow the dynamic model to propagate information to the data-sparse regions. However, this dramatically increases the computational cost, since many oceanic features move very slowly. An alternative solution to this is developed using a mapping method based on complex empirical orthogonal functions (EOF), which utilizes data from a much longer period than the assimilation cycle and deals with the information in space and time simultaneously. This method is applied to map satellite altimeter data from the ground track observation locations and times onto a regular spatial and temporal grid. Three different experiments are designed for testing the assimilation technique: two experiments assimilate SSH data produced from a model run to evaluate the method, and in the last experiment the technique is applied to TOPEX/Poseidon and Jason-1 altimeter data. The assimilation procedure converges in all experiments and reduces the error in the model fields. Since the adjoint, or “backward”, model is two-dimensional, the method is much more computationally efficient than if it were to use a fully three-dimensional backward model.
