Modelling the Ecological Vulnerability to Forest Fires in Mediterranean Ecosystems Using Geographic Information Technologies
Modelling the Ecological Vulnerability to Forest Fires in Mediterranean Ecosystems Using Geographic Information Technologies
复制标题
DOI：
10.1007/s00267-012-9933-3
复制
发表时间：
2012-12-01
影响因子：
3.5
通讯作者：
Vallejo, Ramon V.
中科院分区：
文献类型：
--
作者：
Duguy, Beatriz;Antonio Alloza, Jose;Vallejo, Ramon V.
森林火灾是地中海区域生态系统和地貌变化的主要驱动力。环境特征和植被是估计生态对火灾脆弱性的关键因素;定义为生态系统易受火灾不利影响和无法科普火灾不利影响的程度（如果发生火灾）。鉴于预测的该区域的气候变化，迫切需要验证评估这种脆弱性的空间明确工具，以支持设计新的防火和恢复战略。这项工作提出了一个创新的基于GIS的建模方法来评估生态系统的生态脆弱性火灾，考虑其主要组成部分（土壤和植被）和不同的时间尺度。评价分为三个阶段：短期（侧重于土壤退化风险）、中期（侧重于植被变化）以及短期和中期脆弱性的结合。该模型在两个地区实施：阿拉贡（西班牙东北部内陆）和瓦伦西亚（西班牙东部）。绘制了区域一级的生态火灾脆弱性地图。我们部分验证了模型在一个研究网站相结合的两种互补的方法，重点是测试模型的预测在三个生态系统，所有非常常见的火灾易发景观的西班牙东部：两个灌木丛和松树林的充分性。这两种方法都是基于模型的预测值与归一化植被指数（归一化植被指数），这被认为是一个很好的代理绿色生物量的比较。这两种方法表明，该模型的性能是令人满意的，当应用到三个选定的植被类型。
Forest fires represent a major driver of change at the ecosystem and landscape levels in the Mediterranean region. Environmental features and vegetation are key factors to estimate the ecological vulnerability to fire; defined as the degree to which an ecosystem is susceptible to, and unable to cope with, adverse effects of fire (provided a fire occurs). Given the predicted climatic changes for the region, it is urgent to validate spatially explicit tools for assessing this vulnerability in order to support the design of new fire prevention and restoration strategies. This work presents an innovative GIS-based modelling approach to evaluate the ecological vulnerability to fire of an ecosystem, considering its main components (soil and vegetation) and different time scales. The evaluation was structured in three stages: short-term (focussed on soil degradation risk), medium-term (focussed on changes in vegetation), and coupling of the short- and medium-term vulnerabilities. The model was implemented in two regions: Aragn (inland North-eastern Spain) and Valencia (eastern Spain). Maps of the ecological vulnerability to fire were produced at a regional scale. We partially validated the model in a study site combining two complementary approaches that focused on testing the adequacy of model's predictions in three ecosystems, all very common in fire-prone landscapes of eastern Spain: two shrublands and a pine forest. Both approaches were based on the comparison of model's predictions with values of NDVI (Normalized Difference Vegetation Index), which is considered a good proxy for green biomass. Both methods showed that the model's performance is satisfactory when applied to the three selected vegetation types.