Publicação
Can fire spread simulations contribute to support decisions in a fire suppression context ? An evaluation using MaxEnt, FARSITE and satellite active fire data
| dc.contributor.advisor | Sá, Ana Cristina Lopes de | |
| dc.contributor.advisor | Pereira, José Miguel Cardoso | |
| dc.contributor.author | Baptista, Rafael Coelho | |
| dc.date.accessioned | 2019-09-18T10:43:46Z | |
| dc.date.available | 2019-09-18T10:43:46Z | |
| dc.date.issued | 2018 | |
| dc.description | Mestrado em Engenharia Florestal e dos Recursos Naturais - Instituto Superior de Agronomia / UL | pt_PT |
| dc.description.abstract | This work is of an exploratory nature and describes and evaluates a method to simulate fire spread which, in an operational context, has the potential to be used as a decision-support tool for fire management and suppression. The use of fire spread models has, for the most part, followed a deterministic approach, which does not account for predictions uncertainty. However, fire spread models are subject to assumptions and limitations that inherently produce errors during simulations and so should be integrated in the simulations themselves. For that matter uncertainty was propagated through Farsite fire behavior model by randomly defining 100 different independent combinations of some of the most important input variables. The simulations were run with three different fuel maps, one standardized and two customized. For the evaluation of the fire spread predictions a qualitative and a quantitative analysis were made. Both analyses used MaxEnt derived reference perimeters, and active fire data was used on the qualitative analyses to add temporal depth to the evaluation. Results showed that uncertainties in wind speed and direction, location of ignitions (spatial and temporal), fuel model assignment and typology may have major impact on prediction accuracy. Overall, fuel models presented better results when compared with the standard model and generally showed higher Kappa and burned class agreement values and lower omission errors. This thesis suggests that this method has major potential to optimize fuel management practices, especially if simulations are run with fuel maps derived Portuguese landcover maps | pt_PT |
| dc.description.version | N/A | pt_PT |
| dc.identifier.citation | Baptista, R.C. - Can fire spread simulations contribute to support decisions in a fire suppression context ? An evaluation using MaxEnt, FARSITE and satellite active fire data. Lisboa: ISA, 2018, 65 p. | pt_PT |
| dc.identifier.tid | 203092384 | |
| dc.identifier.uri | http://hdl.handle.net/10400.5/18338 | |
| dc.language.iso | eng | pt_PT |
| dc.publisher | ISA/UL | pt_PT |
| dc.subject | probabilistic simulations | pt_PT |
| dc.subject | active fire data | pt_PT |
| dc.subject | MaxEnt | pt_PT |
| dc.title | Can fire spread simulations contribute to support decisions in a fire suppression context ? An evaluation using MaxEnt, FARSITE and satellite active fire data | pt_PT |
| dc.type | master thesis | |
| dspace.entity.type | Publication | |
| rcaap.rights | openAccess | pt_PT |
| rcaap.type | masterThesis | pt_PT |