Artificial neural networks (ANNs) show a significant ability to discover patterns in data that are too obscure to go through standard statistical methods. Data of natural phenomena usually exhibit significantly unpredictable non-linearity, but the robust behavior of a neural network makes it perfectly adaptable to environmental models such as a wildland fire danger rating system. These systems have been adopted by many developed countries that have invested in wildland fire prevention, and thus civil protection agencies are able to identify areas with high probabilities of fire ignition and resort to necessary actions. Since one of the drawbacks of ANNs is the interpretation of the final model in terms of the importance of variables, this article presents the results of sensitivity analysis performed in a back-propagation neural network (BPN) to distinguish the influence of each variable in a fire ignition risk scheme developed for Lesvos Island in Greece. Four different methods were utilized to evaluate the three fire danger indices developed within the above scheme; three of the methods are based on network's weights after the training procedure (i.e., the percentage of influence-PI, the weight product-WP, and the partial derivatives-PD methods), and one is based on the logistic regression (LR) model between BPN inputs and observed outputs. Results showed that the occurrence of rainfall, the 10-h fuel moisture content, and the month of the year parameter are the most significant variables of the Fire Weather, Fire Hazard, and Fire Risk Indices, respectively. Relative humidity, elevation, and day of the week have a small contribution to fire ignitions in the study area. The PD method showed the best performance in ranking variables' importance, while performance of the rest of the methods was influenced by the number of input parameters and the magnitude of their importance. The results can be used by local forest managers and other decision makers dealing with wildland fires to take the appropriate preventive measures by emphasizing on the important factors of fire occurrence.

Prevention is one of the most important stages in wildfire and other natural hazard management regimes. Fire danger rating systems have been adopted by many developed countries dealing with wildfire prevention and pre-suppression planning, so that civil protection agencies are able to define areas with high probabilities of fire ignition and resort to necessary actions. This present paper presents a fire ignition risk scheme, developed in the study area of Lesvos Island, Greece, that can be an integral component of a quantitative Fire Danger Rating System. The proposed methodology estimates the geo-spatial fire risk regardless of fire causes or expected burned area, and it has the ability of forecasting based on meteorological data. The main output of the proposed scheme is the Fire Ignition Index, which is based on three other indices: Fire Weather Index, Fire Hazard Index, and Fire Risk Index. These indices are not just a relative probability for fire occurrence, but a rather quantitative assessment of fire danger in a systematic way. Remote sensing data from the high-resolution QuickBird and the Landsat ETM satellite sensors were utilised in order to provide part of the input parameters to the scheme, while Remote Automatic Weather Stations and the SKIRON/Eta weather forecasting system provided real-time and forecasted meteorological data, respectively. Geographic Information Systems were used for management and spatial analyses of the input parameters. The relationship between wildfire occurrence and the input parameters was investigated by neural networks whose training was based on historical data.

Fire, a frequent disturbance in the Mediterranean, affects pollinator communities. We explored the response of major pollinator guilds to fire severity, across a fire‐severity gradient at different spatial scales. We show that the abundance of all pollinator groups responded to fire severity, and that bees and beetles showed in addition a significant species‐diversity response. Bees, sawflies, and wasps responded to fire severity at relatively small spatial scales (250–300 m), whereas flies and beetles responded at larger spatial scales. The response of bees, sawflies, and wasps was unimodal, as predicted by the intermediate disturbance hypothesis, whereas flies and beetles showed a negative response. A possible explanation is that the observed patterns (spatial scale and type of response) are driven by taxa‐specific ecological and life‐history traits, such as nesting preference and body size, as well as the availability of resources in the postfire landscape. Our observational study provides an insight into the effect of fire severity on pollinators. However, future research exploring the explicit link between the pre‐ and postfire landscape structure and pollinator traits and responses is required for further establishment and understanding of cause–effect relationships.

Empirical allometric equations relating biomass of aboveground components to dendrometric variables for Pinus brutia Ten. trees are derived in this paper.

No abstract