Have New England heat waves been influenced by forest cover trends?

Have New England Heat Waves Been Influenced by Forest Cover Trends?

In the late 1700s, as the population of southern New England began to grow, large numbers of trees were cut down for building, heating and cooking, and to provide room for agriculture and livestock grazing.  By 1880, forest cover had dropped to less than 50% of what it had been in the middle 1700s (Foster and Aber, 2004).  Beginning in the early 1900s, as reliance on wood diminished, forest cover began to recover, reaching early 1700s levels by 1980. 

Grasslands have a very different surface energy balance compared to forests, which would suggest that heat waves might be measurably affected by the presence or absence of forest cover.  For instance, the summertime albedo of grassland is larger than that of forest cover (19% vs. 16%), while the summertime surface roughness length is smaller (12 cm vs. 50 cm), and the moisture availability is much lower (30% for forest cover, 15% for grassland).  The larger albedo would reflect more energy away from the surface, leading to cooler temperatures, but the larger roughness length would lead to larger surface sensible and latent heat fluxes, and thus warmer surface air temperatures.  The fluxes are also affected by the vertical temperature gradient, which could be smaller given the larger surface albedo for grasslands.  In addition, the surface fluxes will vary with the surface wind speed, which can be affected by the size of the daytime convective mixed layer, which in turn is affected by the surface fluxes.  Evapotranspiration also can vary greatly between forest and field.  Hence, there are many non-linear forcing mechanisms at work.

To determine how these forcing mechanisms interact, we have performed three pairs of exploratory model simulations.  One of each pair uses the current forest cover and the other member of the pair runs without the presence of forests in Massachusetts and southern New Hampshire.  The simulations were initialized at 0000 UTC, with initial and boundary weather conditions taken from the Global Forecast System analysis grib files for three days (May 25, 2007, August 3, 2007, and June 10, 2008) characterized by afternoon temperatures above 33^o C  (91^oF) throughout much of Massachusetts and southern New Hampshire.  The land use characteristics were changed from the present forest cover (deciduous broadleaf forest) to grassland, and the simulations were run for 24 hours.  The temperature differences in the lowest model layer are shown in the Figure for 1900 UTC, August 3, 2007.  The red shaded area (contours in ^oC, positive shaded in red) coincides almost exactly with the region of changed forest cover and the positive sign indicates that the grassland cover was warmer than the forest cover.  This finding agrees with recent observational data from three different ecosystems which showed that forested areas are cooler than open areas (Juang, et. al, 2007).  These temperature changes are part of a complete change in the surface energy balance.  The grassland dewpoints are as much as 1-2^oC lower, and the convective boundary layer is 25 hPa deeper, partly as a result of the lower moisture availability for grassland.  The simulations for the other two days had similar results.  While these results are preliminary, they suggest that forest cover can play an important role in modulating heat waves and that, due to the large historic and continuing changes in New England forests, understanding the role of forest cover is also important to understanding heat wave trends.

References:

Foster, D., and J. Aber, 2004:  Background and framework for long-term ecological research.  Chapter 1 in Forests in Time:  The environmental consequences of 1,000 years of change in New England, D. Foster and J. Aber, Eds.  Yale University Press, 2004, 477pp.

Juang, J.-Y., G. Katul, M. Siqueira, P. Stoy, K. Novick, 2007:  Separating the effects of albedo from eco-physiological changes on surface temperature along a successional chronosequence in the southeastern United States, Geophys. Res. Lett. 34, L21408.