A spatially explicit integrative model for estimating the risk of wildfire impacts in New-Caledonia

ABSTRACT

In this study, we present a spatially explicit bayesian model which is able to predict the distribution of potential fire ignition at a grid resolution of over the entire main island of New-Caledonia. This statistical model, when used in conjunction with a mechanistic fire model (FlamMap), is able to estimate the comparative and spatialized risk of fire impact on specific areas for specific issues such as biodiversity or erosion. The input data includes variables related to the physical environment such as the topography, climate, and some geographical indicators relating to human influences such as population density and type of land property.

Flowchart computation for the specific cell i. The grey rectangles denote the input data, the trapezoids denote the intermediate indices and the rounded corners rectangles denote the integrated wildfire risk indices. The connection legends indicates the corresponding equations and below certain nodes a drawing describes an example related to the main variables and the associated area used.

Graphical Abstract

Keywords:

• Fire risk modelling
• Bayesian network
• integrated modelling
• spatially explicit model
• decision making tool

Notes

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work has been funded by ANR BDIV-07-008 project INC and ‘Programme Pluri Formation Ingienérie des connaissances’ of University of New Caledonia. We are grateful to Hubert Géraux and the WWF for the insights about the fire impacts in New Caledonia and to Meteo France for providing the weather data. We also thanks New Caledonian governmental institutions for data support and Roberta Célimène and Tamara Somers for reviewing the text.