Analysis of the interannual variability of tropical cyclones striking the California coast based on statistical downscaling

Fernando J. Mendez1, Ana Rueda1, Patrick Barnard2, Nobuhito Mori3, Sota Nakajo3, Antonio Espejo4, Manuel del Jesus4, Javier Diez Sierra5, Antonio S Cofino6 and Paula Camus4, (1)University of Cantabria, Ciencias y Tecnicas del Agua y del Medio Ambiente, Santander, Spain, (2)USGS California Water Science Center San Diego, San Diego, CA, United States, (3)Kyoto University, Kyoto, Japan, (4)Environmental Hydraulics Institute, Universidad de Cantabria, Santander, Spain, (5)Enviromental Hydraulics Institute of Cantabria, Santander, Spain, (6)University of Cantabria, Departamento de Matematica Aplicada y Ciencias de la Computación, Santander, Spain
Abstract:
Hurricanes hitting California have a very low ocurrence probability due to typically cool ocean temperature and westward tracks. However, damages associated to these improbable events would be dramatic in Southern California and understanding the oceanographic and atmospheric drivers is of paramount importance for coastal risk management for present and future climates.

A statistical analysis of the historical events is very difficult due to the limited resolution of atmospheric and oceanographic forcing data available. In this work, we propose a combination of: (a) statistical downscaling methods (Espejo et al, 2015); and (b) a synthetic stochastic tropical cyclone (TC) model (Nakajo et al, 2014). To build the statistical downscaling model, Y=f(X), we apply a combination of principal component analysis and the k-means classification algorithm to find representative patterns from a potential TC index derived from large-scale SST fields in Eastern Central Pacific (predictor X) and the associated tropical cyclone ocurrence (predictand Y). SST data comes from NOAA Extended Reconstructed SST V3b providing information from 1854 to 2013 on a 2.0 degree x 2.0 degree global grid. As data for the historical occurrence and paths of tropical cycloneas are scarce, we apply a stochastic TC model which is based on a Monte Carlo simulation of the joint distribution of track, minimum sea level pressure and translation speed of the historical events in the Eastern Central Pacific Ocean. Results will show the ability of the approach to explain seasonal-to-interannual variability of the predictor X, which is clearly related to El Niño Southern Oscillation.

References

Espejo, A., Méndez, F.J., Diez, J., Medina, R., Al-Yahyai, S. (2015) Seasonal probabilistic forecasting of tropical cyclone activity in the North Indian Ocean, Journal of Flood Risk Management, DOI: 10.1111/jfr3.12197

Nakajo, S., N. Mori, T. Yasuda, and H. Mase (2014) Global Stochastic Tropical Cyclone Model Based on Principal Component Analysis and Cluster Analysis, Journal of Applied Meteorology and Climatology, DOI: 10.1175/JAMC-D-13-08.1

Back to: Tropical Cyclone-Ocean Interactions: From Weather to Climate II Posters