26 Aug 2021

26 Aug 2021

Review status: this preprint is currently under review for the journal ACP.

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

Sally S.-C. Wang1, Yun Qian1, L. Ruby Leung1, and Yang Zhang2 Sally S.-C. Wang et al.
  • 1Atmospheric Sciences and Global Change Division, Pacific Northwest National Laboratory, Richland, Washington, 99354, USA
  • 2Department of Civil and Environmental Engineering, Northeastern University, Boston, Massachusetts, 02115, USA

Abstract. Annual burned areas in the United States have increased twofold 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 and to understand the controlling factors of fire emissions. By comparing the predicted fire PM2.5 emissions from the interpretable ML model with the Global Fire Emissions Database (GFED) observations and predictions from process-based models in the Fire Modeling Intercomparison Project (FireMIP), the ML model is also used to diagnose the process-based models 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 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 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 coarse spatial resolutions of the processed-based models or atmospheric forcing data or 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 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 interpretable ML model with process-based fire models highlights their strengths and uncertainties for regional analysis and prediction and provides useful insights on future directions for model improvements.

Sally S.-C. Wang et al.

Status: open (until 07 Oct 2021)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse

Sally S.-C. Wang et al.

Data sets

Replication Data for: Interpretable machine learning prediction of fire emission and comparison with FireMIP process-based models Sally S.-C. Wang, Yun Qian, L. Ruby Leung, and Yang Zhang

Sally S.-C. Wang et al.


Total article views: 343 (including HTML, PDF, and XML)
HTML PDF XML Total Supplement BibTeX EndNote
270 70 3 343 16 2 3
  • HTML: 270
  • PDF: 70
  • XML: 3
  • Total: 343
  • Supplement: 16
  • BibTeX: 2
  • EndNote: 3
Views and downloads (calculated since 26 Aug 2021)
Cumulative views and downloads (calculated since 26 Aug 2021)

Viewed (geographical distribution)

Total article views: 344 (including HTML, PDF, and XML) Thereof 344 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
Latest update: 19 Sep 2021
Short summary
This study develops an interpretable machine learning (ML) model predicting monthly PM2.5 fire emission over the contiguous US at 0.25-degree resolution and compares the prediction skills of the ML and process-based models. The comparison facilitates attributions of model biases and better understanding of the strengths and uncertainties in the two types of models at regional scales, for informing future model development and their applications in fire emission projection.