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.
Engelman, M. B. (2008).
A Validation of the FSU/COAPS Climate Model. Master's thesis, Florida State University, Tallahassee, FL.
Abstract: This study examines the predictability of the Florida State University/Center for Oceanic and Atmospheric Prediction Studies (FSU/COAPS) climate model, and is motivated by the model's potential use in crop modeling. The study also compares real-time ensemble runs (created using persisted SST anomalies) to hindcast ensemble runs (created using weekly updated SST) to asses the effect of SST anomalies on forecast error. Wintertime (DJF, 2 month lead time) surface temperature and precipitation forecasts over the southeastern United States (Georgia, Alabama, and Florida) are evaluated because of the documented links between tropical Pacific SST anomalies and climate in the southeastern United States during the winter season. The global spectral model (GSM) runs at a T63 resolution and then is dynamically downscaled to a 20 x 20 km grid over the southeastern United States using the FSU regional spectral model (RSM). Seasonal, monthly, and daily events from the October 2004 and 2005 model runs are assessed. Seasonal (DJF) plots of real-time forecasts indicate the model is capable of predicting wintertime maximum and minimum temperatures over the southeastern United States. The October 2004 and 2005 real-time model runs both produce temperature forecasts with anomaly errors below 3°C, correlations close to one, and standard deviations similar to observations. Real-time precipitation forecasts are inconsistent. Error in the percent of normal precipitation vary from greater than 100% in the 2004/2005 forecasts to less than 35% error in the 2005/2006 forecasts. Comparing hindcast runs to real-time runs reveals some skill is lost in precipitation forecasts when using a method of SST anomaly persistence if the SST anomalies in the equatorial Pacific change early in the forecast period, as they did for the October 2004 model runs. Further analysis involving monthly and daily model data as well as Brier scores (BS), relative operating characteristics (ROC), and equitable threat scores (ETS), are also examined to confirm these results.
Putnam, W. M. (2007).
Development of the Finite-Volume Dynamical Core on the Cubed-Sphere. Ph.D. thesis, Florida State University, Tallahassee, FL.
Abstract: The finite-volume dynamical core has been developed for quasi-uniform cubed-sphere grids within a flexible modeling framework for direct implementation as a modular component within the global modeling efforts at NASA, GFDL-NOAA, NCAR, DOE and other interested institutions. The shallow water equations serve as a dynamical framework for testing the implementation and the variety of quasi-orthogonal cubed-sphere grids ranging from conformal mappings to those numerically generated via elliptic solvers. The cubed-sphere finite-volume dynamical core has been parallelized with a 2-dimensional X-Y domain decomposition to achieve optimal scalability to 100,000s of processors on today's high-end computing platforms at horizontal resolutions of 0.25-degrees and finer. The cubed-sphere fvcore is designed to serve as a framework for hydrostatic and non-hydrostatic global simulations at climate (4- to 1-deg) and weather (25- to 5-km) resolutions, pushing the scale of global atmospheric modeling from the climate/synoptic scale to the meso- and cloud-resolving scale.
Stewart, M. L. (2007).
Cyclogenesis and Tropical Transition in Frontal Zones. Master's thesis, Florida State University, Tallahassee, FL.
Abstract: Tropical cyclones can form from many different precursors, including baroclinic systems. The process of an extratropical system evolving into a warm core tropical cyclone is defined by Davis and Bosart (2004) as a Tropical Transition (TT) with further classification of systems into Weak Extratropical Cylclones (WEC) and Strong Extratropical Cyclones (SEC). It is difficult to predict which systems will make the transition and which will not, but the description of a common type of TT occurring along a front will aid forecasters in identifying systems that might undergo TT. A wind speed and SST relationship thought to be necessary for this type of transition is discussed. QuikSCAT and other satellite data are used to locate TT cases forming along fronts and track their transformation into tropical systems. Frontal TT is identified as a subset of SEC TT and the evolution from a frontal wave to a tropical system is described in five stages. A frontal wave with stronger northerly wind and weaker southerly wind is the first stage in the frontal cyclogenesis. As the extratropical cyclogenesis continues in the next two stages, bent back warm front stage and instant occlusion stage, the warmer air of the bent back front becomes surrounded by cooler air . Next, in the subtropical stage the latent heat release energy from the ocean surface begins ascent and forms a shallow warm core. As the energy from surface heat fluxes translates to convection within the system, the warm core extends further into the upper levels of the atmosphere in the final, tropical stage of TT. Model data from MM5 simulations of three storms, Noel (2001), Peter (2003) and Gaston (2004) are analyzed to illustrate the five stages of frontal TT. Noel is found to have the most baroclinic origin of the three and Gaston the least.
