03013naa a2200373 a 450000100080000000500110000800800410001902400410006010000200010124501170012126000090023852019280024765000140217565300170218965300280220665300150223470000150224970000230226470000220228770000130230970000200232270000270234270000170236970000150238670000190240170000180242070000190243870000240245770000200248170000250250170000170252670000210254377300750256421591732023-12-05       2022    bl uuuu     u00u1 u    #d7 ahttps://doi.org/10.1002/cli2.192DOI1 aANDERSON, L. O.  aAn alert systemfor Seasonal Fire probability forecast for South American Protected Areas.h[electronic resource]  c2022  aTimely spatially explicit warning of areas with high fire occurrence probability is an important component of strategic plans to prevent andmonitor fires within South American (SA) Protected Areas (PAs). In this study, we present a five-level alert system,which combines both climatological and anthropogenic factors, the two main drivers of fires in SA. The alert levels are: High Alert, Alert, Attention, Observation and Low Probability. The trend in the number of active fires over the past three years and the accumulated number of active fires over the same period were used as indicators of intensification of human use of fire in that region, possibly associated with ongoing land use/land cover change (LULCC). An ensemble of temperature and precipitation gridded output from the GloSea5 Seasonal Forecast System was used to indicate an enhanced probability of hot and dry weather conditions that combined with LULCC favour fire occurrences. Alerts from this system were first issued in August 2020, for the period ranging from August to October (ASO) 2020. Overall, 50% of all fires observed during the ASO 2017?2019 period and 40% of the ASO 2020 fires occurred in only 29 PAs were all categorized in the top two alert levels. In categories mapped as High Alert level, 34% of the PAs experienced an increase in fires compared with the 2017?2019 reference period, and 81% of the High Alert false alarm registered fire occurrence above the median. Initial feedback from stakeholders indicates that these alerts were used to inform resource management in some PAs. We expect that these forecasts can provide continuous information aiming at changing societal perceptions of fire use and consequently subsidize strategic planning andmitigatory actions, focusing on timely responses to a disaster risk management strategy. Further research must focus on the model improvement and knowledge translation to stakeholders.  aWildfires  aConservation  aDisaster risk reduction  aMitigation1 aBURTON, C.1 aREIS, J. B. C. dos1 aPESSÔA, A. C. M.1 aBETT, P.1 aCARVALHO, N. S.1 aSILVA JUNIOR, C. H. L.1 aWILLIAMS, K.1 aSELAYA, G.1 aARMENTERAS, D.1 aBILBAO, B. A.1 aXAUD, H. A. M.1 aRIVERA-LOMBARDI, R.1 aFERREIRA, J. N.1 aARAGÃO, L. E. O. C.1 aJONES, C. D.1 aWILTSHIRE, A. J.  tClimate Resilience and Sustainabilitygv. 1, n. 1, p. 1-19, Feb. 2022.