The Williams Lab

Effects of past and future climate change on the distribution of terrestrial vegetation

Using a combination of future climate scenarios and paleoecological data from the last deglaciation, we are studying:
1) The potential magnitude and velocity of climate change in eastern North America, with particular emphasis in Wisconsin
2) The sensitivity of plant communities to past and future climate changes

In doing so, we address three major components:

Figure 1. Maps of the standardized Euclidean distance (SED) between observed late 20th-century North American climate and the 21st-century model ensemble for the SRES A2 scenario, for Madison, WI. Blue and green colors indicate areas with late-20th- century climates that are similar to the 21st-century projections; while red and orange colors indicate areas with late-20th-century climates that differ from the to the 21st-century projections. The SED between observed late 20th century climate and modeled late 20th century shows the historical climatic dissimilarity across the region and the locations of each city. From Veloz et al. (in press).

1) Climate analogs. Climate-analog analysis offer a place-based approach for identifying and communicating possible climate change impacts that is complementary to the species based approaches of species distributional molds, and assumptions about the characterization and conservatism of species. Using the WICCI downscaled information, at 0.1 decimal degree resolution, and identifies contemporary locations in North America that are the most similar to the projected climates for Wisconsin. Under higher-end emission scenarios (Fig 1) the contemporary climatic analogs for Wisconsin's late-21st-centory climate will be almost entirely outside the sate (Veloz et al. in press). We are also studying the effects of past climates with no modem analog upon the predictive ability of species distribution molds (Veloz et al. in review) and habitat-correlation prioritization models (Williams et al., in revision).




2) Vegetation modeling. We are using the LPJ, a dynamic vegetation model forced by the WICCI downscaled climate projections, to model future changes in vegetation and carbon sequestration in the terrestrial ecosystems of Wisconsin. The mean climate projection is substantial warming and an increase in precipitation. In the LPJ model, drying solid leads to reduced tree cover across southern and western Wisconsin, establishing a more prairie-like environment. Evergreen tree cover declines, and project climate changes produce a large loss of terrestrial carbon, primary from vegetation during the first half of the century and from both vegetation and and soil in the second half (Notaro et al. in review).

3) Climate velocity. It is widely recognized that climate change poses a grave threat to biodiversity, exacerbating existing threats because of land use change, fragmentation, and environmental degradation. This is due to its' impacts on species distributions, abundance and ecological interactions. This impact will be of considerable magnitude if the rate of change exceeds the pace of biological response. Here we examine the multidimensional aspects of climatic change, the effect of this multidimensionality and its' significance for both biodiversity conservation and natural resource management. For this we use the idea of climate change velocity (Loarie et al. 2009) and multivariate climatic aggregations to evaluate the spatial/temporal rate of change in the mean and variability of physiologically meaningful climatic variables (Fig 2). In a similar fashion I evaluate the importance of extreme climatic events. The advances of my work over current analyses on this topic is the use of multiple climate variables and the development of an integrated measure of climatic change that considers both its spatial-temporal dimension and multi-variable composition.


Figure 2. Late 21st-century predictions of mean annual temperature and annual total precipitation change according to WICCI downscaled model ensemble for the SRES A2 scenario. Colors represent the spatial-temporal rate of change in evaluated variables (i.e climate velocity). Blue and green colors represent areas where the movement per year to reach and analog is very short (slow velocities); while red and orange colors represent areas where there movement per year to reach and analog needs to be very long (fast velocity). From Ordonez, unpublished.


Williams, J.W., Kharouba, H. M., Veloz, S., Vellend, M., Mclachlan, J., Liu, Z., Otto-Bliesner, B. & He, F. (in revision) The Ice Age Ecologist: Testing methods for reserve prioritizationduring the last global warming. Global Ecology and Biogeography.

Veloz, S., Williams, J.W., Blois, J.L., He, F., Otto-Bliesner, B., and Z. Liu (in review, Global Change Biology). No-analog climates and shifting realized niches during the late Quaternary: implications for 21st-century predictions by species distribution models.

Veloz, S.,, Williams, J.W., Vimont, D. J., Vavrus, S., Lorenz, D. J. & Notaro, M. (in press) Identifying climatic analogs for Wisconsin under 21st-century climate-change scenarios. Climatic Change.


Climate Research Unit (CRU)
IPCC - Intergovernmental Panel on Climate Change
Center for Climatic Research
WICCI- Wisconsin Iniative on Climate Change Impacts
NOAA- National Oceanic and Atmospheric Administration
NOAA Climate Services
Human Footprint (Wildlife Conservation Society)
GEO Data Portal
ISRIC - World Soil Information

Website feedback or accesibility issues | copyright 2007 The Williams Lab

LinkHome LinkGeography