Atlantic Basin Tropical Cyclone Storm Track Analysis Desmond

Atlantic Basin Tropical Cyclone Storm Track Analysis Desmond Carroll Advisor: Frank Hardisty

Goals • Analyze Tropical Cyclone susceptibility in the Caribbean. • Enhance seasonal forecasts by examining correlation between basin wide activity and small island risk. • Develop a web based visualization of the results.

Background

N. Atlantic Tropical Cyclone Season Not exclusively an American problem. Cayman Cuba The Bahamas

The North Atlantic Basin (Google Earth, 2010)

The NOAA HURDAT Dataset (NOAA, 2010)

The Extended Best Track Dataset

Synthetic Track Generation • Augment the existing data. • Analyze the extreme cases. • Compare probabilistic versus deterministic.

Problem • Small densely populated islands. • A line track may be too general. • Neither the HURDAT nor EBT alone sufficient.

Research Questions • Will the inclusion of storm size in track modeling problems improve the quality of the results in the caribbean? • Is there any link between the total number of named storms in the Atlantic basin and the susceptibility of individual islands in The Bahamas?

Objectives

Buffer Analysis • Generate a method for creating buffers from EBT data (ESRI, 1997)

Modeling Asymmetry • Get closer to simulating the real variability in storm paths. (NASA, 2006)

Track Analysis Augment the HURDAT with the data similar to EBT (Rogers and Spirnak, 2006)

Visualization • Provide a public outlet. • Place the focus on the less studied islands. • Explore HTML 5’s Canvas element.

Methods

Libraries • Geospatial Libraries – Open Source solutions – Solve well known problems – Python bindings

Buffers GDAL – Geospatial Data Abstraction Library – Data manipulation – Projections

Buffers • Shapely – Post. GIS Inspired – Convex Hull – Cascading Union Algorithms

Initial Run

Markov Chain Monte Carlo • Unknown processes. • Probability functions. • Many applications.

MCMC Random Walk (Patil et al. , 2010)

Application to HURDAT • Back test model for validity. • Augment storm tracks with asymmetry. • Extreme events can be examined in more. detail.

Current Web Visualization Technologies • Javascript client – HTML 5 Canvas Element • Openlayers framework • Geoserver middleware • Post. GRESQL/Post. GIS backend

Site Layout http: //loggedout. org

Future Work • • Conduct analysis for various storm strengths. Investigate temporal effects. Couple with storm track model. Employ more deterministic factors.

Questions?

References Demuth, J. , M. De. Maria, and J. A. Knaff, 2006: Improvement of advanced microwave sounder unit tropical cyclone intensity and size estimation algorithms. J. Appl. Meteor. , 45, 1573 -1581. Emanuel et al. A statistical deterministic approach to hurricane risk assessment. Bulletin of the American Meteorological Society (2006) vol. 87 (3) pp. 299 -314 Hall and Jewson. Statistical modelling of North Atlantic tropical cyclone tracks. Tellus A (2007) vol. 59 (4) pp. 486 -498 Jiechen Wang et al. Review of Buffer Generation Algorithm Studies. Intelligent Information Technology Application, 2008. IITA '08. Second International Symposium on (2008) vol. 2 pp. 911 – 917 National Oceanic and Atmospheric Administration. (2005, December 9). National Hurricane Center. Retrieved May 25, 2010, from National Hurrican Center: http: //www. nhc. noaa. gov/

References Patil, A. , D. Huard and C. J. Fonnesbeck. 2010. Py. MC: Bayesian Stochastic Modelling in Python. Journal of Statistical Software, 35(4), pp. 1 -81. Rumpf et al. Stochastic modelling of tropical cyclone tracks. Mathematical Methods of Operations Research (2007) vol. 66 (3) pp. 475 -490 Rumpf et al. Tropical cyclone hazard assessment using model-based track simulation. Natural hazards (2009) vol. 48 (3) pp. 383 -398 Wang Jiechen et al. A Novel Method of Buffer Generation Based on Vector Boundary Tracing. Information Technology and Applications, 2009. IFITA '09. International Forum on (2009) vol. 1 pp. 579 - 582