Predicting fire behaviour (in-)directly via plant traits and Remote Sensing

Blauw, Luuk; Logtestijn, R.S.P. van; Bartholomeus, H.M.; Kooistra, L.; Cornelissen, Hans; Aerts, R.


Wildfires have been thoroughly investigated over the past decades and its drivers are well known. That knowledge is important to accurately define carbon emission models and conduct efficient fire management. Although both of these operate on a large scale, most fire research is based on smallscale laboratory experiments and occasionally large-scale field experiments. Yet, satellite RS images are already used to determine global and local fire frequencies, but satellites have a low spatial resolution and cannot be used in fire management. Therefore, airborne RS is used to accurately estimate plant traits and vegetation indices and that could be used to determine plant flammability traits. Accurate RS estimates of flammability traits could provide indirect predictions of fire behaviour. In dry and fire-prone areas, airborne RS could measure plant traits and estimate its sensitivity to fire and associated fire behaviour. We carried out 12 prescribed fires in young and old Scottish heathland, similar to Dutch heathlands, to determine the relation between plant traits, fire behaviour and remote sensing. Our results indicate a strong effect of plant traits on fire behaviour and that fire behaviour could be estimated based on the RS images.