AI For Crisis Communication: A Revolutionary Strategy To Disaster Management

http://amresearchreview.com/index.php/Journal/about Volume 3, Issue 11 (2025) Online ISSN Print ISSN . . 3007-3197 3007-3189 http://amresearchreview.com/index.php/Journal/about Page 398 AI For Crisis Communication: A Revolutionary Strategy To Disaster Management Itba Tahreem Department of Media and development Communication, The University of Punjab, Lahore, 54000, Pakistan Email: [email protected] ORCID id: https://orcid.org/0009-0006-9979-7620 Laraib Noor Department of Media and development Communication, The University of Punjab, Lahore, 54000, Pakistan Aneeba Amin Department of Journalism, The University of Punjab, Lahore, 54000, Pakistan The advancement of technical methods and research approaches in the field of hazards and disaster studies is considered a pivotal aspect of disaster management. Technologies such as artificial intelligence (AI) are increasingly employed in areas like tracking, mapping, geospatial analysis, remote sensing, robotics, drone operations, machine learning, telecommunications, hotspot detection, urban planning for smart cities, transportation, and environmental impact assessments. These technological tools play a significant role in shaping societal responses to hazards and disasters, influencing research across social sciences. This paper explores how AI is currently utilized in all four phases of disaster management and underscores its importance in enhancing the efficiency, precision, and preparedness of responses. The integration of Geographic Information Systems (GIS) and Remote Sensing (RS) further strengthens disaster management by improving planning, analysis, situational awareness, and recovery efforts. GIS and RS are widely recognized as essential tools in supporting disaster management, with visualization techniques, satellite imagery, and AI-driven analyses that enable governments to make prompt, informed decisions in the wake of natural disasters. Key Words : Disaster Management, Artificial Intelligence, Geographical Information System, Crisis Communication Introduction Geographic Information Systems (GIS) and artificial intelligence (AI) have evolved very much over the recent years and become inseparable in emergency situations. Disasters are events that lead to death, damage or loss of property or adverse environmental impacts (Ajzen, 2020). According to the Center of Investigation on A B S T R A C T http://amresearchreview.com/index.php/Journal/about Volume 3, Issue 11 (2025) Online ISSN Print ISSN . . 3007-3197 3007-3189 http://amresearchreview.com/index.php/Journal/about Page 399 Epidemiology and Natural Disasters, the economic damage to disaster affected countries was estimated at approximately 3 trillion between 1998 and 2017 (Rehman et al., 2019). The United States had the highest casualties with losses amounting to approximately $1 trillion followed by China, Japan and India. The UN Refugee Agency has noted that the rate of catastrophe has nearly doubled over the past 2 decades with the Asia-Pacific region particularly prone since 1995 (Abid et al., 2020). Many variables also significantly influence the process of disaster response planning, including the geography, climate, ecology, and resources of the region. It is recommended to apply the concepts of management science and operations research to enhance the resilience of emergency response and take into consideration the population impact on the allocation of resources (Ogie, R. I., Rho, J. C., & Clarke, R. J., 2018). There have been several scholarly studies evaluating the application of AI in disaster relief (Nunavath & Goodwin, 2019). The information related to significant disasters caused by natural disasters should be selected and ranked the highest, as the crisis management setting in other countries may be vastly dissimilar to the crisis management setting in India (Abid et al., 2021). Strong approaches to reducing disaster impacts are the construction of resilience, preparedness to vulnerability, and preventive action (Yang et al., 2020; Costache, 2019). Many researchers investigate the spatial distribution of flood risks and vulnerability to disasters with the help of AI and GIS (Arinta et al., 2019; Chen et al., 2020). GIS is a facilitator, which is used to manage the geographic data in order to provide a quick and effective response to deal with the disaster of floods and mitigate hazards in realtime relying on the input, storage, sharing, management, and delivery of geographic