CISL Seminar- Chemical Data Assimilation for Atmospheric Composition and Climate Science

12/07/2015 - 10:00am
Foothills Laboratory 2 - Main Auditorium

Jérôme Barré
NCAR ACOM

Data assimilation of the atmospheric chemical composition has various applications such as air quality forecasting, global model sensitivity analysis, long-range transport of pollution and emission estimation. Data assimilation finds its roots on in numerical weather prediction. Although data assimilation science can be applied to many fields of science, a particular current focus is application to Earth system components including the atmosphere, ocean, land, ice and rivers. Data assimilation techniques for atmospheric chemistry have been developed over the last two decades.

In this talk, I will present results from my research on chemical data assimilation. Variational data assimilation of satellite tropospheric ozone retrievals in a regional scale air quality model (MOCAGE) show the importance of boundary conditions, especially the constraint of stratospheric ozone on the tropospheric ozone estimate. Ensemble Kalman filter techniques will be also covered with a focus on global satellite carbon tropospheric monoxide data assimilation within the CESM-DART system. In addition to assessing the impact of chemical observations from different satellite instruments, the potential of ensemble techniques for inferring the unobserved part of the chemical state of the atmosphere will be emphasized. Data assimilation has found an important application in Observational System Simulation Experiments (OSSEs) both for numerical weather prediction, and more recently, to assess the potential of the current and future chemical observations to constraint atmospheric pollutant predictions. I will also describe ensemble, adjoint and hybrid sensitivity analysis capability for efficient and accurate OSSE framework development.

Finally, I will discuss some personal views on the challenges and opportunities for data assimilation science applied to Earth system prediction both within NCAR and the wider community.

Date: December 7, 2015

Time: 10:00am

Location: Foothills Laboratory - Main Auditorium