July 14, 2016
Statistics associate professor gives invited talk at International Society for Non-Parametric Statistics
Juan Du, associate professor of statistics, delivered an invited talk at the third Conference of the International Society for Non-Parametric Statistics in Avignon, France in June.
"Spatial Functional Modeling of Weather Change Impact on Corn Yield in Kansas" described an approach to modeling the relationship between weather outcomes and corn yields in Kansas over time accounting for functional nonlinearity and the underlying spatial autocorrelations. This model is intended to understand how the climate is changing in the U.S. and in Kansas, and Du hopes to help assess the impacts on regional patterns of land cover, water, agricultural systems, and thus the Kansas economy.
Du's model efficiently takes into account the fact that the weather data were recorded daily, whereas crop yield data are available only yearly, as well as complex dependence structure of crop yields that are gathered at different locations. The model also must overcome nonlinearity, because the relationship between temperature and corn yield is not linear. Accurate models could help identify the actual influential factors and the critical periods of weather change that could affect crop yields during the growing season.
Du's travel was supported by a Faculty Development Award from the Office of Research and Sponsored Programs in fall 2015 as well as by the statistics department.