data (Abid et al., 2021). The purpose of this study is to evaluate how Geographic Information Systems (GIS) and artificial intelligence (AI) might enhance crisis communication tactics in the context of flood management and response.The main objective of this study is: To evaluate AI’s utility in enhancing crisis communication during flood scenarios. To determine the advantages and drawbacks of using AI in crisis communication To investigate how AI could potentially help to enhance the effectiveness and relevance of communication in emergencies. To explore how GIS could be combined with AI. This current study is based on some research questions like, RQ1.How does AI enhance the efficiency and reliability of crisis messages during the flood calamities? RQ2.How is it possible to implement AI into GIS for improving crisis communication? RQ3.What are the critical factors and issues faced in the implementation of AI and GIS during disaster management? RQ4.How will the integration of GIS with artificial intelligence improve public awareness and flood management? http://amresearchreview.com/index.php/Journal/about Volume 3, Issue 11 (2025) Online ISSN Print ISSN . . 3007-3197 3007-3189 http://amresearchreview.com/index.php/Journal/about Page 400 Literature Review Artificial Intelligence and Disaster Management Unlike human intelligence, artificial intelligence (AI) demonstrates intellect through technological advancements (Kumar & Sud, 2020). AI involves the integration of various machines to replicate human behaviors, focusing on information technology activities aimed at developing intelligent machines. Over the past decade, innovations in AI have significantly enhanced our capacity to forecast disasters and provide assistance during such events (Canon et al., 2018; Saravi et al., 2019). Notable advancements in AI are evident in disaster preparedness, crowdsourcing systems for information, and the distribution of rescue and humanitarian aid (Kumar et al., 2020; Park et al., 2018). While AI encompasses various forms, this paper concentrates on the applications of robotics, drones, data mining, deep learning, sensors, and methodologies to predict disasters and expedite rescue and relief efforts (Mosavi et al., 2018; Chakraborty et al., 2020). Although automation and robotics have been around for decades, new advancements in sensor technology and processing capacity have made robots into completely autonomous and intelligent machines rather than just simple decision-making tools (Park et al., 2017). Machine learning is a relatively new aspect of artificial intelligence compared to automated systems, which have been around for a while. Machine learning algorithms are classified as artificial intelligence (AI) since they are able to carry out specified tasks without explicit instructions by using patterns and conclusions found in the given data (Noymanee et al., 2017). It is possible to develop models that identify the changes in the satellite image and calculate the location that can be affected by a disaster and orchestrate the actions to rescue people with the help of convolutional neural networks (CNN). Aerial robots are able to fly long distances into the disaster fields to assess the damage and offer help (Axel et al., 2021). Machine learning is advanced software that is capable of studying by observing patterns within the text, images, videos, numbers and other forms of data and then applying that learning to predict consequences in situations that it has already encountered (Sulaiman et al., 2021; Lee et al., 2020). The primary goal of artificial intelligence is to enhance disaster management processes and automatize them. Artificial intelligence (AI) technologies, such as sensors, are useful in disaster response by improving the sharing of data through ontologies, supplying information to the disaster response organizations, and offering multi-agent solutions to real-world assistance and simulated scenarios. Emergency management is the ability to access information from different sources, synthesize it, and reach justifiable conclusions (Erdelj et al., 2017). Digital tools and the social media platform provide efficient and effective means by which the population can access and disseminate information (Villodre and Criado, 2020). Social media has been rapidly adopted by millions of people to share information in both natural and manmade disasters (Aisha et al., 2015; Shaluf et al., 2002). One of the artificial intelligence techniques that are adopted to predict hazards, risk assessment, and examination of the potential consequences of these hazards is regression analysis including linear, nonlinear, and logistic regression. Moreover, the http://amresearchreview.com/index.php/Journal/about Volume 3, Issue 11 (2025) Online ISSN Print ISSN . . 