Statistical Learning of California’s Greenhouse Gas Emissions Based on Atmospheric Models

Statistical Learning of California’s Greenhouse Gas Emissions Based on Atmospheric Models

Seminar Abstract 

Since California’s legislation on climate change (AB-32), which requires greenhouse gas (GHG) emissions to be reduced to 1990 levels by 2020, California has established a more ambitious plan aiming for 80% emission reduction below 1990 levels by 2050. It is essential to verify the implementation of the progressive targets through science-based independent research. Recent studies on carbon dioxide (CO2) emissions including work from LBNL suggest anthropogenic CO2 from California's official inventory is accurate within ~10%. However, methane (CH4) and nitrous oxide (N2O) emissions are more uncertain than CO2. This talk describes statistical learning methods developed to quantify GHG emissions based on atmospheric models and measurements focusing on CH4 and N2O. In recent full annual analyses, the LBNL team finds that the CH4 and N2O estimates by the state inventory are low by 50 – 100% suggesting that these two major non-CO2 GHGs likely account for ~20% of California’s total GHG. These results further indicate that non-CO2 GHGs are important in reducing emissions to 20% of 1990 levels by 2050. In closing, this talk identifies research areas where more comprehensive efforts are needed to accurately assess California’s progress in meeting its bold climate goals and offers research directions for future GHG emission monitoring.

 

Seminar Speaker(s) 

Seongeun Jeong
Principal Scientific Engineering Associate, Sustainable Energy Systems Group, Sustainable Energy & Environmental Systems Department, Energy Analysis & Environmental Impacts Division

Date 

Mar 14, 2019 -
12:00pm to 1:00pm

Location 

90-3122