Wildfire spreading across the urban area: definition and mapping of vegetation connectivity

Climate change, increased frequency of natural disasters and their anomaly in the modern world demand review existing knowledge about environment and approaches for its management. In recent decades, one of the phenomena that humanity met with is urban wildfires. Experience has shown that existing approaches from forest fire management are not effective in a built-up environment. In practice, studying this phenomenon is limited due to lack of empirical knowledge.

To bridge this knowledge gap, the fire dynamic patterns in an urban area was reconstructed based on fire service data on the example of wildfire in Haifa (Israel) in November 2016. The results of reconstruction show that unlike forest fires, where fuel is almost unlimited around the ignition point, urban areas are sensitive to wildfire because of 'fire connectivity' through vegetation: fire moves from one green patch to another, bypassing nonflammable structures.

Although the role of urban vegetation in the spread of fires is obvious, it is difficult to define a suitable term in case of fire management. Today, cities and vegetation are studied together in the urban planning and ecosystem services and have different definitions, depends on study propose: ‘Urban Green Spaces’, ‘Green Infrastructure’, ‘Urban Forestry’, etc. The closest term in fire management for vegetation interaction with structures is ‘Wildland Urban Interface’ (WUI). However, WUI considers vegetation surrounding the city and excludes urban green spaces. Since before this time the inner-city vegetation was not considered in the fire management and wasn’t estimated its essential parameters for this case, it does not have an exact term and definition. The reconstructed wildfire allows to analyze and define vegetation in context of fire connectivity.

The results show that wildfire spreads in the urban area through ember attacks. Meanwhile, the embers led to ignition only in some patches and only in a few cases the energy amount was enough for new embers emission. Thus, the fire vegetation connectivity in urban areas can be defined by its ability to ignite and reproduce new embers.

To support fire management and risk assessment in urban areas, it is important to map vegetation based on its ignitability and potential energy emission. The task is complicated due to the uncertainty of surface fuel (including both litter and human waste). Citywide, detecting and monitoring such patches through field surveys is time-consuming work. To solve this objective by remote sensing technique, we analyzed series of Landsat 8 images for 2015 to identify spectral and temporal features of vegetation related to its flammability. The proposed approach supports estimation and mapping of vegetation connectivity in case of urban wildfire based on its multi-temporal spectral signature.