Understanding the Nature of Abrupt Decadal Shifts in a Changing Climate

RICKETTS James-thesis_nosignature.pdf - Submitted Version (17MB) | Preview

Ricketts, James Henry (2019) Understanding the Nature of Abrupt Decadal Shifts in a Changing Climate. PhD thesis, Victoria University.


Planning for future climate risk tends to incorporate assumptions of smoothly accelerating climate change, around which unchanged variability constitutes the risk boundaries. This constitutes hypothesis H1 – forced warming and natural variability evolve gradually and independently and the climate response is trend-like. Against this, there is evidence for H2 – forced warming interacts with natural variability and the climate response includes abrupt steps. Earlier than expected breaching of risk bounds follows from H2. New automation tools, and post-detection tests find and characterise step-like regime onsets in temperatures. With these tools I show that step-like temperature regime shifts are detectable at all spatial scales at the land and ocean surface, and in the vertical temperature structure of the ocean. Based on published climate models shifts respond to warming by becoming more intense, wider-spread and more frequent. Regimes are regional, differ qualitatively between land and ocean, align with natural variability coincident with known bio-physical shifts. Two, circa 1976 and 1996, align with the Pacific Decadal Oscillation, involving rapid vertical ocean restructuring. One, 1968 in the Southern Hemisphere ocean does not, and 1986 in the Northern Hemisphere reflects atmospheric reorganisation. H2 is strongly supported by the findings. Step-like warming dominates trends, increasingly so at finer scale.

Item type Thesis (PhD thesis)
URI https://vuir.vu.edu.au/id/eprint/40470
Subjects Historical > FOR Classification > 0104 Statistics
Historical > FOR Classification > 0401 Atmospheric Sciences
Current > Division/Research > Institute for Sustainable Industries and Liveable Cities
Keywords climate change; climate data; temperature; spatial analyses; multistep bivariate test
Download/View statistics View download statistics for this item

Search Google Scholar

Repository staff login