Development of moderate-resolution gridded monthly air temperature and degree-day maps for the Labrador-Ungava region of northern Canada

Abstract

Detailed climatological grids are needed for many applications, including permafrost prediction, ecological modelling and infrastructure planning. This study describes the creation of moderate-resolution gridded climate datasets covering the entire Labrador-Ungava region (50°–63°N) for a series of climate indices, including monthly air temperature, annual air temperature, freezing degree-days (FDDs) and thawing degree-days. Using a recently developed spatiotemporal infilling technique, temporally consistent climate grids spanning the 1948–2014 period were derived at a monthly resolution. Comparison against within-sample and out-of-sample climate stations revealed thin plate spline smoothing as more accurate for modelling air temperatures than regression, kriging and co-kriging. Evaluation of derived air temperature grids across a wide range of environments and scenarios shows an overall accuracy of 0.8 ± 0.3 °C. Spatially distributed air temperatures were converted to thawing and FDDs using an empirical transfer function that compensates for the impacts of continentality and coastal proximity. These climate datasets will form the basis of inputs for future ecological and environmental modelling in the eastern Labrador-Ungava region.