Interpreting machine learning prediction of fire emissions and comparison with FireMIP process-based models

Abstract

Abstract. Annual burned areas in the United States have increased 2-fold during the past decades. With more large fires resulting in more emissions of fine particulate matter, an accurate prediction of fire emissions is critical for quantifying the impacts of fires on air quality, human health, and climate. This study aims to construct a machine learning (ML) model with game-theory interpretation to predict monthly fire emissions over the contiguous US (CONUS) and to understand the controlling factors of fire emissions. The optimized ML model is used to diagnose the process-based models in the Fire Modeling Intercomparison Project (FireMIP) to inform future development. Results show promising performance for the ML model, Community Land Model (CLM), and Joint UK Land Environment Simulator-Interactive Fire And Emission Algorithm For Natural Environments (JULES-INFERNO) in reproducing the spatial distributions, seasonality, and interannual variability of fire emissions over the CONUS. Regional analysis shows that only the ML model and CLM simulate the realistic interannual variability of fire emissions for most of the subregions (r>0.95 for ML and r=0.14∼0.70 for CLM), except for Mediterranean California, where all the models perform poorly (r=0.74 for ML and r<0.30 for the FireMIP models). Regarding seasonality, most models capture the peak emission in July over the western US. However, all models except for the ML model fail to reproduce the bimodal peaks in July and October over Mediterranean California, which may be explained by the smaller wind speeds of the atmospheric forcing data during Santa Ana wind events and limitations in model parameterizations for capturing the effects of Santa Ana winds on fire activity. Furthermore, most models struggle to capture the spring peak in emissions in the southeastern US, probably due to underrepresentation of human effects and the influences of winter dryness on fires in the models. As for extreme events, both the ML model and CLM successfully reproduce the frequency map of extreme emission occurrence but overestimate the number of months with extremely large fire emissions. Comparing the fire PM2.5 emissions from the ML model with process-based fire models highlights their strengths and uncertainties for regional analysis and prediction and provides useful insights into future directions for model improvements.

Projection of future fire emissions over the contiguous US using explainable artificial intelligence and CMIP6 models

“…Table 2 shows the predictors in the model with their sources and original spatial and temporal resolutions. Compared to the variables used in Wang et al (2022), this study used a subset of the variables that are accessible in the CMIP6 outputs.…”

“…Therefore, we include predictors representing the synoptic patterns driving fire emission variability, constructed using the singular value decomposition (SVD) method (Wang et al, 2022). The leading nodes of SVDs were identified for the three regions where large fires periodically occur, including northern California, southern Rocky Mountains, and the southeastern US, as defined in Wang et al (2021).…”

“…Finally, we compute the monthly standard deviation of the daily SVD time series for the first two SVD modes, representing the month-to-month variations of synoptic fluctuations and atmospheric instability. Note that we use U-and V-wind at the surface rather than at 850 hPa used in Wang et al (2022) due to data availability in the CMIP6 archive. The detailed methods and discussions about the SVDs are provided in Wang et al (2021) and Wang et al (2022).…”

“…Note that we use U-and V-wind at the surface rather than at 850 hPa used in Wang et al (2022) due to data availability in the CMIP6 archive. The detailed methods and discussions about the SVDs are provided in Wang et al (2021) and Wang et al (2022).…”

A machine learning approach targeting parameter estimation for plant functional type coexistence modeling using ELM-FATES (v2.0)

“…XGBoost has been successfully applied in a variety of fields within Earth and Environmental Sciences, such as urban temperature emulation (Zheng et al, 2021b), wildfire burned area (Wang et al, 2021), and emissions prediction (Wang et al, 2022), flash flood risk assessment (Ma et al, 2021), and aerosol property estimation (Zheng et al, 2021a, c).…”