Team #14 // Fire Index

Executive Summary

Natural wild fires are extremely costly both financially and in terms of lives lost. We propose a Fire Index (FI) that can predict areas with a high risk of fire. The FI will be based on various data sources such as vegetation dryness, ground dryness, demographics, meteorological forecasts, climatological observations and several other parameters. A predictive FI has the potential to reduce the number and severity of fires, inform the population about hazards and reduce emergency management expenses. The potential value of our proposed FI is high both in economic and environmental terms as it meets the Climate Action and Life on Land Sustainable Development Goals.

Problem

Natural wild fires are caused by a combination of land management, human activities, cultural traditions, climate- and weather conditions. According to the European Forest Fire Information System for 2017[1], wildfires burned over 1.2 million ha of natural lands in the EU and killed 127 fire fighters and civilians. Estimated loss was around 10 billion euros caused by these fires and it is expected that many of Europe’s forests will be affected more frequently by extreme weather conditions such as drought, extended heat waves and strong winds as a consequence of climate change[1].

Target user / Customer

Short-term: Danish Metrological Institute (DMI), Danish Emergency Managements Agency.

Mid-term: National Meteorological and Hydrological Services (NMHS), Scandinavian Emergency Management Agencies, Agricultural knowledge center (SEGES), Insurance & Pension Industry organization (F&P), Vestas wind turbine solutions and services.

Long-term: European Organization for the Exploitation of Meteorological Satellites (EUMETSAT), European Environmental Agency (EEA).

Your solution and how the concept is feasible

A lot of parameters such as demographic data, land cover, ground elevation, soil type, soil moisture, dryness of vegetation, and meteorological parameters are important for predicting fire risk. Our initial solution will be to combine the Drought Index (DI) and dryness of vegetation. The DI is provided by DMI and dryness of vegetation can be obtained using remote sensing data, more specifically by using the Temperature Vegetation Dryness Index (TVDI). The TVDI index is estimated from Land Surface Temperature (LST) and Vegetation Index (NDVI) available from the Sentinel 3 satellite sensors. The combination of these two indexes will be an Improved Drought Index (IDI).

Below is first attempt at calibrating the DI with TVDI in Denmark, two days in the summer 2018, and the result looks very promising. Here we get very detailed information about sandy/rocky costlines, artificial irrigation of fields, indication of soil types and forest areas that have not necessarily depleted their water supply in the ground as first indicated in the original DI.

This IDI is just a first step towards creating an operational FI. The next step is to create a model that predicts the fire risk using historical data of meteorological parameters, DI, TVDI and soil moisture. These inputs will then be used to predict historical fire locations. The historical records of reported fires in Denmark are registered in a statistics bank called ODIN and managed by the Danish Emergency Management Agency. The final FI will offer detailed information and global overview in one solution, with a user-friendly interface and improved predictions of fire hazards in near real time data delivery.

Four unique value propositions

  1. Provide increase predictability of natural wild fires to all EU countries, including those not covered by geostationary satellites.
  2. Crowd location services to identify increased fire hazard locations in near real time (e.g. Major gathering i.e. concerts, mid summer night, etc).
  3. Customized user interface tailored to fit specific customer systems and workflows.
  4. Configurable system services and data sources based on availability in individual countries.


Financials

To become a key provider of fire risk predictions in the EU, and possibly the world, it is critical to gain credibility from the other EU countries. We will demonstrate that our solution can outperform the market.

1. Partnership

It is often the government that provides information about fire risks, in many scenarios it is the metrological institutes (NMHS National Meteorological and Hydrological Services). We will not compete against the NMHS but instead seek to obtain partnerships with these entities, as it will increase credibility and rubber stamp our product/service. We will start with Denmark, as we have direct contact with the Danish Metrological Institute (DMI). The Emergency Management Agency (EM) is our potential customer and we will seek a close collaboration with them during development, as they hold knowledge of key requirements.

2. Fundraising
In order to get the project of the ground, initial funding is needed. There are several options, such as Innobooster, funds and investors. Forsikring & Pension (F&P) is an industry organization for insurance and pension companies in Denmark who might have an interest in funding this project, as natural forest fires have director impact on costs to insurance companies.

3. Credibility
Gaining credibly is needed, not only to succeed in Denmark and the rest of EU, but also to ensure fundraising. ESA and Copernicus are major rubber stamps that can be obtained as part of Copernicus Accelerator Program and the new ESA BIC starting in Denmark 2020. One key milestone is to go live in Denmark by 1. March 2021 when the potential dry season starts. A financial scenario includes a subscription-based setup with the Danish EM, where DMI delivers the fire risk map and warns of fire risks.

4. Scandinavia
When our solution is successfully operational in Denmark, we will approach the Nordic meteorological institutes, SMHI, MET Norway and FMI. A successful case in Denmark, using DMI as a channel, will ease entry to the Nordic market, as there is close collaboration between all Nordic NMHSes. At the same time, we will approach other companies in different areas, as the fire index solution will result in several sub products with relevance for other parties, such as the agricultural knowledge center (SEGES), Insurance & Pension Industry organization (F&P) and Vestas wind turbine solutions and services.

5. EU
With successful implementations in Nordic countries, our service and company will stand strong with a sustainable solution that can cover all EU countries. While implementing our solution in collaboration with interested north EU countries that are not covered by geostationary satellites, we will begin adapting GEO data to obtain much higher product quality. We will be in direct competitions with Land SAF, but with increased prediction of natural wild fires. This may provide a unique opportunity to sign a contract with EUMETSAT, to further distribute the fire index to all EU metrological institutes. In addition, our service can be used as a future information tool about fire risk situation to policy makers in EU if EEA (European Environmental Agency) could be convinced of the value of our services.

Impact

Global warming can have an increased impact on the number of natural forest fires, and forest fires also contribute directly to the effects of global warming – in other words, this is a self-powering effect that directly addresses the Sustainable Development Goals SDG-13, Climate Action. Furthermore it addresses SDG-15 Life on Land, as forest fires have a negative effect on biodiversity. The solution we are suggesting will lead to a reduction in global wildfires and fatal disasters, more sustainable ecosystems, and improved agriculture.

Links for more information

[1] EU report 2017: forest_fires-180920_da

[2] DMI new fire index model: summary

[3] Climate change and natural disasters: climate-change-will-bring-more-frequent-natural-disasters-weigh-on-economic-growth

[4] Climate change and wildfires: wildfires


Team

Our team consists of experts in the fields of data science, climatology and remote sensing which are exactly the fields required in order to develop the Fire Index. Dan has a PhD in Machine Learning and several years experience with modeling and applied mathematics. Thomas works at SoftSingularity and has a background from DMI and several years experience with climatological and forecasts data. Georgios works at Sandholt and has several years experience with remote sensing and earth observations in general.

https://softsingularity.com

https://www.sandholt.eu

Group Members


Credit

Forest trees by webstockreview.net
Fundraising icon by MRFA
Credibility icon by Scott Baker

Share

Share on facebook
Share on twitter

More related