NOAA/WDS Paleoclimatology - Marcinkowski - Thornton Lakes - TSME - ITRDB WA148

A stationary response of tree radial growth to climatic variables is assumed as a basis for climatic reconstructions and future growth projections in response to climate change. Mountain hemlock (Tsuga mertensiana (Bong.) Carriere) trees on the western slopes of the North Cascade Range (Washington, USA) were examined for stability in growth response to climatic influences at a small spatial scale. Moving correlation functions demonstrate that climate-growth interactions are nonstationary over time, alternating between periods of significant and nonsignificant responses. Correlations between growth and winter precipitation have weakened, becoming statistically insignificant in the last decade, but correlations with spring temperature and previous-year summer temperature have strengthened, becoming statistically significant. The Pacific Decadal Oscillation influences patterns in climate-growth correlations but does not seem to account for the most recent changes in correlation strength. At an interannual scale, growth differs between El Nino Southern Oscillation phases, specifically between El Nino and La Nina years and between La Nina and neutral phase years. The variability in growth response to climate at interannual and interdecadal time frames, especially with the climate changes emerging in recent decades, will challenge the reliability and accuracy of reconstruction and predictive models.

本研究以树木径向生长对气候变量的稳态响应（stationary response）为基础，开展气候重建及气候变化背景下未来树木生长预测相关研究。针对美国华盛顿州北喀斯喀特山脉西坡的山地铁杉（Tsuga mertensiana (Bong.) Carriere），本研究在小空间尺度下探究其生长对气候影响的响应稳定性。滑动相关函数（Moving correlation functions）分析结果显示，气候-生长交互作用随时间呈现非稳态特征，在显著响应与非显著响应阶段交替出现。生长与冬季降水的相关性已逐渐减弱，近十年已无统计学显著性；而生长与春季气温、前一年夏季气温的相关性则不断增强，达到统计学显著水平。太平洋年代际振荡（Pacific Decadal Oscillation）会影响气候-生长相关性的格局，但似乎无法解释近期相关性强度的变化。在年际尺度上，树木生长在厄尔尼诺-南方涛动（El Nino Southern Oscillation）不同相位间存在显著差异，具体表现为厄尔尼诺（El Nino）年与拉尼娜（La Nina）年、拉尼娜（La Nina）年与中性相位年之间的生长差异。在年际和年代际时间尺度上，树木生长对气候的响应存在变异，尤其是近几十年愈发凸显的气候变化，将对气候重建与预测模型的可靠性和准确性构成挑战。