资源环境科技发展态势分析平台

Regional climate variability is strongly related to large-scale circulation modes. However, little is known about changes in their spectral characteristics under climate change. Here, we introduce piecewise evolutionary spectra to quantify time-varying variability and co-variability of climate variables, and use ensemble periodograms to estimate these spectra. By employing a large ensemble of climate change simulations, we show that changes in the variability and relationships of the North Atlantic Oscillation (NAO) and regional surface temperatures are disparate on individual timescales. The relation between NAO and surface temperature over high-latitude lands weakens the most on 20-year timescales compared to shorter timescales, whereas the relation between NAO and temperature over subtropical North Africa strengthens more on shorter timescales than on 20-year timescales. These projected evolution and timescale-dependent changes shed new light on the controlling factors of circulation-induced regional changes. Accounting for them can lead to the improvement of future regional climate predictions.