Predictability of summer droughts over the US: Multiscale coupling between land, convection, and large-scale circulation
Summer droughts over the US central plains are a result of complex interactions and feedbacks between multiple components of the climate system including atmosphere, land, and ocean. In this project, I am applying exploratory data analysis to various geo-spatial data products in order to understand the GP drought onset drivers and identify unknown relationships between hydroclimate parameters that can increase the drought predictability skill.
Here is a list of some of the techniques/concepts used in this research:
- Data mining, machine learning, large data, dimension reduction
- Deep learning, self organizing maps, SVD, EOF, regression, correlation
- Statistical/Dynamic drought prediction, forecast skill, model uncertainty
- Seasonal and inter-annual variability, Atmospheric variability, climate change
- Land-atmosphere interactions, deep and shallow convection, Atmospheric circulation