Hurricane Isaac: A Longitudinal Analysis of Storm Characteristics and Power Outage Risk

Area of Science:

• Environmental Science
• Disaster Management
• Data Science

Background:

• Hurricane Isaac made landfall in August 2012, causing significant power outages in Louisiana.
• The storm brought high winds, storm surge, and flooding, leading to prolonged power disruptions across the state.

Purpose of the Study:

• To analyze hourly power outage data in Louisiana during Hurricane Isaac.
• To correlate power outages with storm conditions (wind, rainfall, storm surge) geographically.
• To understand the causal drivers of power outages during hurricanes.

Main Methods:

• Collected hourly power outage data by zip code for Louisiana.
• Utilized a nonparametric ensemble data mining approach to correlate outages with storm conditions.
• Analyzed the geographical variation in the importance of storm characteristics.

Main Results:

• The statistical importance of storm characteristics for power outages varied geographically.
• Wind speed, precipitation, and previous outages were generally highly important.
• Storm surge showed lower importance, even in areas with significant surge.

Conclusions:

• Geographical variations in storm characteristic importance influence power outages.
• Precipitation is a significant factor in power outages, alongside wind.
• Including precipitation and other covariates can improve power outage forecasting models' accuracy and robustness.