A51G-3111:
Impact of Land Use Change over North America as simulated by the Canadian Regional Climate Model

Friday, 19 December 2014
Arlette Chacon1, Laxmi Sushama1 and Hugo Beltrami2, (1)University of Quebec at Montreal UQAM, Montreal, QC, Canada, (2)St. Francis Xavier University, Antigonish, Canada
Abstract:
This study investigates the biogeophysical impacts of human-induced land cover change, particularly crops, on the regional climate of North America, using the fifth generation Canadian Regional Climate Model (CRCM5). To this effect, two simulations are performed with CRCM5 with different land cover datasets – one corresponding to the potential vegetation (i.e. without land use change) and the other corresponding to current land use. Most of the land use changes are concentrated over the US mid-west and south-central Canada, where forests and grasses have been replaced by crops. This transformation changes the surface parameters, particularly vegetation fractional area, leaf area index, albedo, roughness length and rooting depth among other variables, in the regions where land cover change takes place in these simulations. Both simulations span the 1988–2012 period and are driven by ERA-Interim at the lateral boundaries. The sea surface temperature and sea ice cover that vary inter-annually are also taken from ERA-Interim. Results suggest that regions where forests/grasses were replaced by crops generally show increases in albedo, particularly during the spring, fall and winter seasons, with the increase in albedo being largest for winter. This higher increase in albedo during winter is due to a snow-mediated positive feedback. The increased albedo values during winter, spring and fall are reflected in the cooler 2 meter temperature obtained in the simulation with land use change, compared to that with potential vegetation. Some cooling is observed in the summer for the simulation with land use change, mostly due to the increased latent heat fluxes. Increases in precipitation are noted for these regions, but are not statistically significant.