Abstract
Aim Because of the low albedo of forests and other biophysical factors, most scenario‐based climate modelling studies indicate that removal of temperate forest will promote cooling, indicating that temperate forests are a source of heat relative to other classes of land cover. Our objective was to test the hypothesis that US temperate forests reduce surface temperatures. Location The continental U nited S tates. Methods Ordinary least squares regression was used to develop relationships between forest extent and surface temperature. Forest extent was derived from the 900 m 2 2001 N ational L and C over D atabase ( NLCD 2001) and surface temperature data were from the MODIS 1 km 2 8‐day composite ( MYD11A2 ). Forest–surface temperature relationships were developed for winter, spring, summer, autumn and annually using 5 years of MODIS land surface temperature data (2007–11) across six spatial scales (1, 4, 9, 16, 25 and 36 km 2 ). Regression models controlled for the effects of elevation, aspect and latitude (by constraining the regressions within a 1° range). Results We did not find any significant positive slopes in regressions of average annual surface temperatures versus the proportion of forest, indicating that forests are not a source of heat relative to other types of land cover. We found that surface temperatures declined as the proportion of forest increased for spring, summer, autumn and annually. The forest–surface temperature relationship was also scale dependent in that spatially extensive forests produced cooler surface temperatures than forests that were dominant only locally. Main conclusions Our results are not consistent with most scenario‐based climate modelling studies. Because of their warming potential, the value of temperate afforestation as a potential climate change mitigation strategy is unclear. Our results indicate that temperate afforestation is a climate change mitigation strategy that should be implemented to promote spatially extensive forests.