Investigating the Role of the Atlantic and Pacific in the Early 20th Century Warming

Lea Svendsen1,2, Noel S Keenlyside2,3, Ingo Bethke2,4 and Yongqi Gao1,2, (1)Nansen Environmental and Remote Sensing Center, Bergen, Norway, (2)Bjerknes Centre for Climate Research, Bergen, Norway, (3)Geophysical Institute Bergen, Bergen, Norway, (4)Uni Climate, Uni Research AS, Bergen, Norway
Abstract:
Instrumental records show that there have been two periods of enhanced global warming in the 20th century, the early warming from 1910-1940 and a later period from the end of the 1970s. State-of-the-art coupled models are not able to simulate the early century global warming, and the causes of these variations are not well understood. To improve our understanding of the early warming, we have performed several ensemble experiments with a global coupled model, the CMIP5 version of the Norwegian Earth System Model (NorESM). The experiments consists of 6-member ensembles of historical all-forcing 20th century simulations where daily momentum flux anomalies from the 20th century reanalysis are prescribed to the ocean globally, or only to the Atlantic or the Indo-Pacific region. We find that we are able to constrain sea surface temperature variability to observations in the regions where we prescribe momentum flux. ENSO events are reproduced, as well as multi-decadal variability in the North Atlantic, which neither the pre-industrial control nor the fully coupled 20th century historical simulations of NorESM can simulate. However, only the experiments where momentum flux is prescribed in the Indo-Pacific are able to simulate the early 20th century warming, globally and in the Arctic.