Putting analytics into disaster preparedness and response

Big Data Evangelist, IBM

Natural disasters are among the biggest risks that confront communities everywhere, and weather is one of the primary causes of these disasters. Global warming is going to increase the frequency and severity of adverse weather events around the world for the rest of our lives.

In Boston this past year, for example, the city spent over $50 million on snow removal when they only spent $5 million a couple years earlier. Increasing levels of heat-related deaths are likely to occur. More than 2,500 people perished during a recent heat wave in India, and more than 1,250 people succumbed to a recent heat wave in Karachi. Extreme heat events such as these are 50 times more likely to occur than they did a half century ago, according to The Weather Company.

In 2014, weather events also claimed more than 12,500 lives worldwide and cost societies nearly $110 billion.

Many countries have central agencies for predicting, preventing and remediating disasters, with regional and local jurisdictions often required to shoulder much of the operational responsibility for onsite first response. In addition, nongovernmental and volunteer organizations are often key participants in emergency response, assisting government agencies with rescue, medical, food, shelter, logistics and other time-sensitive, high-priority requirements.

Putting analytics to work for response teams

Though uncontrollable, adverse weather conditions—hurricanes, floods and blizzards—need not be catastrophic in their impact on society. To protect lives, property, infrastructure and the common welfare, governments at all levels need to plan for all plausible adverse weather scenarios. For emergency planning, preparedness and response, managers in the public sector require weather data analytics that are predictively accurate, up to the second and geospatially targeted down to the neighborhood and street level.

When atmospheric conditions cause or are contributing factors in disasters, government agencies require weather data analytics to plan and manage effective responses. A proactive and integrated response improves efficiencies when responding to immediate emergencies, sustaining lives or initiating recovery efforts. Weather data analytics provide several principal benefits.

Contingency planning

Having the ability to model and simulate many disaster scenarios enables government agencies to develop contingency plans suited to whatever nasty weather may come. Historical weather data, geospatial intelligence and predictive analytics enable planners to formulate preparedness, mitigation and response plans for a wider range of plausible scenarios than if only a single forecast were used.

Early warning

Being able to forecast weather-driven disasters hours or days in advance can make all the difference when lives and property are at stake. Predictive weather data analytics, such as the probabilistic 15-day tropical forecast enabled by WSI’s tools, can provide disaster preparedness coordinators with a day or two of extra lead time. They can use this extra time to coordinate comprehensive and effective responses by multiple organizations in public and private sectors.

Strike-zone severity profiling

Emergency planners can direct resources more effectively if they can use interactive map-based visualizations to profile disaster impact zones. Real-time, predictive and post hoc analyses can be used to mitigate or reduce the severity of disasters on lives, health, safety, property and infrastructure. Within seconds of a disaster’s occurrence, analytics-driven situational awareness enables emergency response teams to visualize the hardest hit areas and avoid guesswork. In addition, weather data analytics can be useful in identifying collateral impacts of disasters on crime rates, resource shortages and other critical matters that demand rapid response.

Blending weather data analyses with emergency planning

Public sector stakeholders responsible for emergency planning, disaster preparedness and response stand to benefit from these data-driven weather analyses. These analyses need to incorporate both the agency’s own emergency planning and monitoring data sets as well as weather data that enables them to comprehensively model all relevant scenarios. In addition to disaster emergency response, public sector agencies can also benefit from weather data analytics for various operational functions:

  • Predicting the impact of various weather scenarios on government equipment and other asset failures
  • Optimizing the deployment of fire, police, hospital and other key services during various weather scenarios
  • Assessing the impact of adverse weather on public safety during closings of schools, government offices and other facilities
  • Issuing public safety alerts and hyperlocal event-management alerts during adverse weather
  • Predicting weather-related demand on public resources, services and infrastructure—for example, water, roadways and sanitation 

Public sector administrators can learn how to do these weather-driven analyses for themselves by attending IBM Insight 2015, 25–29 October 2015, in Las Vegas, Nevada. At the event, IBM and The Weather Company demonstrate how to use The Weather Company’s weather data packages and IBM data science tools—such as Apache Spark as a service on the Bluemix cloud platform—to address highly urgent weather-related business challenges. These challenges include a powerful workbench and Spark as a service for building advanced, smart data applications for meteorological forecasting. In addition, learn how to use Spark as a Service to hack weather-related analytics challenges as well as all your analytics-related business challenges.