Researchers from Tomsk Polytechnic University together with colleagues from Russian universities propose a new concept for estimating the probability of wildfires caused by people. They develop mathematical models that make it possible to predict such fires based not only on statistics but also on a number of other factors – the characteristics of forest species, the physics of fire spreading, and anthropogenic activity in a particular area. The researchers chose two villages in the vicinity of the Lake Baikal – Gorychinsk and Khuramsha as areas for testing their models. The test of the models in this area is planned for 2019.
The research project is supported by the Russian Foundation for Basic Research (grant No. 17-29-05093).
The project leader Nikolai Baranovskiy, associate professor at the TPU Butakov Research Center told TASS:
‘The anthropogenic factor is one of the crucial wildfire factors. At the same time, it is the most difficult factor to predict and to estimate. Now specialists take into account only the statistics of fires. If we conditionally divide the forest territory into large areas, we may not find information in every particular point of the area. There may not be fires there. It does not mean that there is no anthropogenic factor in this area and fires are impossible in the future there. Therefore, we propose such a term as virtual – possible – number of wildfires which is calculated via mathematical models. This data mapped can be used for estimating fire-hazardous situations.’
The researchers are developing mathematical models for three types of objects, i.e. point, linear, and area. For example, campgrounds and settlements refer to point objects, roads are linear objects.
‘On such territories, the anthropogenic load is distributed differently. At the point objects, it is radial around the settlement, along the road it is linear. We still take the statistics of wildfires on a particular area as a basis, then pure mathematics comes into play which, according to the probability theory and mathematical statistics, helps to determine the likelihood of a fire at a particular point. The mathematical models take into account trees and characteristics of soil in the area. These are parameters to affect the physics of fire spreading. For example, coniferous forests burn much more often than mixed ones, and even more than deciduous,’ explained the project leader.
The data on tree species are delivered by images from the Landsat satellite and mathematically processed.
‘We started this project in 2018 and we have already developed a number of models. Our colleagues from Ulan-Ude suggested two settlements in the area of the Lake Baikal – the villages of Goryachnisk and Khuramsha – as experimental sites. There are stationary measurement complexes which allow estimating the peculiarities of soil and meteorological conditions in a certain area. The most important is the fact that they are located in the Baikal basin that is a unique natural object requiring special attention,’ says the scientist.
The project should result in a geoinformation system prototype. It will be an electronic map indicating areas where, taking into account the anthropogenic load, a fire is more likely to occur. The scientists expect that in the future such information will help to strengthen the monitoring of specific areas of the forest and responding more quickly to emergencies.
TPU researchers are conducting the project jointly with colleagues from ITMO University, Gorno-Altaysk State University, the Institute of Physics Materials Science SB RAS (Ulan-Ude), the Institute of General and Experimental Biology SB RAS (Ulan-Ude), and V.R. Filippov Buryat State Agricultural Academy (Ulan-Ude).
For the last 20 years, the forest territories around Baikal have reduced by 5 %. Meantime, forests in the basin of the lake annually suffer from wildfires. The Republic Agency of Forests of Buryatia calls the anthropogenic factor as the main reason for wildfires. Ignites occur at picnic sites, during a hunt, from an abandoned burning match, an extinguished cigarette, due to glass fragments, in sunny weather focusing the sun's rays as incendiary lenses.