We have submitted a new paper that documents the important role of vegetation on long term near surface temperature trends, as well as the actual observed trends that have been occurring. The paper is
Fall, S., D. Niyogi, R.A. Pielke Sr., A. Gluhovsky, and E. Kalnay, 2007: Impacts of land surface properties on temperature trends using North American regional reanalysis over the USA. Int. J. Climatol., submitted,
with the abstract
“Recent studies have confirmed the impacts of land surface processes on surface temperature trends. We investigate this relationship over the conterminous United States (CONUS) by using the observation minus reanalysis (OMR) approach. We derive OMR trends for the 1979-2003 period from two datasets: the US Historical Climate Network (USHCN observations), and the NCEP-NCAR North American Regional Reanalysis (NARR).
We use the mean square differences (MSDs) for the comparisons between temperature anomalies from station observations (both unadjusted and adjusted) and interpolated reanalysis data. Trends of monthly mean temperature anomalies at individual station level and over the CONUS show the agreement between USHCN and NARR and demonstrate that NARR captures the climate variability at different time scales. Temperature anomalies exhibit a spatial variability, with amplitudes increasing from south to north. As a further evaluation of spatial patterns, the RMS differences also depict a good agreement over the eastern CONUS (0.29°C to 0.6°C).
OMR trend results suggest that unlike findings from studies based on the global reanalysis, NARR often has a larger warming trend than observations, but 10-year moving window trends reveal that this situation varies considerably over time, from one station to another. Most of the warming accounts for large winter increases (adjusted USHCN: 0.46°C/decade; NARR: 0.5°C/decade), especially over the Midwest. In contrast, other seasons –in particular summer and fall- exhibit low trends. This difference in winter and other season trends is a prominent feature.
OMR trends are sensitive to land cover types. Evergreen needleleaf forests, open shrublands, bare soils and urban areas exhibit the largest warming. Moreover, the OMR method captures the trend variability of land types that are subject to seasonality (e.g. croplands, deciduous forests). Overall the results obtained with the regional reanalysis are consistent with findings from global reanalysis (R1 and ERA 40).”