Smith, R. A. (2007).
Trends in Maximum and Minimum Temperature Deciles in Select Regions of the United States. Master's thesis, Florida State University, Tallahassee, FL.
Abstract: Daily maximum and minimum temperature data from 758 COOP stations in nineteen states are used to create temperature decile maps. All stations used contain records from 1948 through 2004 and could not be missing more than 5 consecutive years of data. Missing data are replaced using a multiple linear regression technique from surrounding stations. For each station, the maximum and minimum temperatures are first sorted in ascending order for every two years (to reduce annual variability) and divided into ten equal parts (or deciles). The first decile represents the coldest temperatures, and the last decile contains the warmest temperatures. Patterns and trends in these deciles can be examined for the 57-year period. A linear least-squares regression method is used to calculate best-fit lines for each decile to determine the long-term trends at each station. Significant warming or cooling is determined using the Student's t-test, and bootstrapping the decile data will further examine the validity of significance. Two stations are closely examined. Apalachicola, Florida shows significant warming in its maximum deciles and significant cooling in its minimum deciles. The maximum deciles seem to be affected by some localized change. The minimum deciles are discontinuous, and the trends are a result of a minor station move. Columbus, Georgia has experienced significant warming in its minimum deciles, and this appears to be the result of an urban heat-island effect. The discontinuities seen in the Apalachicola case study illustrate the need for a quality control method. This method will eliminate stations from the regional analysis that experience large changes in the ten-year standard deviations within their time series. The regional analysis shows that most of the region is dominated by significant cooling in the maximum deciles and significant warming in the minimum deciles, with more variability in the lower deciles. Field significance testing is performed on subregions (based on USGS 2000 land cover data) and supports the findings from the regional analysis; it also isolates regions, such as the Florida peninsula and the Maryland/Delaware region, that appear to be affected by more local forcings.
Arrocha, G. (2006).
Variability of Intraseasonal Precipitation Extremes Associated with ENSO in Panama. Master's thesis, Florida State University, Tallahassee, FL.
Abstract: Extensive analysis has been conducted over past decades showing the impacts of El Niño-Southern Oscillation (ENSO) on various regions throughout the world. However, these studies have not analyzed data from many stations in Panama, or they have not analyzed long periods of observations. For these reasons, they often miss climatological differences within the region induced by topography, or they do not possess enough observations to adequately study its climatology. Accordingly, the current study focuses on ENSO impacts on precipitation specific to the Isthmus of Panama. Results will be useful for agricultural and water resources planning and Panama Canal operations. Monthly total precipitation data were provided by Empresa de Transmisión Eléctrica S.A., which includes 32 stations with records from 1960 to 2004. The year is split into three seasons: two wet seasons (Early and Late Wet), one dry season (Dry). The country is also divided into regions according to similarities in the stations' climatology and geographic locations. Upper and lower precipitation extremes are associated with one of the three ENSO phases (warm, cold or neutral) to estimate their percentages of occurrences. The differences between each ENSO phases' seasonal precipitation distributions are statistically examined. Statistical analyses show effects of ENSO phases that vary by season and geographical region. Cold and warm ENSO years affect the southwestern half of the country considerably during the Late Wet season. Cold ENSO phases tend to increase rainfall, and the warm phase tends to decrease it. The opposite is true for the Caribbean coast. The Dry season experiences drier conditions in warm ENSO years, and the Early Wet season does not show any statistically significant difference between ENSO years' rainfall distributions.
DiNapoli, S. (2010).
Determining the Error Characteristics of H*WIND. Master's thesis, Florida State University, Tallahassee, FL.
