Abstract
Here we present a new R-package (KrigR) for acquiring and statistically downscaling climate data for ecological applications. The package is principally designed to make use of two of the most recent global reanalysis climate products from the European Centre for Medium Range Weather Forecasting ERA5 and ERA5-land. These reanalysis products include numerous climate variables relevant for ecological applications including air temperature, precipitation, and soil moisture at hourly resolution and spatial resolutions of 30 by 30km and 9 by 9 km respectively. Reanalysis products resolve issues of biases, discontinuities and inconsistencies within individual observational products (e.g. WorldClim, CRU), and represent the state-of-the-art knowledge on historical climate. While the hourly temporal resolution of the ERA5 data family marks a great improvement over many other climate data sets, this can be further improved using statistical downscaling to match the spatial resolution of such legacy datasets. Our package uses kriging to downscale the reanalysis output to a user-specified resolution, reliably up to one order of magnitude finer than the reanalysis product. Our approach in designing KrigR was to give the user as much freedom as possible while making the existing downloading and kriging methodology more streamlined. KrigR allows one to download any ERA5-family variable, at any given temporal resolution, in any chosen region (rectangular or as a shape) across the globe. Furthermore, while we supply the user with a downloading function for the covariates used for kriging, our kriging method does allow for user-input not generated by our download functions.
Erik Kusch
Senior Engineer & Statistical Consultant
In my research, I focus on statistical approaches to understanding complex processes and patterns in biology using a variety of data banks.