Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Improve wildfire predictability driven by extreme water cycle with interpretable physically-guided ML/AI

Technical Report ·
DOI:https://doi.org/10.2172/1769720· OSTI ID:1769720
 [1];  [1];  [2]
  1. Lawrence Berkeley National Lab. (LBNL), Berkeley, CA (United States)
  2. Univ. of California, Irvine, CA (United States)
Focal Area(s): Predictive modeling using AI techniques and AI-derived model components; use of AI and other tools to design a prediction system comprising of a hierarchy of models (Primary); insight gleaned from complex data (both observed and simulated) using AI, big data analytics, and other advanced methods, including explainable AI and physics or knowledge-guided AI (Secondary). Science Challenge: Wildfires modify land surface characteristics, such as vegetation composition, soil and litter carbon stocks, and surface albedo, with significant consequences for the regional carbon cycle. For example, tropical regions (i.e., African and South America) are particularly vulnerable to wildfire, account for more than 80% of the global burned area, and emit ~1.4 PgC y-1 into the atmosphere together with other dust and aerosols that strongly affect regional climate.
Research Organization:
Artificial Intelligence for Earth System Predictability (AI4ESP) Collaboration (United States)
Sponsoring Organization:
USDOE Office of Science (SC), Biological and Environmental Research (BER)
OSTI ID:
1769720
Report Number(s):
AI4ESP--1154
Country of Publication:
United States
Language:
English