Utilization of a High Resolution Weather and Impact Model to Predict Hurricane Irene

For several years the Consolidated Edison Company of New York has been working with the IBM 2km Deep Thunder weather model and team from T.J. Watson Research Center. The project is to link the Deep Thunder weather model to an impact prediction tool. This tool would be able to provide advance notice of significant weather and its resultant impact on the overhead electrical system in Westchester County so that company operations may adequately prepare for the event.

The weather and impact model run twice daily (00z/12z) and provide a 3 day outlook for weather and impact. The Deep Thunder WRF model forecast provides temperature (dry & wet bulb), wind (sustained & gust), and precipitation at a 2km resolution across the Consolidated Edison service territory. The impact prediction tool is a statistical model that was developed using historical outage data from the company known as “job tickets” and relate them to the observed weather from Westchester County. The impact model receives its weather forecast data from the Deep Thunder version of WRF. It then provides a probabilistic forecast of “jobs” that will be caused as a result of the forecasted weather conditions.

In 2011 the weather and impact model had one of its most significant tests as Hurricane Irene barreled up the coast toward the NYC area. This presentation will review the Deep Thunder weather model forecasts leading up to the event and show how they affected the impact prediction tool. How these forecasts were used, lessons learned, potential adjustments to the model, and the challenges to working with this type of tool will also be discussed.