Modelling studies on the probability and predictability of future climate change

Knutti, Reto (2002). Modelling studies on the probability and predictability of future climate change (Unpublished). (Dissertation, Universität Bern, Philosophisch–naturwissenschaftliche Fakultät, Physikalisches Institut, Abteilung für Klima– und Umweltphysik)

Text knutti02phd.pdf - Other Restricted to registered users only Available under License BORIS Standard License. Download (3MB) | Request a copy

There is strong evidence that the Earth’s climate has changed significantly over the last century. Direct observations of near-surface atmospheric temperature and sub-surface ocean temperature as well as numerous indirect observations of sea ice extent, snow cover and glacier retreats all point to a warming world. The topic of climate change has thus attracted worldwide attention and alarmed the public as well as the scientific community seeking for explanations. The recent Third Assessment Report of the Intergovernmental Panel on Climate Change [IPCC, 2001] has further strengthened our confidence in the widely accepted view of a strong anthropogenic influence on the climate system during the last century. Human activities affect the climate system primarily through the emission of greenhouse gases and aerosols that change the radiation balance of the atmosphere. Greenhouse gas concentrations are projected to increase further over the next century due to anthropogenic emissions, and will thus cause an acceleration of the rates of surface warming and sea level rise, probably accompanied by widespread changes in the hydrological cycle and in the frequency of extreme events.
Although some aspects of anthropogenic climate change are well understood, many uncertainties remain. For the evolution of future climate, they are mainly associated with our limited understanding of the physical climate system, the limited representations of processes and feedbacks in climate models, but also with large uncertainties related to the economical, technical and social development of the World population. Of special concern is the fact that many uncertainties of key numbers in climate change are not rigorously quantified, e.g., the uncertainties of surface warming projections are generally based on expert judgement rather than on clear quantitative statistical methods. Therefore, there is a need for a quantitative assessment of the risk of climate change, and this requires probabilities that can be assigned to different scenarios and to uncertainties of future projections.
Forecasts of future climate change can only come from numerical climate models, and are inevitably uncertain. The publications presented in this thesis all deal with uncertainties in climate change, with probabilities, and with the question of how far the future of the climate system is predictable. A very efficient climate model of reduced complexity that represents the ocean, the atmosphere and the sea ice is used to assess questions regarding the probability and predictability of future warming and of the evolution of the large-scale deep ocean circulation. The combination of ensemble simulations and simplified climate models proves to be a powerful approach to tackle problems which cannot be answered with complex models due to computational constraints.
The deep ocean circulation on the largest scale is driven by temperature and salinity at the surface. This thermohaline circulation is currently in an active circulation mode where warm surface water flows northward in the Atlantic, sinks into the depth around Greenland, flows southward as a deep western boundary current, and finally recirculates by broad upwelling in the Pacific and Indian Ocean. Chapter 2 discusses the future of the thermohaline circulation in the context of a warming world, the feedbacks that affect the strength and stability of this active circulation mode, and reviews the current state of knowledge about a possible reversal of the thermohaline circulation into a passive state without deep water formation in the Atlantic. Such a transition would be relatively abrupt, nonlinear, possibly irreversible, and would have widespread consequences for the climate in the North Atlantic region. In chapter 3, the parameter space close to this instability threshold for a transition of the circulation into a different regime is investigated using ensemble simulations forced with stochastic noise. The predictability of the future of the thermohaline circulation is found to be strongly limited when such a threshold is approached. The study presented in chapter 4 shows that for a global warming scenario, sea level rise from thermal expansion of seawater would be significantly different if the ocean circulation switched into a qualitatively different state.
In chapters 5 and 6, ensembles of many thousand model simulations are used to assess the uncertainty of future climate change. Probability density functions are derived for important parameters in the climate system as well as for the future warming, for changes in sea level and for the thermohaline circulation. For these studies, all relevant uncertainties are systematically taken into account and the observed warmings of ocean and atmosphere are used as independent constraints. In contrast to earlier studies where the accuracy of model projections was often judged by experts, this new method provides an objective statistical assessment of the uncertainties associated with global warming and provides a synthesis of probabilities related to climate change which is consistent with the warming we have experienced so far.

Interest & Impact

Downloads

0 since deposited on 08 May 2024
0 in the past 12 months

Citations

5 Citations in Web of Science ®
6 Citations in Scopus

Search

in Google Scholar™

Services

Actions (login required)

Edit item

Actions (login required)

Edit item