Non-stationarities in the relationships of heavy precipitation events in the Mediterranean area and the large-scale circulation in the second half of the 20th century
Non-stationarities in the relationships of heavy precipitation events in the Mediterranean area and the large-scale circulation in the second half of the 20th century
复制标题
20世纪下半叶地中海地区强降水事件与大范围环流关系的非平稳性
DOI：
10.1016/j.gloplacha.2016.10.009
复制
发表时间：
2017
影响因子：
3.9
通讯作者：
Jacobeit
中科院分区：
文献类型：
--
作者：
Merkenschlager;Hertig;Jacobeit
在分析地中海地区强降水事件与大尺度环流异常的时间变化关系的基础上，建立了分位数回归模型。在700 hPa和850 hPa水平上使用不同的环流和热力学变量作为预报因子，以及在地中海地区不同站点的日降水时间序列作为预报对象，对模型进行了校准。分析是在20世纪世纪后半叶进行的。在评估预测因子-预测因子关系的非平稳性的范围内，将时间序列分为校准期和验证期。使用100个随机子样本在固定条件下校准/验证模型。100个随机样本的最高和最低技能得分用于确定随机变异性的范围。非平稳条件下的模型性能来自交叉验证运行子区间的技能得分。如果连续几年的技能分数超出随机变化范围，则称为非平稳性。特别是伊比利亚半岛和黎凡特地区的影响，非平稳性，前者与显着的正偏差的技能得分，后者与显着的负偏差。通过对黎凡特地区的一个案例研究，我们确定了三个可能的原因，非平稳行为的预测-预测和关系。地中海涛动作为一个超坐标系统影响地中海盆地的气旋活动和塞浦路斯低压的位置和强度。总体而言，它表明，非平稳性必须考虑到统计降尺度模型的发展。
In the context of analyzing temporal varying relationships of heavy precipitation events in the Mediterranean area and associated anomalies of the large-scale circulation, quantile regression models were established. The models were calibrated using different circulation and thermodynamic variables at the 700 hPa and 850 hPa levels as predictors as well as daily precipitation time series at different stations in the Mediterranean area as predictand. Analyses were done for the second half of the 20th century. In the scope of assessing non-stationarities in the predictor-predictand relationships the time series were divided into calibration and validation periods. 100 randomized subsamples were used to calibrate/validate the models under stationary conditions. The highest and lowest skill score of the 100 random samples was used to determine the range of random variability. The model performance under non-stationary conditions was derived from the skill scores of cross-validated running subintervals. If the skill scores of several consecutive years are outside the range of random variability a non-stationarity was declaimed. Particularly the Iberian Peninsula and the Levant region were affected by non-stationarities, the former with significant positive deviations of the skill scores, the latter with significant negative deviations. By means of a case study for the Levant region we determined three possible reasons for non-stationary behavior in the predictor-predictand relationships. The Mediterranean Oscillation as a superordinate system affects the cyclone activity in the Mediterranean basin and the location and intensity of the Cyprus low. Overall, it is demonstrated that non-stationarities have to be taken into account within statistical downscaling model development.
登录
查看更多内容
气候和土地利用变化对地中海生态系统植被动态的影响：古生态学和动态建模的见解
影响因子：
5.2
作者：
Paul D. Henne;C. Elkin;D. Colombaroli;S. Samartin;H. Bugmann;O. Heiri;W. Tinner
通讯作者：
W. Tinner
地中海地区的极端降水和相关的高层天气尺度流结构
影响因子：
4.6
作者：
A. Toreti;Paraskevi Giannakaki;O. Martius
通讯作者：
O. Martius
约旦河上游流域的基于物理的分布式水文建模和有效模型方程的研究
DOI：
--
发表时间：
2006
期刊：
影响因子：
--
作者：
H. Kunstmann;A. Heckl;A. Rimmer
通讯作者：
A. Rimmer
影响因子：
4.6
作者：
Xoplaki, E;González-Rouco, JF;Wanner, H
通讯作者：
Wanner, H
DOI：
10.1002/joc.973
发表时间：
2003-12-01
期刊：
INTERNATIONAL JOURNAL OF CLIMATOLOGY
影响因子：
--
作者：
Dünkeloh, A;Jacobeit, J
通讯作者：
Jacobeit, J