Date:
Tuesday, March 6, 2018, 3:00pm
Location:
Seminar Room MCZ, 429
Speaker: Aleyda Trevino
Aleyda will lead a discussion on using machine learning techniques and data assimilation to improve the parameterization schemes in climate models.
In the discussion, Aleyda will lead us through the following key points:
- A potential of using the techniques in data assimilation and machine learning to improve the current parameterization schemes in climate models.
- As a proof of concept, Schneider et al show that the computational efficiency can be improved using a low-order model (e.g. Lorenz-96) and methods like Bayesian inversions and ensemble Kalman inversions.
- We will discuss some detailed examples for adopting machine learning techniques in the climate and Earth sciences modeling.
schneider_et_al-2017-geophysical_research_letters.pdf | 1.46 MB |