Progress in the remote sensing of veld fire occurrence and detection: A review

Abstract

Our research provides a detailed overview of the progress in remotely sensed fire monitoring techniques, which have been developed and employed for fire occurrence and detection. Our overview is provided from a literature search of English-peer reviewed articles, conference proceedings and scientific book sections published between the periods of 1980 and 2019. Literature reveals that historically, fire detection through remotely sensed techniques has mainly occurred through ground-based, airborne and satellite systems. Mathematical models, such as decision tree models, Species Distribution Models, Dynamic Global Vegetation Models and Global Climate Models, have also been employed alongside satellite systems to facilitate a greater understanding of fire dynamics and its susceptibility to changes in ecological and climatic variables. Fire frequency and severity are known to be influenced by atmospheric conditions, fuel load and ignitions. However, the literature suggests that targeting inappropriate wildfires with these techniques may still result in wildfires outside of the natural fire regime. Most studies regarding fire occurrence and/or monitoring focus on satellite-based techniques as they provide the greatest coverage of wildfires at varying spatial and temporal resolutions depending on the sensor used. Satellite systems are advantageous for fire monitoring as they provide extensive coverage inexpensively. Finally, fire occurrence is explicitly influenced by moisture-limited climatic conditions and/or fuel load in the form of leaf-litter or water-stressed plants.