Abstract: The HRD Real-time Hurricane Wind Analysis System (H*Wind) is a software application used by NOAA's Hurricane Research Division to create a gridded tropical cyclone wind analysis based on a wide range of observations. One application of H*Wind fields is calibration of scatterometers for high wind speed environments. Unfortunately, the accuracy of the H*Wind product has not been studied extensively, and therefore the accuracy of scatterometer calibrations in these environments is also unknown. This investigation seeks to determine the uncertainty in the H*Wind product and estimate the contributions of several potential error sources. These error sources include random observation errors, relative bias between different data types, temporal drift resulting from combining non-simultaneous measurements, and smoothing and interpolation errors in the H*Wind software. The effects of relative bias between different data types and random observation errors are determined by performing statistical calculations on the observed wind speeds. We show that in the absence of large biases, the total contribution of all error sources results in an uncertainty of approximately 7% near the storm center, which increases to nearly 15% near the tropical storm force wind radius. The H*Wind analysis algorithm is found to introduce a positive bias to the wind speeds near the storm center, where the analyzed wind speeds are enhanced to match the highest observations. In addition, spectral analyses are performed to ensure that the filter wavelength of the final analysis product matches user specifications. With increased knowledge of these error sources and their effects, researchers will have a better understanding of the uncertainty in the H*Wind product, and can then judge the suitability of H*Wind for various research applications
Boisserie, M. (2010).
Generation of an empirical soil moisture initialization and its potential impact on subseasonal forecasting skill of continental precipitation and air temperature. Ph.D. thesis, Florida State University, Tallahassee, FL, FL.
Abstract: The effect of the PAR technique on the model soil moisture estimates is evaluated using the Global Soil Wetness Project Phase 2 (GSWP-2) multimodel analysis product (used as a proxy for global soil moisture observations) and actual in-situ observations from the state of Illinois. The results show that overall the PAR technique is effective; across most of the globe, the seasonal and anomaly variability of the model soil moisture estimates well reproduce the values of GSWP-2 in the top 1.5 m soil layer; by comparing to in-situ observations in Illinois, we find that the seasonal and anomaly soil moisture variability is also well represented deep into the soil. Therefore, in this study, we produce a new global soil moisture analysis dataset that can be used for many land surface studies (crop modeling, water resource management, soil erosion, etc.). Then, the contribution of the resulting soil moisture analysis (used as initial conditions) on air temperature and precipitation forecasts are investigated. For this, we follow the experimental set up of a model intercomparison study over the time period 1986-1995, the Global Land-Atmosphere Coupling Experiment second phase (GLACE-2), in which the FSU/COAPS climate model has participated. The results of the summertime air temperature forecasts show a significant increase in skill across most of the U.S. at short-term to subseasonal time scales. No increase in summertime precipitation forecasting skill is found at short-term to subseasonal time scales between 1986 and 1995, except for the anomalous drought year of 1988. We also analyze the forecasts of two extreme hydrological events, the 1988 U.S. Drought and the 1993 U.S. flood. In general, the comparison of these two extreme hydrological event forecasts shows greater improvement for the summertime of 1988 than that of 1993, suggesting that soil moisture contributes more to the development of a drought than a flood. This result is consistent with Dirmeyer and Brubaker  and Weaver et al. . By analyzing the evaporative sources of these two extreme events using the back-trajectory methodology of Dirmeyer and Brubaker , we find similar results as this latter paper; the soil moisture-precipitation feedback mechanism seems to play a greater role during the drought year of 1988 than the flood year of 1993. Finally, the accuracy of this soil moisture initialization depends upon the quality of the precipitation dataset that is assimilated. Because of the lack of observed precipitation at a high temporal resolution (3-hourly) for the study period (1986-1995), a reanalysis product is used for precipitation assimilation in this study. It is important to keep in mind that precipitation data in reanalysis sometimes differ significantly from observations since precipitation is often not assimilated into the reanalysis model. In order to investigate that aspect, a similar analysis to that we performed in this study could be done using the 3-hourly Tropical Rainfall Measuring Mission (TRMM) dataset available for a the time period 1998-present. Then, since the TRMM dataset is a fully observational dataset, we expect the soil moisture initialization to be improved over that obtained in this study, which, in turn, may further increase the forecast skill.
Stauffer, C. L. (2018).
Air-sea coupling dependency on sea surface temperature fronts as observed by research vessel data. Bachelor's thesis, Florida State University, Tallahassee, FL.
Morey, S., Koch, M., Liu, Y., & Lee, S. - K. (2017). Florida's oceans and marine habitats in a changing climate. In E. P. Chassignet, J. W. Jones, V. Misra, & J. Obeysekera (Eds.),
Florida's climate: Changes, variations, & impacts (pp. 391–425). Gainesville, FL: Florida Climate Institute.