资源环境科技发展态势分析平台

A methodological framework based on nonlinear recurrence analysis is proposed to examine the historical data evolution of extremes of maximum and minimum daily mean areal temperature patterns over time under different climate scenarios. The methodology is based on both historical data and atmospheric General Circulation Model (GCM) produced climate scenarios for the periods 1961-2000 and 2061-2100 which correspond to 1 x CO2 and 2 x CO2 scenarios. Historical data were derived from the actual daily observations coupled with atmospheric circulation patterns (CPs). The dynamics of the temperature was reconstructed in the phase-space from the time series of temperatures. The statistically comparing different temperature patterns were based on some discriminating statistics obtained by the Recurrence Quantification Analysis (RQA). Moreover, the boot-strap method of Schinkel et al. (2009) was adopted to calculate the confidence bounds of RQA parameters based on a structural preserving resampling. The overall methodology was implemented to the mountainous Mesochora catchment in Central-Western Greece. The results reveal substantial similarities between the historical maximum and minimum daily mean areal temperature statistical patterns and their confidence bounds, as well as the maximum and minimum temperature patterns in evolution under the 2 x CO2 scenario. A significant variability and non-stationary behaviour characterizes all climate series analyzed. Fundamental differences are produced from the historical and maximum 1 x CO2 scenarios, the maximum 1 x CO2 and minimum 1 x CO2 scenarios, as well as the confidence bounds for the two CO2 scenarios. The 2 x CO2 scenario reflects the strongest shifts in intensity, duration and frequency in temperature patterns. Such transitions can help the scientists and policy makers to understand the effects of extreme temperature changes on water resources, economic development, and health of ecosystems and hence to proceed to effective proactive management of extreme phenomena. The impacts of the findings on the predictability of the extreme daily mean areal temperature patterns are also commented.