A USE CASE STUDY OF HURRICANE MATTHEW OPERATIONAL

A USE CASE STUDY OF HURRICANE MATTHEW OPERATIONAL READINESS LEVELS AND USE CASE TEMPLATES – 3 DM WORKSHOP Kari Hicks GIS Data Analyst Duke Energy

OBJECTIVE & BACKGROUND • 3 DM Workshop (Data-Driven Decision Making) • All Hazards Consortium, Fleet Response Working Group, ESIP • Objectives OF 3 DM: • Identify key “sector operational issues” • Develop new sector Use Cases • Identify government and private sector data sets, information, solutions • Develop or link to solutions • Leverage the SISE (Sensitive information Sharing Environment) • 3 Primary Use Cases: • Situational Awareness • Resource Movement • Damage Assessment

RESOURCE MOVEMENT/MANAGEMENT USE CASE HURRICANE MATTHEW CASE STUDY

HURRICANE MATTHEW

HURRICANE MATTHEW “Total outages reached their peak level on October 9, with roughly 2. 5 million residential, commercial, and industrial electricity customers without service across five states” (eia. gov).

USE CASE STUDY: THE ACTORS Howard Fowler VP Engineering & Technical Customer Relations Ron Osborne Emergency Preparedness Manager Storm: MW Incident Commander Jack Gardner Manager EAM Business System Support

INCIDENT COMMAND TIMELINE • October 5 th • Carolinas: • “Compared to yesterday’s forecast, the timing of impact from Hurricane Matthew on the Carolinas service area has slowed and with the more favorable storm track, the overall impact appears to be lower. ” • “We should have enough native resources to support the storm. ” • Florida: • “Hurricane Matthew will bring significant impacts to the Florida service area starting late Thursday night and lasting into the early morning, pre-dawn hours on Saturday as the storm lifts northward. The greatest impacts felt from hurricane strength winds will be limited to the immediate coastal counties and across the North Central zone of the Florida service area. ” • “We are acquiring 550 additional line resources in coordination with CDO to bolster the total non-native contractor line complement to 1, 000” • • Favorable Storm track Enough native resources Significant impacts to Florida Greatest impact from wind 12

13

ISSUES IDENTIFIED (INTERVIEWS AND ICS SUMMARIES) • Less emphasis of flood risk and forecast. This ended up becoming our biggest hurdle (more emphasis on wind and rain) • Eye wall had broken down and weakened by the time Matthew hit the Carolinas, but rain band had broadened • https: //www. youtube. com/watch? v=u. P 9 Czo. B 73 Gg • Challenges with hurricane forecast interpretation • All agents communicated a need for better damage prediction modeling • Underestimated # outages 23

THE DATA

THE ACTORS Ron Osborne: “We lack effective tools to balance resources based on predicted damage. ” Howard Fowler: “We could not infer from our flood datasets how long the flooding would last. ” Jack Gardner “I reached out to everyone I knew outside the company, just trying to see what datasets I could get my hands on. ”

USE CASE DATASETS • Based on user interviews, the following datasets are critical for this use case: • • Flood Prediction and Assessment Wind Live Traffic Damage Prediction Outages* -internal Facilities* -internal Staging Sites* -internal 26

FLOODING • Datasets Used • FIMAN map (NC Flood Inundation Mapping and Alert Network) • Rain Gauges (Advanced Hydrologic Prediction Service, NOAA/NWS) • Flood Plain • Issues • Broadened rain band • Trucks and Crews getting trapped by receding flood waters • Model coverage • When will waters recede?

WIND • Datasets Used: • Wind gust forecast (NOAA) • Live wind readings from stations • Issues: • Wind data more effective when contextualized • Role of vegetation • Sustained wind vs gust • Better as part of damage prediction product, rather than alone • Understanding Saffir-Simpson Hurricane Wind Scale

DAMAGE PREDICTION • Data “product” rather than dataset • Existing damage prediction models • Ability to balance resources

http: //www. tdworld. com/distribution/d amage-prediction-model-improvesstorm-response Connecticut Light and Power 30

ORLS The most important functionality of the ORLs is the ability to communicate the reliability of a dataset to people without the background or expertise to interpret the reliability themselves.

ORLS • Communication and guidance for interpretation • Contact? • Automated emails when changes are made? • Interpretation: • Hurricane track • Metadata

ORLS • Improved efficiency and readiness • No ad hoc efforts to scramble for data • Users know how to interpret data to apply to operational settings and situations

CONCLUSION • Underlying theme: interpretation and communication of raw scientific data • Tackle an improved damage prediction model • Operational Readiness Levels – help facilitate translation from provider to user