Climatic comparisons with tree-ring data from montane forests: are the climatic data appropriate?

Abstract

The appropriateness of relating spatially proximate (40-km radius) temperature and precipitation data from different elevations to montane forest growth patterns was investigated for Mount Washington, New Hampshire. Monthly mean temperature and total precipitation data (1933–1983) were correlated (p < 0.05) among all pairs of meteorological stations (280, 420, 610, 1915 m and regional averages) on or near Mount Washington. The unexplained variance (1 − r2) for precipitation comparisons between meteorological stations was greater relative to temperature. When correlated with the average tree-ring index chronology of 90 red spruce trees on Mount Washington (800–1200 m), the monthly temperature data yielded similar correlative patterns among the four meteorological stations. However, the monthly temperature data from the meteorological stations (610 and 1915 m) most proximate to the montane forest study site were correlated (p < 0.10) with the tree-ring indices for two to three times as many months as the temperature data from the lower elevations. There was no consistency in correlative results of tree-ring indices with monthly precipitation data among the four meteorological stations. However, precipitation measurements and Palmer drought indices are poor indicators of moisture availability in montane forests. We conclude that spatially proximate, low elevation temperature data can underestimate correlative relationships between temperature and montane tree-ring data in the northeastern United States.