Characteristics and GFS Forecast Accuracy of Intraseasonal Shifts in the Arctic Oscillation Index

• Other Authors: Visco, Travis Connor (authoraut), Fuelberg, Henry E. (professor directing thesis), Hart, Robert E. (committee member), Sura, Philip (committee member), Department of Earth, Ocean and Atmospheric Sciences (degree granting department), Florida State University (degree granting institution)
• Format: Text
• Language: English
• Published: Tallahassee, Florida: Florida State University 2012
• Subjects: Earth sciences; Oceanography; Atmospheric sciences; Geophysics; Arctic
• Online Access: https://diginole.lib.fsu.edu/islandora/object/fsu%3A183323/datastream/TN/view/Characteristics%20and%20GFS%20Forecast%20Accuracy%20of%20Intraseasonal%20Shifts%20in%20the%20Arctic%20Oscillation%20Index.jpg

This study evaluates the characteristics and forecast accuracy of the Arctic Oscillation (AO) Index on an intraseasonal time scale. The Arctic Oscillation is a natural pattern of time varying sea-level pressure anomalies that is one of the leading modes of weather variability in the Northern Hemisphere. Sustained shifts in the AO Index can lead to pronounced and sudden changes in weather patterns that can have dramatic economic and social impacts. Previous studies have described characteristics and trends in the AO, but on seasonal and decadal time scales. Focusing on short time scales that can be depicted by Numerical Weather Prediction models, this study describes the AO's influence on surface temperature and the ability of the Global Forecast System (GFS) and Global Ensemble Forecast System (GEFS) numerical models to forecast changes in the AO index. Forecast performance is investigated over a range of atmospheric conditions from 2000-2011. Evaluation metrics include Probability of Detection, False Alarm Rate, and Critical Success Index. In addition, average forecast error is quantified through the use of absolute error calculations. Together, it is presented which evaluation techniques best enhance the AO Index forecast accuracy of the GFS and GEFS models, along with the expected forecast error that the models and methodologies provide. Results conclude that shorter period forecasts that utilize smoothing filters produce the best model performance with the least forecast error. The GFS and GEFS models have enhanced performance when the strength of the shift in the AO Index is sufficiently large (> 2 standard deviations). In addition, during the highly variable winter, forecast performance is largely diminished. Submitted Note: A Thesis submitted to the Department of Earth, Ocean, and Atmospheric Sciencein partial fulfillment of the requirements for the degree of Master of Science. Degree Awarded: Fall Semester, 2012. Date of Defense: August 7, 2012. Keywords: Arctic Oscillation, Cold air ...