Observed Changes in Summer Season Length over the Northern Hemisphere

Abstract

This study analyzed the observed long-term variations in the summer season timing and length over the Northern Hemisphere (NH) lands and its sub-regions using temperature-based indices. The climatological mean showed coastal-land contrast as summer starts and ends earlier in-lands than coastal areas due to different heat capacity. During the past 60 years (1953-2012), observations show the lengthening of summer season with the advanced summer onset and the delayed summer withdrawal over the all NH lands. The trend of summer onset has contributed more to the observed increase in the summer season length in many sub-regions. In order to understand anthropogenic and natural contributions to the observed change, we analyzed summer season trends from multi-model CMIP5 simulations simulated under different forcing factors (ALL: anthropogenic plus natural forcing, NAT: natural forcing only, and GHG: greenhouse gas only forcing). ALL and GHG simulations are found to reproduce the overall observed global and regional lengthening trends but NAT cannot, implying that the greenhouse gas increase is the main cause of the observed change of summer season indices. ALL runs, however, tend to underestimate the observed summer onset trend and overestimate the withdrawal trend, indicating overall phase delay in models compared to the observations.. Further, we assessed possible contribution of multi-decadal variabilities such as Pacific Decadal Oscillation (PDO) and Atlantic Multi-decadal Oscillation (AMO) to the regional trends in summer season length. Results suggest that the multi-decadal variability can explain small but considerable portion (up to around ±20%) of the observed and simulated trends in summer season length, more strongly over Europe (PDO) and North America (AMO).