3007-3197 3007-3189 http://amresearchreview.com/index.php/Journal/about Page 401 support vector machines can be applied in risk assessment and accurate forecasting in time. Also, methods such as neural networks, principal component analysis, kmeans clustering, fuzzy clustering, hierarchical clustering and neural networks have been applied in the generation and assessment of training plans, mitigation tactics, and disaster evacuation plans (Sulistijono et al., 2016;Ci et al., 2019). Examples of popular deep learning methods that are applied in catastrophe mapping include convolutional neural networks, deep neural networks, and multilayer perceptrons. Interagency cooperation, damage assessment and disaster information systems are also necessary during emergencies (Fu et al., 2020;Baldazo et al., 2019). Table 1. Explains how different studies used AI for disaster management.It includes (Doshi et al., 2018;Chen et al., 2017;Amit et al., 2017; Ahmad et al., 2017;Vetrivel et al., 2018; Duarte et al., 2020; Nex et al., 2019; Syifa et al., 2019). Table 1. Previous studies which used AI for crisis communication and disaster management Theoretical Framework Information and Communication Technology for Development (ICT4D) is a theoretical framework that examines possible ways to use technology (e.g. artificial intelligence (AI) and communication methods) to promote economic and social development in regions where technological resources are often limited. The point is that, when implemented in a proper and successful way, it could become the means of bridging the gap in knowledge and making the resources more readily available. The ICT4D approach promotes inclusivity.It has ensured that the application of technology is in a wide range of settings, which include rural areas with poor infrastructure and low-income or developing countries.It has also ensured sustainability and requires development of technology that can be sustained and incorporated in the long run. For this current study, ICT4D is highly relevant because a crisis management system with ai tools must be easily accessible to all, no matter what class or geographical location they belong.With the help of AI it is possible to convey real time crisis communication alerts through social media even to the areas with low technological development.Its main focus is to ensure that important information is properly conveyed to the marginalized population or communities who would have been overlooked from standard form of communication during disaster. Methodology A comprehensive literature review was conducted to assess the role of artificial intelligence in disaster management, focusing on studies published between 2015 and 2020. A keyword search string was employed to identify relevant literature, utilizing specific keywords associated with the application of artificial intelligence in disaster management contexts. The search was restricted to articles published in journals indexed in reputable databases, including Scopus, Web of Science, and ScienceDirect. This approach facilitated the selection of diverse studies that addressed various aspects of AI in disaster management. A systematic list of keywords was curated to ensure a thorough evaluation of the literature, encompassing multiple global perspectives on the http://amresearchreview.com/index.php/Journal/about Volume 3, Issue 11 (2025) Online ISSN Print ISSN . . 3007-3197 3007-3189 http://amresearchreview.com/index.php/Journal/about Page 402 integration of AI technologies in disaster response and recovery efforts. Original research articles had an initial number of 112, found across all databases. The 61 selected articles included after the duplicate article removal and the title and abstract screening. Full-text review led to the final reviewing of 42 studies that met all inclusion criteria and were incorporated in the thematic analysis. Table 2. Search string (keyword analysis in international journals, 2015–2020) Figure 1 visually presents the keywords and methodologies identified from the search results within the database for the period 2015 to 2020.It demonstrates the change in research towards proactive crisis management based on machine learning and big data analytics. The majority of studies were found