Statistical downscaling is a useful technique to localize global or regional climate model projections to assess the potential impact of climate changes. It requires quantifying a relationship between climate model projections and local observations. The usual regression techniques are not applicable, because the climate model projections are meant to reflect the distributions of relevant variables but not to provide daily forecasts. In the case of univariate downscaling, a simple quantile-matching approach with asynchronous measurements often works well, but challenges remain for downscaling bivariate data. In this talk, we discuss a new bivariate downscaling method for asynchronous measurements based on a notion of bivariate ranks and positions. The proposed method is preferable to univariate downscaling, because it is able to preserve general forms of association between two variables (e.g., temperature and precipitation) in statistical downscaling.
This talk is made possible by scholarly contributions from several UIUC faculty and students. It is an example of how interdisciplinary research in the spirit of Bohrer can be fun and valuable.