Skip to main content

Geospatial based Environment for Optimisation Systems Addressing Fire Emergencies

Periodic Reporting for period 2 - GEO-SAFE (Geospatial based Environment for Optimisation Systems Addressing Fire Emergencies)

Reporting period: 2018-05-01 to 2020-04-30

The EU and Australia are prone to destructive wildfires, affecting thousands of square kilometres each year causing significant human, economic and ecological losses. And the annual fire season in both regions is worsening, as are the associated human and economic costs. Collaboration between European and Australian researchers and firefighters enables them to acquire new knowledge and develop computational tools to assist in wildfire management. Understanding wildfire enables society to better address the risks posed and so protect lives, infrastructure and the environment. It is vital that European fire managers have opportunities to develop their wildfire skills by exchanging knowledge with Australian firefighting organisations and with researchers from different disciplines. Developing computational and training tools to assist with managing wildfires will ultimately result in reduced losses and improved use of resources. GEO-SAFE has three main objectives: wildfire knowledge creation; sharing knowledge amongst European/Australian academics and firefighters; skills development of researchers specialising in the development of firefighting tools. Specific objectives include:
1.Knowledge creation: 1. Develop tools (e.g. dynamic risk cartography, fire spread, fire suppression, evacuation, etc), to support decision making to enable optimal mitigation, response and deployment of resources during wildfire. 2. Understand how people behave when faced with wildfire and how this impacts evacuation decisions. 3. Understand end-user requirements and provide a means for testing models and methodologies. 4. Understand relevant management processes and develop training tools to facilitate the implementation of proposed management solutions.
2.Sharing of know-how: Creation of a research and development network through secondments and workshops between universities, labs and firefighting organisations in Australia and Europe.
3.Skills development: training of doctoral students and post-doctoral scientists and training of professional staff located in firefighting organisations.
While the frequency and size of wildfires around the world is growing, increasing the threat to human life, property and the environment, much can be achieved through risk reduction and mitigation. GEO-SAFE has demonstrated through developments in risk cartography, decision making models for prevention and mitigation and evacuation planning tools that society is not helpless when faced by the devastating power of wildfire. Using these tools and methods it is possible to take decisive actions to reduce the consequences of wildfires long before the fire starts. However, further financial and intellectual investment is essential to turn these exciting prospects into reality.
1. Knowledge Creation: Several mathematical and computer models have been developed including: 1) a model for controlled or prescribed burning taking into account environmental aspects to reduce fire risk; 2) models for fire spread using data gathered from historical fires, 3) a theoretical model for the firefighter problem on initial attack, multiple ignition and resource allocation; 4) a model for resource allocation related to asset protection; 5) an urban-scale agent based evacuation model including interaction of pedestrians, vehicles and wildfire spread; 6) a model for resource allocation in supported evacuation; 7) a deep learning model to assess the risk of fire ignition.
In addition, data were collected and analysed on historical wildfires to develop case studies for model testing and scenarios for fuel management, firebreak location, and resource allocation for initial attack and lives and property protection. Data on human response to wildfire and how this affects evacuation decisions were collected.
2. Sharing of know-how: Knowledge exchange was facilitated through the successful delivery of: 222 pm of secondments by European partners involving 17 European and 3 Australian organisations; 33 deliverables; 10 workshops and 1 conference.
3. Dissemination and Communication: A total of 107 journal/conference papers or technical reports, 35 open lectures and 16 presentations at conferences/workshops were delivered. The estimated number of persons reached is: 9735 in the Science Community; 340 in Industry; 6,100 members of the General Public; 750 in the Civil Society category, 635 Policy Makers and 633606 via Social Media.
4. Skills development: 107 researchers and 49 professional staff from 17 organisations have gained knowledge in wildfire issues. 21 PhD students were associated with GEO-SAFE, 17 of which undertook secondments. Of these, 6 have completed their PhDs, 2 have submitted their thesis and 9 are continuing their doctoral studies. All researchers gained invaluable experience and knowledge through close interaction with end-users facilitating a focused targeting of their research, thereby improving impact. The end-users have also developed an understanding of the capabilities of advanced modelling tools and assisted in their development.
Neural network models were developed that demonstrated better predictive accuracy and map quality for prediction of daily maps of wildfire burn probability. A good understanding was obtained on human behaviour in response to wildfires of wildfire evacuation. A large scale pedestrian evacuation model was integrated with traffic and wildfire spread models. Efficient algorithms were designed for solving general problems such as fuel management, evacuation, firebreak locating and firefighting. These developments have progressed beyond the state of the art, allowing decision makers and practitioners to better plan prescribed burns, identify location for firebreaks, allocate resources for initial attack and the protection of life and property.
A significant long-term GEO-SAFE impact has been the establishment of close contact between wildfire researchers from a wide range of disciplines and end-users of the research, the firefighters and incident commanders from, Australia, France, Italy, Netherlands, Spain, Switzerland, and the UK. Collaboration between these researchers and end-users will continue beyond the life of the project. The knowledge and data produced by GEO-SAFE (107 publications) will also have an impact on the rapidly developing area of wildfire research. The most important long-lasting impact of the GEO-SAFE project is the number of young scientists it has nurtured - 21 doctoral students. These publications, their wide spread dissemination and the next generation of scientists that GEO-SAFE has nurtured will open new lines of research and help to improve life, property and environmental safety into the future.
Economic: Advances in modelling technology achieved by GEO-SAFE will contribute to saving financial resources for public firefighting organisations in Europe and elsewhere and will contribute to reducing losses due to wildfires.
Practitioners and professional services: Contributions are provided to improve management processes, firefighter training programmes and upskilling for wildfire control. This is being achieved through knowledge exchange between firefighters and researchers.
Societal: The scientific knowledge gained and the modelling tool sets developed - once fully exploited via IP in future projects - will ultimately safeguard human lives, the infrastructure, livestock, the natural environment, and other natural resources.
Neural network integrates wildfire predictions&historical data to generate wildfire probability maps
Taking pictures with a drone to produce 3D images for locating firebreaks
Integration of [pedestrian, traffic and wildfire models applied to wildfire evacuation - urbanEXODUS
Data collection exercise in Italy on behavioural itineraries when people evacuate from a wildfire
Maps of vegetation age before/after applying prescribed burns in red dots obtained from a model
Prediction of bushfire occurrence rating maps for Victoria based on models developed in GEO-SAFE
Integration of pedestrian, traffic and wildfire models applied to wildfire evacuation by urbanEXODUS
Investigating the impact on vegetation at different times after a fire using satellite images
Graph theoretical application developed to evaluate the impact of firebreak locations on burn risk