Wildland fires are becoming more destructive and costly in the United States, posing increased environmental, social, and economic threats to fire-prone regions. Quantifying current wildfire risk by considering a wide range of multi-scale, and multi-disciplinary variables such as socio-economic and biophysical indicators for resiliency and mitigation measures, deems inherently challenging. To systematically examine wildfire threats amongst humans and their physical and social environment on multiple scales, a livelihood vulnerability index (LVI) analysis can be employed. Therefore, we produce a framework needed to compute the LVI for the top 14 American States that are most exposed to wildfires, based on the 2019 Wildfire Risk report of the acreage size burnt in 2018 and 2019: Arizona, California, Florida, Idaho, Montana, Nevada, New Mexico, Oklahoma, Oregon, Utah, Washington, and Wyoming. The LVI is computed for each State by first considering the State’s exposure, sensitivity, and adaptive capacity to wildfire events (known as the three contributing factors). These contributing factors are determined by a set of indicator variables (vulnerability metrics) that are categorized into corresponding major component groups. The framework structure is then justified by performing a principal component analysis (PCA) to ensure that each selected indicator variable corresponds to the correct contributing factor. The LVI for each State is then calculated based on a set of algorithms relating to our framework. LVI values rank between 0 (low LVI) to 1 (high LVI). Our results indicate that Arizona and New Mexico experience the greatest livelihood vulnerability, with an LVI of 0.57 and 0.55, respectively. In contrast, California, Florida, and Texas experience the least livelihood vulnerability to wildfires (0.44, 0.35, 0.33 respectively). LVI is strongly weighted on its contributing factors and is exemplified by the fact that even though California has one of the highest exposures and sensitivity to wildfires, it has very high adaptive capacity measures in place to withstand its livelihood vulnerability. Thus, States with relatively high wildfire exposure can exhibit relatively lower livelihood vulnerability because of adaptive capacity measures in place. On the other hand, States can exhibit a high LVI (such as Arizona) despite having a low exposure, due to lower adaptive capacities in place. The results from this study are critical to wildfire managers, government, policymakers, and research scientists for identifying and providing better resiliency and adaptation measures to support the American States that are most vulnerable to wildfires.

Wildfire indices are used globally to quantify and communicate a wide range of fire characteristics, including fire danger and fire behaviour. Wildfire terminologies, definitions and variables used to compute fire indices vary broadly. This makes it difficult to compare them under a common framework for regional assessment and for future improvements under changing climate and land-use/land-cover conditions. This paper reviews 24 fire indices used worldwide and proposes a simple framework within which they can be classified based on constitutive inputs used for their computation. We differentiate between constitutive inputs that are raw or directly measurable variables such as fuel, weather and topography (referred to as Level 1 inputs) and calculated constitutive inputs such as fuel moisture (as a function of ecology and hydrometeorology); fire behaviour (as a function of spread, energy, and ignition); and dynamic meteorology. These six calculated constitutive inputs are referred to as Level 2 inputs. Based on this classification, our findings indicate that the Burning Index from the United States National Fire Danger Rating System (NFDRS) and the Fire Weather Index from the Canadian Forest Fire Danger Rating System (CFFDRS), used by many countries worldwide, utilize the most comprehensive set of Level 2 inputs. In addition, the Level 2 input that is most frequently used by all fire indices is fuel moisture as a function of hydrometeorology and the least integrated input is that of fire ignition. We further group the fire indices in three types: fire weather, fire behaviour, and fire danger indices, according to the open literature definition of their predictant outputs and examine the specific constitutive inputs used in their computation. Most fire indices are based on Level 2 inputs (which use Level 1 inputs) and are predominantly fire danger and fire behaviour indices. This is followed by fire indices that use a combination of both Level 1 and Level 2 inputs, separately and are mostly fire danger indices. Only a few fire indices are computed solely with raw Level 1 inputs and are mainly fire behaviour indices. Providing a comprehensive view of the existing wildfire indices’ utilization and computational structure is expected to be a helpful resource for wildfire researchers and operational experts worldwide. 2