Fire weather and likelihood: characterizing climate space for fire occurrence and extent in Puerto Rico
Assessing the relationships between weather patterns and the likelihood of fire occurrence in the Caribbean has not been as central to climate change research as in temperate regions, due in part to the smaller extent of individual fires. However, the cumulative effect of small frequent fires can shape large landscapes, and fire-prone ecosystems are abundant in the tropics. Climate change has the potential to greatly expand fire-prone areas to moist and wet tropical forests and grasslands that have been traditionally less fire-prone, and to extend and create more temporal variability in fire seasons. We built a machine learning random forest classifier to analyze the relationship between climatic, socio-economic, and fire history data with fire occurrence and extent for the years 2003–2011 in Puerto Rico, nearly 35,000 fires. Using classifiers based on climate measurements alone, we found that the climate space is a reliable associate, if not a predictor, of fire occurrence and extent in this environment. We found a strong relationship between occurrence and a change from average weather conditions, and between extent and severity of weather conditions. The probability that the random forest classifiers will rank a positive example higher than a negative example is 0.8–0.89 in the classifiers for deciding if a fire occurs, and 0.64–0.69 in the classifiers for deciding if the fire is greater than 5 ha. Future climate projections of extreme seasons indicate increased potential for fire occurrence with larger extents.
We thank Joel Figueroa and Iván Cruz from the Fire Department of Puerto Rico and Luis Rosa, Gary Votaw and Shawn Rossi from the National Weather Program, NOAA/National Weather Service, San Juan for providing data on fire occurrence. Eric Harmsen provided the initial processing of the daily wind speed data. This research was supported by grant DEB 1546686 from US National Science Foundation to the Institute for Tropical Ecosystem Studies, University of Puerto Rico, and to the International Institute of Tropical Forestry (IITF) USDA Forest Service, as part of the Luquillo Long-Term Ecological Research Program. The US Forest Service (Department of Agriculture) Research Unit, the Luquillo Critical Zone Observatory (EAR-1331841), and the University of Puerto Rico gave additional support. We thank Dr. Ariel E. Lugo for providing useful comments on an earlier version of the manuscript. All research at the International Institute of Tropical Forestry is done in collaboration with the University of Puerto Rico.
- Chawla NV (2005) Data mining for imbalanced datasets: an overview. In: Maimon O, Rokach L (eds) Data mining and knowledge discovery handbook. Springer, New York, pp 853–867Google Scholar
- Dwyer E, Grégoire J, Malingreau J (1998) A global analysis of vegetation fires using satellite images: spatial and temporal dynamics. Ambio 27:175–181Google Scholar
- Fernández-Delgado M, Cernadas E, Barro S, Amorim D (2014) Do we need hundreds of classifiers to solve real world classification problems? J Mach Learn Res 15:3133–3181Google Scholar
- Forster P, Ramaswamy V, Artaxo P, Berntsen T, Betts R, Fahey DW, Haywood J, Lean J, Lowe DC, Myhre G, Nganga J, Prinn R, Raga G, Schulz M, Van Dorland R (2007) Changes in atmospheric constituents and in radiative forcing. In: Solomon S, Qin D, Manning M, Chen Z, Marquis M, Averyt KB, Tignor M, Miller HL (eds) Climate change 2007: the physical science basis. Contribution of working group I to the 4th assessment report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, pp 129–234Google Scholar
- Gould WA, Alarcón C, Fevold B, Jiménez M, Martinuzzi S, Potts G, Quiñones M, Solórzano M, Ventosa E (2008) The Puerto Rico gap analysis project. Volume 1: land cover, vertebrate species distributions, and land stewardship. US Dep Agric For Serv Int Inst Trop For. IITF-GTR-39, 165 pp. Río Piedras, Puerto RicoGoogle Scholar
- Gregorutti B, Michel B, Saint-Pierre P (2017) Correlation and variable importance in random forests. Stat Comput 27(3):659–678Google Scholar
- Martinuzzi S, Gould WA, Ramos González OM (2007) Creating cloud-free Landsat ETM+ data sets in tropical landscapes: cloud and cloud-shadow removal. US Dep. Agric. For. Serv. Int. Inst. Trop. For. IITF-GTR-32, 12 pp. Río Piedras, Puerto RicoGoogle Scholar
- Méndez-Tejeda R, Santos-Corrada M, Ortiz-Morales S, Claudio-Vargas O (2015) Environmental and economic impact of Forest fires in Puerto Rico 2013-2014. Open J For 5:353Google Scholar
- van der Werf GR, Randerson JT, Giglio L, Gobron N, Dolman AJ (2008) Climate controls on the variability of fires in the tropics and subtropics. Glob Biogeochem Cycles 22:GB3028Google Scholar
- Yu YP, Omar R, Harrison RD, Sammathuria MK, Nik AR (2011) Pattern clustering of forest fires based on meteorological variables and its classification using hybrid data mining methods. J Comput Biol Bioinforma Res 3:47–52Google Scholar