Back to search

National and Kapodistrian University of Athens

Citizen Science and dynamic data assimilation methods for wildfire mitigation

Abstract

dc:description

Wildfires are natural hazards with severe consequences worryingly worsening for many climate-change-affected regions of our planet. Never before was it of such paramount importance to develop novel methods and tools to empower our communities in combating the evermore dangerous wildfires more effectively. The focus of this Ph.D. dissertation was to introduce such methods and implement them in new software tools and services to help us mitigate the growing wildfires risk. One of the most crucial aspects of wildfire risk mitigation is predicting the wildfire’s course accurately and as soon as a fire event is detected. The prediction can be accomplished by employing a wildfire model and propagation simulation algorithms that use complex input data layers describing the weather and the forest landscape. We have developed FLogA (Fire Logic Animation), a set of methods and a web-based user-friendly software tool allowing the user to draw a forest area anywhere in Europe, insert fire ignition points, and generate all input data layers required on the fly. FLogA then simulates and animates the behavior of the evolving fire line under different weather conditions. FLogA is now the first wildfire simulator to run entirely on mobile platforms, and to provide a way to discover actual wildfire events from satellite hotspot services and perform analyses using dynamic data-driven simulations with the option to choose among multiple data assimilation methods.Another critical aspect of wildfire risk mitigation is monitoring the wildfire’s course. Unfortunately, technologies that provide real-time fireline information, such as satellites, in-field sensors, and social media texts exhibit low spatial/temporal resolution or cannot be deployed cost-effectively in widespread geographical areas. In this dissertation, we introduce methods that, for the first time, enable citizens-volunteers to provide real-time wildfire hotspot reports to benefit their community’s wildfire resilience and become active geographic data providers during crisis events. To enable our Citizen Science approach to wildfires mitigation, we designed, developed, and implemented algorithms and a novel software service called CITISENS (citizens as sensors). By exploiting commodity smartphone sensors, registered citizen-volunteers can easily georeference from a safe distance to a fireline and report its coordinates as they photograph it. While the citizen targets a wildfire hotspot with her smartphone camera from a safe distance, our georeferencing mobile algorithm consumes data generated from the location and orientation sensors of the smartphone to compute the view-ray of the device and employs a digital elevation model to estimate the coordinates where the view-ray intersects with the topography. We have tested the georeferencing accuracy, and it is on par with, or even better, than existing polar-orbit satellite wildfire hotspot services. Monitoring a wildfire’s course is not only essential to us humans. Indeed, wildfire simulators also need ground truth data when predicting the course of large wildfires where simulation drift inevitably comes into play, and the wildfire model needs periodic recalibration. The concept of Dynamic Data-Driven Assimilation Systems (DDDAS) has been shown to combat simulation drift effectively, at least when assimilating high-accuracy, laboratory-grade ground truth data. However, to operationalize wildfire DDDASs, the effectiveness of the systems must be proved in the real world. So far, limited efforts to validate wildfire DDDAS with real-time and Real-World Data (RWD) had limited success. We used FLogA and CITISENS as an end-to-end system to design and evaluate alternative DDDAS approaches using real-time RWD from satellite hotspot services and citizen-contributed synthetic hotspot reports generated based on RWD from two large fires in Greece. Our results suggest a marked improvement over non-DDDAS approaches when assimilating noisy RWD, showing how a citizenship engagement concept in wildfire monitoring and hotspots reporting has the potential to contribute high-quality ground truth data and thus quality wildfire model calibrations to form an effective crowdsourced DDDAS system. We intend to extend this research to offer real-time prediction with dynamic assimilation of satellite-based hotspots augmented with citizen-reported hotspots for improved predictive accuracy. We envision a decision support system promoting safe citizen-volunteer participation in wildfires mitigation with many benefits for the involved communities when a wildfire is approaching. Such a dynamic closed-loop trustworthy system may become a model for developing more citizen science-inspired cyber-physical systems to improve citizen communities’ resilience against devastating environmental hazards.

Degree

thesis:*
Grantor dc:publisher
National and Kapodistrian University of Athens
Year dc:date
2022

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Bogdos, Nikolaos
  • Μπόγδος, Νικόλαος

Subjects

dc:subject × 23

Rights

Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/52255

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Bogdos, Nikolaos; Μπόγδος, Νικόλαος. Citizen Science and dynamic data assimilation methods for wildfire mitigation. National and Kapodistrian University of Athens, 2022. http://hdl.handle.net/10442/hedi/52255