Forceful Disasters and Omnipotent Data

by Maria Antoniou, ICONHIC Research Team
3min read

Natural Disaster Management is now in the midst of a revolutionary transformation. This rapid shift has been driven largely by technological changes – Big Data and Advanced Analytics, the Internet of Things (IoT), Robotics, and Artificial Intelligence – collectively described as the Fourth Industrial Revolution. Arriving at phenomenal speed, its consequences are already evident across all four pillars of Natural Disaster Management: Mitigation, Preparedness, Response & Recovery [1]. The Big Data ecosystem has come to radically change the ways we form our Natural Disaster Management strategies, in order to reduce human and economic losses. But how?

New Orleans flooding during hurricane Katrina in 2015
(Photo credits: pixy.org)

Through monitoring: New generation seismographs, UAVs, satellites, or airborne and terrestrial LiDAR offer improved remote sensing capabilities, making it possible to survey and assess critical areas post- or pre-disaster. One of the most remarkable contributions of remote sensing imagery, for example, involves rapid post-disaster damage assessment through change detection [2]. By monitoring and evaluating a flooded area, data regarding the flood-prone zone or maximum water level rise can emerge and be used in creating new prevention strategies next time a similar situation arises. Crowdsourcing data from citizens’ smartphones (e.g. vibration data during an earthquake) or information from social media (e.g. twitter posts) also offer new potential for monitoring hazards such as earthquakes, hurricanes, or floods. The USGS ‘Did You Feel It?’ application is a great example of gathering crowdsourcing data; the application collects information from people who felt an earthquake and creates maps that show what people experienced and the extent of the damage.

Through precise forecasting & hazard vulnerability assessments: Satellite data, for instance, enable scientists to identify natural hazard risks in spatial terms. A number of real-time, or quasi-real-time early warning systems that have emerged the past few years, utilize such data in order to timely inform stakeholders on appropriate disaster management strategies. The NOAA’s National Hurricane Center (NHC) issues warnings, forecasts, and analyses of hurricanes or other hazardous tropical weather, making use of satellite images and geospatial data, and informs citizens and the government by sharing information on its website or Twitter account. Similarly, spatially distributed earthquake hazard maps can facilitate governments in ensuring that buildings at earthquake-prone areas are constructed in such a way, that can withstand earthquakes of small magnitudes without causing any damage to life or property. Moreover, phone metadata, i.e. mobile GPS and CDR (Call Data Records), can be used to estimate population exposure to a specific hazard or natural disaster event, that can be then determined via estimation of the population distribution.

With natural disasters striking at an increased frequency and a dazzling abundance of data being generated in their aftermath, comes a new paradigm: one in which we are called to rapidly and efficiently harness a small fraction of meaningful data to better prepare and respond to disaster.

By guiding disaster response: Remote sensing can be used to provide rapid real-time mapping and damage assessment caused by hurricanes and earthquakes. UAV aerial imagery, with a higher spatial resolution and much faster processing compared to satellite imagery, can form a valuable tool in post-disaster rescuing operations, by quickly tracking survivors [3]. Moreover, social media can be monitored to provide early warning or guide rescue teams on threats ranging from disease outbreaks to earthquakes [4], while mobile phone data may provide important information on population movements after a disaster.

By raising awareness to communities: Resilient communities should be able to strengthen their infrastructure, manage their natural ecosystems, and maintain the social networks that make them strong. The longer-term potential of Big Data lies in their capacity to raise societal awareness and urge citizens to take action before, during, and after a natural disaster strikes. The MyShake app for example (University of Berkeley), uses smartphones as earthquake sensors and envisages to build a worldwide earthquake early warning network, so that communities can reduce the impact of earthquakes. The app keeps users informed about earthquakes and monitors for them using their phones’ data.

Data-Driven Decision Support in Natural Disaster Management (Image credits: Adobe Stock)

From enhanced monitoring to predictive mapping, and from crisis management to increased community resilience, the power of Data Analytics represents a tremendous opportunity to drill down and tap into critical and -otherwise- perishable insights of the cosmos of Natural Disasters globally. Each time a catastrophe occurs, data of immense value are generated, and with the right tools they can be collected, analyzed, and utilized to better understand the nature of hazards, to accurately assess the vulnerability of our infrastructure, to more effectively inform decisions on emergency operations, and to eventually fuel simulations, predictive planning, and all others means of proactive and preemptive Disaster Management. From a business standpoint, the value is also great: from creation to analysis and to archiving, disaster-related data produce information of high integrity, reproducibility, and reliability, that drive decision making and build ever-growing steam of intelligence. Without them, enterprises and organizations are ill-prepared to make sound decisions, and are facing the risk of becoming obsolete.

With natural disasters striking at an increased frequency and a dazzling abundance of data being generated in their aftermath, comes a new paradigm: one in which we are called to rapidly and efficiently harness a small fraction of meaningful data to better prepare and respond to disaster.

References

1. Alliance, D. P. (2015). Big Data for climate change and disaster resilience: Realising the benefits for developing countries.

2. Pradhan, B.; Tehrany, M.S.; Jebur, M.N. (2016). A new semiautomated detection mapping of flood extent from TerraSAR-X satellite image using rule-based classification and Taguchi optimization techniques. IEEE Trans. Geosci. Remote Sens., 54, 4331–4342.

3. Ofli, F.; Meier, P.; Imran, M.; Castillo, C.; Tuia, D.; Rey, N.; Briant, J.; Millet, P.; Reinhard, F.; Parkan, M.; et al. (2016). Combining human computing and machine learning to make sense of big (aerial) data for disaster response. Big Data, 4, 47–59.

4. Muniz-Rodriguez, K., Ofori, S. K., Bayliss, L. C., Schwind, J. S., Diallo, K., Liu, M., … & Fung, I. C. H. (2020). Social media use in emergency response to natural disasters: a systematic review with a public health perspective. Disaster medicine and public health preparedness, 14(1), 139-149.