in the field of Natural Hazards, followed closely by research in Sustainability, Disasters, and the Journal of Natural Disasters. This research was based on a comprehensive review process aimed at evaluating the role of AI and GIS in disaster management. The Scopus database served as a primary source for identifying relevant studies due to its extensive coverage of the target domain. Additional literature was extracted from Science Direct and Web of Science, ensuring that the most current and reliable studies on AI applications in disaster management were included in the review. A broader selection of peer-reviewed journals and their associated studies was chosen to provide a comprehensive overview of research on artificial intelligence and disaster management (as shown in Figures 1, 2, and 3). The majority of these studies were published by researchers from China, the United States, South Korea, Iran, Australia, and Italy, covering the period from January 2015 to December 2020. http://amresearchreview.com/index.php/Journal/about Volume 3, Issue 11 (2025) Online ISSN Print ISSN . . 3007-3197 3007-3189 http://amresearchreview.com/index.php/Journal/about Page 403 Figure 1.Keywords assessed within different international journals(2015-2020). http://amresearchreview.com/index.php/Journal/about Volume 3, Issue 11 (2025) Online ISSN Print ISSN . . 3007-3197 3007-3189 http://amresearchreview.com/index.php/Journal/about Page 404 Figure 2. Distribution of studies published by different countries(2015-2020) As demonstrated in Figure 2, most of the reviewed research was done in technologically advanced countries especially China and the United States. The respective nations have powerful AI ecosystems and emergency facilities. Nevertheless, there was a notable underrepresentation of areas at high risk of disasters, including South Asia and Sub-Saharan Africa, which gives rise to doubt in the geographical balance of AI-based disaster technology. In the meantime, Figure 3 shows that the majority of studies used AI and GIS to discuss early warning systems and spatial flood risk analysis. The forecasting and mapping technologies were extensively studied, whereas public communication strategies and simulation training received less attention, implying that the overall research on disaster preparedness needs to be improved. Figure 3. Identified studies on artificial intelligence and Disaster management. Four Phases Of Disaster management Disaster management cycles (DMC) help to minimize concussion.The disaster management process unfolds in four distinct stages, each comprising interconnected activities. These stages are further categorized into three key phases: pre-disaster, during the disaster, and post-disaster. In the Disaster Management Cycle (DMC), these activities work together to mitigate the risk of human and material losses. Predisaster planning focuses on mitigation efforts, which include establishing preventive laws and implementing standards to address potential emergencies. Preparing based on available resources is critical, as preparedness is one of the most crucial stages in the DMC to minimize disaster impact. During a disaster, response and rescue operations take center stage, providing essential first aid, humanitarian support, and conducting initial damage assessments. Once the disaster passes, the recovery phase begins, which involves community rehabilitation, restoring livelihoods, and carrying out detailed damage evaluations to support long-term recovery efforts. http://amresearchreview.com/index.php/Journal/about Volume 3, Issue 11 (2025) Online ISSN Print ISSN . . 3007-3197 3007-3189 http://amresearchreview.com/index.php/Journal/about Page 405 Figure 4. Phases of Disaster Management(Before,during and after a disaster).(Sun, W., Bocchini, P., & Davison, B. D. (2020). Geographic Information System (GIS) and flood management There are many methods and techniques developed to investigate flood risks and risk evaluation. These are flood zoning, statistical index, random forest, logistics regression, the analytic network method and the AHP method. Flood zoning maps and the flood damages assessment with the help of geometric methods are replete with hydraulic science processes. The maps indicate the location of an occurrence of a flood and they help to calculate the level of damage (Fernandez, P., Mourato, S., & Moreira, M., 2016).Flood zoning has become a very crucial element of human safety and security. It also strengthens flood reduction strategies and reduces flood-related