http://repositorio.unb.br/handle/10482/37878| Fichier | Description | Taille | Format | |
|---|---|---|---|---|
| 2005_WalterNascimentoNeto.pdf ???org.dspace.app.webui.jsptag.ItemTag.accessRestricted??? | 5,32 MB | Adobe PDF | ???org.dspace.app.webui.jsptag.ItemTag.restrito??? |
| Titre: | Modelagem ecológica de queimadas usando inteligência artificial |
| Auteur(s): | Nascimento Neto, Walter |
| Orientador(es):: | Miranda, Heloisa Sinátora |
| Assunto:: | Queimada - Cerrados Inteligência artificial Sistemas especialistas (Computação) Raciocínio qualitativo Fogo - Cerrados |
| Date de publication: | 26-mai-2020 |
| Data de defesa:: | 2005 |
| Référence bibliographique: | NASCIMENTO NETO, Walter. Modelagem ecológica de queimadas usando inteligência artificial. 2005. 315 f., il. Tese (Doutorado em Ecologia)—Universidade de Brasília, Brasília, 2005. |
| Abstract: | The Cerrado is one of the best exemples of a Brazilian biome with high fire dependence. As a savanna vegetation, the Cerrado is fire-prone and the effects of fire will depends on several physical, chemical and ecological factors. The use of many types of models in ecology has facilitated the comprehension of ecosystems processes. Artificial Intelligence tries to mimic the way how humans think by means of reasoning systems based on computer programming logic. These systems are able to build models and computer simulations in areas of scientific interest, including ecological sciences. The aim of this study was to use Artificial Intelligence tools – the Expert Systems and the Qualitative Reasoning – to model the behavior of fire and its direct and indirect effects on vertical structure of vegetation and on the carbon and nitrogen balance in areas of campo sujo, one of the many physiognomic forms of cerrado. The Expert System built is able to interact with the user in order to give advice on how to manage fire in campo sujo areas based on the model output. This system uses the qualitative simulator GARP as an inference engine for varied conditions of temperature, amount of dead and live herbaceous biomass, fire line direction with reference to the wind and time since the last burn. The others two models were built using the Qualitative Reasoning approach based on the GARP structure and on the causal relationships between the many variables involved with the fire effects on cerrado. The first model simulates changes in the vegetation structure and in the main variables related to the carbon and nitrogen in areas of campo sujo burned at different frequencies. The second model makes the same type of simulation, considering the effect of precipitation in areas protected against fire. The output generated for each parameter in the three models was validated by a vast bibliography research about the effects of fire in Cerrado and other environments related. The qualitative modeling of ecological phenomena has proved to be strong and efficient to deal with situations in which there is little availability of empirical data although there were some difficulties in fitting it to ecological knowledge. |
| metadata.dc.description.unidade: | Instituto de Ciências Biológicas (IB) |
| Description: | Tese (doutorado)—Universidade de Brasília, Instituto de Ciências Biológicas, Programa de Pós-Graduação em Ecologia, 2005. |
| metadata.dc.description.ppg: | Programa de Pós-Graduação em Ecologia |
| Collection(s) : | Teses, dissertações e produtos pós-doutorado |
Tous les documents dans DSpace sont protégés par copyright, avec tous droits réservés.