Use of Emerging Technologies like Artificial Intelligence/Machine Learning in Disaster Risk Reduction: Opportunities, challenges, and prospects

In the recent era of digital world, Artificial intelligence has become most popular and innovative technology, which is playing a critical role in managing disaster risk and predicting the extreme events. It has a wide range of applications from the forecasting the extreme weather events, to development of hazard maps, to the detection of events in real time basis, and integrated with spatial technology can provide the real-time situational awareness for any disaster event, and beyond.

When we talk about AI let’s try to understand the AI form technology perspectives and how this can improve the disasters management practices.  AI refers to emerging technologies that mimic or even outperform the human intelligence while performing certain tasks or even to automate the things. AI tools and techs offer new opportunities related to applications in, for instance, observational data pre-processing as well as forecast model output for post-processing.

For instance, a literature survey of recent of recent research (2018–2023) shows that AI and machine learning technologies are widely being used to improve disasters early warning and alert systems and also helps in generating hazards and susceptibility maps.

Now the emerging technologies clearly demonstrates that AI-related tools are being applied to help us in better and rapid way of managing the impacts of types of natural hazards and disasters.

However, developing these tools and technology stack is very challenging, given the lack of on-site observation networks across the country.

Therefore, the experts from all over the world, are now using AI to create a tool that predicts the probability of observing the extreme events for a specific day at a given location under certain meteorological and climatological conditions. As a products various new AI and ML based models have been developed by scientific communities which are now able to predict severe convective conditions or even those models can detects days with a high potential of severe hail- or windstorms

These tools are now used to observed the historical data from data-rich regions to extrapolate to other locations worldwide with limited data availability using transfer of technology and available learnings.

Finally, the scientist has developed the downscaling approaches to simulate and analyze these events with the Weather Research and Forecasting (WRF), numerical based weather prediction models. This has shown great potential of AI/ML technologies for forecasting severe storms and producing hazard maps for any locations.

There are many more AI based tools are being used in geodesy to detect tsunamis and advanced Global Navigation Satellite System (GNSS) real-time processing data for positioning and ionospheric imaging provides very significant improvements to Tsunami Disaster Early Warning. Now GNSS technology is used in seismology to study ground displacements as well as to monitor perturbations in ionospheric total electron content (TEC) that commonly follow seismic events.

Now the world has witnessed that the Earth observations data combined with AI and ML tools, can be used to assess disasters impacts and prepare ahead of time.

Event many scientists are exploring that, how AI technology can be used to provide effective communication in the case of disasters and how AI can help disaster responders to assess the severity of risk and prioritize when and where to respond.

 Challenges to the use of AI for DRR

When we applying the AI/ML based tools and technology in Disaster Risk Reduction or disaster response, there are many challenges can appear at any stage of disasters cycle at various stages like reliable data availability, model development, training of AI models, or operational implementation of models in various stages.

Especially during the collection and handling of vast datasets, it is very important to consider: few thing-like biases in training/testing datasets, the distributed AI technologies within the data domain and there are some ethical issues as well.

In terms of biases in the collection and training of models using datasets, it is very important to ensure that datasets are correctly sampled and also to ensure the data is sufficiently represented of each pattern for the problem in question. Because providing biased date to AI models can provide the wrong predictions or biased outcomes.

Prospects in Emerging technology in DRR

In the field of Disaster Risk Reduction, there is considerable interest by scientific community in exploring the benefits of using AI/ML to find the better and trained models which can sustain in the long term to develop batter disaster management planning and strategies.

In particular, we believe that scientific innovations and many efforts are still needed to improve the AI based models also put as more as knowledge and learning to be circulated among the masses in the field. More educational materials should be provided by technologies and scientist to support capacity building programs, for ensuring the availability of computational resources, knowledge, and other hardware tools and for bridging the digital divide.

Opinion and Edited by Dr. Brijendra Kumar Mishra, (Disaster Risk Reduction Expert)