damage(Rumson, A. G., & Hallett, S. H. 2019). GIS combined with hydrologic engineering models (HEC-RAS) is customized to provide river maps. These models have been successfully applied in Warsaw, Columbia, and Dhaka, as well as many other flood-prone states (Ongdas et al., 2020; Vojtek et al., 2016).In South Asia, Malaysia, Indonesia, and Thailand also use HEC-RAS models for flood zoning and to successfully mitigate flood hazards (Curebal et al., 2016;Demir et al., 2016). GIS has also been proven to identify areas suitable for developing flood mitigation systems and evaluate the effectiveness of the available flood mitigation systems(Sulaiman et al., 2020).The research by Puttinaovarat and Horkaew 2020 addressed assisting flood disaster mitigation via an internetworking system using remote sensing, GIS, and deep learning (DL). http://amresearchreview.com/index.php/Journal/about Volume 3, Issue 11 (2025) Online ISSN Print ISSN . . 3007-3197 3007-3189 http://amresearchreview.com/index.php/Journal/about Page 406 Figure 5 Online Flood Information Management system (Puttinaovarat, S., & Horkaew, P., 2020) Disaster management is a strategic and multi-faceted procedure for mitigation, preparedness, response, and recovery to protect the vulnerable community and critical intrastate from any disaster. Working in the field of catastrophe risk reduction, researchers, policymakers, and government representatives have a shared understanding of disasters and adopt preventative measures before they happen. All disasters, though, are connected to how people deal with their aftermath. Thus, the preparation and execution of efficient disaster management procedures determine whether a project succeeds or fails. Additionally, a hazard might lead to a secondary hazard, like a tsunami that produces coastal flooding, which can have a huge impact. Therefore, AI is unquestionably the future of disaster management and a major force multiplier in the ability to safeguard people and property in the case of disaster. http://amresearchreview.com/index.php/Journal/about Volume 3, Issue 11 (2025) Online ISSN Print ISSN . . 3007-3197 3007-3189 http://amresearchreview.com/index.php/Journal/about Page 413 convolutional neural network. In Proceedings of the 2017 International Electronics Symposium on Knowledge Creation and Intelligent Computing (IES-KCIC) (pp. 239–245). Surabaya, Indonesia. https://doi.org/10.1109/IESKCIC.2017.8259732 Ahmad, K., Konstantin, P., Riegler, M., Conci, N., & Holversen, P. (2017). CNN and GAN based satellite and social media data fusion for disaster detection. CEUR Workshop Proceedings, 1984, 13–15. http://ceur-ws.org/Vol-1984/paper3.pdf Vetrivel, A., Gerke, M., Kerle, N., Nex, F., & Vosselman, G. (2018). Disaster damage detection through synergistic use of deep learning and 3D point cloud features derived from very high resolution oblique aerial images, and multiple-kernel learning. ISPRS Journal of Photogrammetry and Remote Sensing, 140, 45–59. https://doi.org/10.1016/j.isprsjprs.2018.03.016 Duarte, D., Nex, F., Kerle, N., & Vosselman, G. (2020). Satellite image classification of building damages using airborne data. ISPRS Annals of Photogrammetry, Remote Sensing and Spatial Information Sciences, IV, 4–7. https://doi.org/10.5194/isprs-annals-IV-4-2020 Nex, F., Duarte, D., Steenbeek, A., & Kerle, N. (2019). Towards real-time building damage mapping with low-cost UAV solutions. Remote Sensing, 11(3), 287. https://doi.org/10.3390/rs11030287 Syifa, M., Kadavi, P. R., & Lee, C.-W. (2019). An artificial intelligence application for post-earthquake damage mapping in Palu, Central Sulawesi, Indonesia. Sensors, 19(3), 542. https://doi.org/10.3390/s19030542 Sun, W., Bocchini, P., & Davison, B. D. (2020). Applications of artificial intelligence for disaster management. Natural Hazards, 103, 2631–2689. https://doi.org/10.1007/s11069-020-04263-0 Canon, M. J., Satuito, A., & Sy, C. (2018). Determining disaster risk management priorities through a neural network-based text classifier. In Proceedings of the 2018 International Symposium on Computer, Consumer and Control (IS3C) (pp. 237–241). Taichung, Taiwan. https://doi.org/10.1109/IS3C.2018.00067 Kumar, T. V. V., & Sud, K. (2020). AI and robotics in disaster studies. Palgrave Macmillan. Villodre, J., & Criado, J. I. (2020). User roles for emergency management in social media: Understanding actors’ behavior during the 2018 Majorca Island flash floods. Government Information Quarterly, 37(1), 101521. https://doi.org/10.1016/j.giq.2020.101521