Dynamics of co-behaviour of climate processes over Southern Africa

Doctoral Thesis

2021

Permanent link to this Item
Authors
Journal Title
Link to Journal
Journal ISSN
Volume Title
Publisher
Publisher
License
Series
Abstract
Large-scale climate processes such as El NiƱo-Southern Oscillation (ENSO), Antarctic Oscillation (AAO), and many others, play varying roles in regional climate variability across the world. While the role of singular processes have been explored in many studies, the combined influence of multiple large-scale processes has received far less attention. Key to this is the challenge of developing methodologies to support the analysis of multiple processes interacting in potentially non-linear ways (co-behaviour) in a particular region. This study details the development of such a methodology and demonstrates its utility in the analysis of the co-behaviour of largescale process interactions on regional precipitation and temperature variability over southern Africa. The study defines co-behaviour as the interaction of large-scale processes that may influence regional circulation leading to climate variability. A novel methodology which involves a combination of analysis techniques such as Self-Organizing Maps (SOM) and Principal Component Analysis (PCA) is developed to identify and quantify such co-behaviour which accommodates potentially non-linear interactions. This methodology is evaluated in the context of southern African regional climate using three key processes, namely ENSO, AAO and Inter-tropical Convergence Zone (ITCZ), and characterizations of regional circulation, and temperature and rainfall variability. Analysis of co-behaviour under observed conditions identifies results that concur with prior studies, in particular the dominant regional response to ENSO, but also establishes key examples of co-behaviour such as the role of the AAO in moderating and altering the regional response to ENSO which is important for understanding regional climate variability. Application of the approach to Global Climate Model (GCM) simulations of past climate reveals that while many GCMs are able to capture individual processes, in particular ENSO, they fail to adequately represent regional circulation variability and key observed co-behaviour. The study therefore clearly demonstrates the importance of co-behaviour in understanding regional climate variability as well as showing the usefulness of the new methodology in investigating co-behaviour. Finally, the new insights into evaluating model performance through the lens of core climate processes and their interaction provides a significant step forward in both model development and application for decision making.
Description

Reference:

Collections