Using machine learning to predict fire‐ignition occurrences from lightning forecasts

Coughlan, Ruth; Giuseppe, Francesca Di; Vitolo, Claudia; Barnard, Christopher; Lopez, Philippe; Drusch, Matthias

doi:10.1002/met.1973

Cited by 38 publications

(33 citation statements)

References 54 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Both LIW and LCC (>18 ms) lightning are rare with respect to typical lightning. Long-term efforts to develop a solid database of wildfires in the CONUS have significantly contributed to the identification of preferential meteorological conditions of LIW (e.g., [31,[68][69][70]). However, the detection of LCC (>18 ms) lightning by the typical Lightning Location Systems (LLS) is difficult due to the weak radiation emitted by the continuing phase of the discharge.…”

Section: Relationship Between Liw and Lcc (>18 Ms) Lightning Occurrencementioning

confidence: 99%

Meteorological Conditions Associated with Lightning Ignited Fires and Long-Continuing-Current Lightning in Arizona, New Mexico and Florida

Pérez‐Invernón

Huntrieser

Moris

2022

Fire

View full text Add to dashboard Cite

Lightning is the main precursor of wildfires in Arizona, New Mexico, and Florida during the fire season. Forecasting the occurrence of Lightning-Ignited Wildfires (LIW) is an essential tool to reduce their impacts on the environment and society. Long-Continuing-Current (LCC) lightning is proposed to be the main precursor of LIW. The long-lasting continuing current phase of LCC lightning is that which is more likely to ignite vegetation. We investigated the meteorological conditions and vegetation type associated with LIW and LCC lightning flashes in Arizona, New Mexico, and Florida. We analyzed LIW between 2009 and 2013 and LCC lightning between 1998 and 2014 and combined lightning and meteorological data from a reanalysis data set. According to our results, LIW tend to occur during dry thunderstorms with a high surface temperature and a high temperature gradient between the 700 hPa and the 450 hPa vertical levels for high-based clouds. In turn, we obtained a high lightning-ignition efficiency in coniferous forests, such as the ponderosa pine in Arizona and New Mexico and the slash pine in Florida. We found that the meteorological conditions that favor fire ignition and spread are more significant in Florida than in Arizona and New Mexico, while the meteorological conditions that favor the occurrence of LIW in Arizona and New Mexico are closely related with the meteorological conditions that favor high lightning activity. In turn, our results indicate high atmospheric instability during the occurrence of LIW. Our findings suggest that LCC (>18 ms) lightning tends to occur in thunderstorms with high relative humidity and ice content in the clouds, and with low temperature in the entire troposphere. Additionally, a weak updraft in the lower troposphere and a strong one in the upper troposphere favor the occurrence of LCC (>18 ms) lightning. We found that the meteorological conditions that favor the occurrence of LCC (>18 ms) lightning are not necessarily the preferential meteorological conditions for LIW.

show abstract

Section: Relationship Between Liw and Lcc (>18 Ms) Lightning Occurrencementioning

confidence: 99%

Meteorological Conditions Associated with Lightning Ignited Fires and Long-Continuing-Current Lightning in Arizona, New Mexico and Florida

Pérez‐Invernón

Huntrieser

Moris

2022

Fire

View full text Add to dashboard Cite

show abstract

“…In 2021, the availability of lightning forecasts at ECMWF led to the development of a data driven modeling framework to provide a quantitative measure to identify those lightning flashes that are conducive of fires (Coughlan et al., 2021). Different binary classifiers were developed to attribute a binary (“yes” or “no”) ignition label to lightning flashes.…”

Section: Introductionmentioning

confidence: 99%

“…Thus, they also posed the limit to what achievable by those models when applied to real events. When the models were employed to predict 147 real cases collected through lookout towers and spotter aircraft with a manual cause‐classification for fire events in West Australia, the prediction accuracy of all binary classifiers decreased to just 50% (Coughlan et al., 2021). The comparison to real cases highlighted that the lack of skill in the lightning forecast could play a crucial role.…”

Section: Introductionmentioning

confidence: 99%

The Value of Probabilistic Prediction for Lightning Ignited Fires

Giuseppe

2022

Geophysical Research Letters

Self Cite

View full text Add to dashboard Cite

Human activity is widely recognized as the primary cause of wildfires ignition (Amatulli et al., 2007;Archibald et al., 2009;Ganteaume et al., 2013). Still lightning-ignited fires are not uncommon and the majority of burned areas in remote regions of Boreal and arctic biomes have this as origin (Granström, 1993;Krider et al., 1980). Lightning is a major cause of wildland fires in Canada (Hanes et al., 2019) and in Alberta, for example, it is responsible for igniting 45% of reported wildfires and 71% of area burned (Blouin et al., 2016). Similar statistics are reported in Australia (Dowdy & Mills, 2012).In recent years, efforts have been devoted to the forecast of lightning activity both using physical models of various complexity (

show abstract

“…Beyond lightning, GAMs have proven their capabilities for severe weather trends (Rädler et al, 2018). Similar machine learning approaches have been used to generate wildfire danger maps (Vitolo et al, 2020;Coughlan et al, 2021). Also the field of postprocessing numerical weather predictions leverages these approaches.…”

Section: Introductionmentioning

confidence: 99%

Amplification of Annual and Diurnal Cycles of Alpine Lightning

Simon

Mayr

Morgenstern

et al. 2021

Preprint

View full text Add to dashboard Cite

The response of lightning to a changing climate is not fully understood. Historic trends of proxies known for fostering convective environments suggest an increase of lightning over large parts of Europe. Since lightning results from the interaction of processes on many scales, as many of these processes as possible must be considered for a comprehensive answer. Recent achievements of decade-long seamless lightning measurements and hourly reanalyses of atmospheric conditions including cloud micro-physics combined with flexible regression techniques have made a reliable reconstruction of lightning down to its seasonally varying diurnal cycle feasible. To include a large variety of land-cover, topographical and atmospheric circulation conditions, the European Eastern Alps and their surroundings are our reconstruction region over a period of four decades. The most intense changes occurred over the high Alps where lightning activity doubled in the past decade compared to the 1980s. There, the lightning season reaches a higher maximum and starts one month earlier, likely due to the earlier snow melt. Diurnally, the peak is up to 50% stronger with more lightning strikes in the afternoon and evening hours. Signals along the southern and northern alpine rim are similar but weaker whereas the flatlands north of the Alps have no significant trend.

show abstract

Using machine learning to predict fire‐ignition occurrences from lightning forecasts

Cited by 38 publications

References 54 publications

Meteorological Conditions Associated with Lightning Ignited Fires and Long-Continuing-Current Lightning in Arizona, New Mexico and Florida

Meteorological Conditions Associated with Lightning Ignited Fires and Long-Continuing-Current Lightning in Arizona, New Mexico and Florida

The Value of Probabilistic Prediction for Lightning Ignited Fires

Amplification of Annual and Diurnal Cycles of Alpine Lightning

Contact Info

Product

Resources

About