Climate Change effects on Storm Surge in Coastal Waters Surrounding New York and New Jersey

The figure above depicts the typical definition of storm surge as the difference between observed and predicted water levels.

  • Dynamic regression and ARMA model for water level/storm surge
  • Predictive methods for heavy-tailed data
  • Numerical modeling of storm surge
  • Ensemble-based forecasting techniques
  • Joint probability method/risk assessments

The coastal regions of the mid-Atlantic states lay within a region that experiences frequent and, often times, powerful coastal storms. One significant and costly affect these storms have on the coastal infrastructure is the result of storm surge (depicted in the image above). In recent years, events such as Hurricane Irene (2011) and Hurricane Sandy (2012) have brought to light the incredible destructive force coastal storms can have.  The region also encounters frequent extra-tropical cyclones known as nor’easters during the late fall and winter months due to it’s location near a significant gradient in temperature between the gulf stream and continental air. Coastal flooding is a factor in both types of storms (tropical and extra-tropical).

An important issue this region is currently trying to understand is how these storms will change in frequency and intensity given a changing climate and subsequent rising sea level.  The vast majority of the infrastructure in lower Manhattan and along the south shore of Long Island is a couple meters above sea level. Work on the topic of storm surge includes the utilization of statistical models  to predict surge through 10 meter winds and sea level pressure. By forcing statistical models with global climate atmospheric data, regional variations in storm surge climatology can be analyzed at particular coastal stations with minimal computational effort and reasonable accuracy. Of primary focus is how the change in the surge climatology in the NY/NJ coastal region will vary in the near future and what are the leading causes of this change (rising sea level, cyclone density variations, etc.).

Future projects could involve:

  • Creating ensemble based methods for probabilistic forecasts of storm surge. Validation of probabilistic forecast guidance.
  • Numerical atmospheric-surge modeling systems.
  • Budget modeling of storm surge. Understanding the role of specific forcing mechanisms over time and spatially in a piecewise sense.
  • Impacts of waves on storm surge

Surge forecasts made by ADCIRC for Hurricane Gloria

Forecasts of storm surge at Battery Park, New York at a 24-h lead time to Hurricane Sandy using dynamic regression.

An example of heavy tailed data (surge). The most critical events occur less than < 1.0 % of the time (black line = 99th percentile).

Students Involved

Keith Roberts (2012-2015)

Contact | Storm Surge Page

Publications

Roberts, Keith J.; Colle, Brian A.; Georgas, Nickitas; Munch, Steve: “A Regression-based Approach for Cool-Season Storm Surge Predictions along the New York/New Jersey Coast” (submitted on December 5 to Journal of Applied Meteorology).

Roberts, Keith J.; Colle, Brian A.; Zhang, Zhenhai; Korfe,Nathan : “21st Century Projections of storm surge in the New York/New Jersey Bight consistent with the RCP8.5 CMIP5 global warming scenario.”  (in draft, intended journal: Journal of Climate.

Presentations

Roberts, K., Colle, B.A., A Statistical Approach to Understanding the Long-Term Variability of Storm Surge in the New York Bight