Disentangling the Multi-Scale Effects of Sea-Surface Temperatures on Global Precipitation: a Coupled Networks Approach
Total Page:16
File Type:pdf, Size:1020Kb
Originally published as: Ekhtiari, N., Agarwal, A., Marwan, N., Donner, R. V. (2019): Disentangling the multi-scale effects of sea-surface temperatures on global precipitation: A coupled networks approach. - Chaos, 29. DOI: http://doi.org/10.1063/1.5095565 Disentangling the multi-scale effects of sea- surface temperatures on global precipitation: A coupled networks approach Cite as: Chaos 29, 063116 (2019); https://doi.org/10.1063/1.5095565 Submitted: 11 March 2019 . Accepted: 31 May 2019 . Published Online: 20 June 2019 Nikoo Ekhtiari, Ankit Agarwal , Norbert Marwan , and Reik V. Donner COLLECTIONS This paper was selected as Featured ARTICLES YOU MAY BE INTERESTED IN Synchronization behavior in a ternary phase model Chaos: An Interdisciplinary Journal of Nonlinear Science 29, 063115 (2019); https:// doi.org/10.1063/1.5097237 Locating the source node of diffusion process in cyber-physical networks via minimum observers Chaos: An Interdisciplinary Journal of Nonlinear Science 29, 063117 (2019); https:// doi.org/10.1063/1.5092772 Temnothorax albipennis migration inspired semi-flocking control for mobile sensor networks Chaos: An Interdisciplinary Journal of Nonlinear Science 29, 063113 (2019); https:// doi.org/10.1063/1.5093073 Chaos 29, 063116 (2019); https://doi.org/10.1063/1.5095565 29, 063116 © 2019 Author(s). Chaos ARTICLE scitation.org/journal/cha Disentangling the multi-scale effects of sea-surface temperatures on global precipitation: A coupled networks approach Cite as: Chaos 29, 063116 (2019); doi: 10.1063/1.5095565 Submitted: 11 March 2019 · Accepted: 31 May 2019 · Published Online: 20 June 2019 View Online Export Citation CrossMark Nikoo Ekhtiari,1,2,a) Ankit Agarwal,1,3,4 Norbert Marwan,1 and Reik V. Donner1,5,b) AFFILIATIONS 1Potsdam Institute for Climate Impact Research, Telegrafenberg A31, 14473 Potsdam, Germany 2Department of Physics, Humboldt University, Newtonstr. 15, 12489 Berlin, Germany 3Institute of Earth and Environmental Science, University of Potsdam, Karl-Liebknecht-Str. 24-25, 14476 Potsdam-Golm, Germany 4GFZ German Research Centre for Geosciences, Section 5.4: Hydrology, Telegrafenberg, 14473 Potsdam, Germany 5Department of Water, Environment, Construction and Safety, Magdeburg–Stendal University of Applied Sciences, Breitscheidstr. 2, 39114 Magdeburg, Germany a)Electronic addresses: [email protected] and [email protected] b)Electronic mail: [email protected] ABSTRACT The oceans and atmosphere interact via a multiplicity of feedback mechanisms, shaping to a large extent the global climate and its variability. To deepen our knowledge of the global climate system, characterizing and investigating this interdependence is an important task of contemporary research. However, our present understanding of the underlying large-scale processes is greatly limited due to the manifold interactions between essential climatic variables at dierent temporal scales. To address this problem, we here propose to extend the application of complex network techniques to capture the interdependence between global elds of sea-surface temperature (SST) and precipitation (P) at multiple temporal scales. For this purpose, we combine time-scale decomposition by means of a discrete wavelet transform with the concept of coupled climate network analysis. Our results demonstrate the potential of the proposed approach to unravel the scale-specic interdependences between atmosphere and ocean and, thus, shed light on the emerging multiscale processes inherent to the climate system, which traditionally remain undiscovered when investigating the system only at the native resolution of existing climate data sets. Moreover, we show how the relevant spatial interdependence structures between SST and P evolve across time-scales. Most notably, the strongest mutual correlations between SST and P at annual scale (8–16 months) concentrate mainly over the Pacic Ocean, while the corresponding spatial patterns progressively disappear when moving toward longer time-scales. Published under license by AIP Publishing. https://doi.org/10.1063/1.5095565 The study of the climate system using complex networks pro- climate variables and their temporal evolution, which might be vides new insights into spatiotemporal climate dynamics. Most overlooked when focusing only at the native resolution of existing previous studies have focused on a single climate variable only. climate data sets. Accounting for the multivariate and multiscale nature of climate variability introduces a new challenging perspective that could help improve our understanding of the underlying physical mech- anisms. In this study, we focus on the aforementioned two aspects I. INTRODUCTION of multiple variables and time-scales contributing to the variabil- The mutual interdependence between ocean and atmospheric ity of the climate system and show that cross-variable statistical variability, which is mediated via various dierent processes, is a relations evolve dierently at dierent time-scales. Considera- key ingredient in global climate dynamics across a vast range of tion of this previously widely disregarded factor provides a more spatial and temporal scales.1–3 In particular, in the tropics and sub- explicit picture of scale-dependent covariability patterns among tropics coupling between both systems via wind stress, dierential Chaos 29, 063116 (2019); doi: 10.1063/1.5095565 29, 063116-1 Published under license by AIP Publishing. Chaos ARTICLE scitation.org/journal/cha heating and evaporation gives rise to important phenomena such as has been motivated by numerous examples of successful applications monsoons or the El Niño–Southern Oscillation (ENSO).4 Thereby, of complex network approaches across scientic disciplines and to understanding the interactions between ocean and atmosphere in problems of great societal relevance.22 Here, each grid point of a given space and time across scales is of fundamental socioeconomic rel- climate data set corresponds to a node of a spatially embedded net- evance, including elds like agriculture or risk assessment in the work, and links represent strong statistical associations between the context of the insurance sector. The latter relates mainly to the tim- variability observed at the respective grid points. ing and magnitude of climate extremes like heatwaves, oodings, Previous applications of functional network analysis to climate or droughts, which play a vital role for daily life and prosperity of science have mainly focused on individual climatological elds rep- the society yet are hard to assess due to their strong spatiotemporal resenting the dynamics of a single variable or atmospheric layer. heterogeneity. Recently, a multivariate extension of this framework has been intro- In this work, we focus on the interdependence between sea- duced in terms of interdependent or coupled climate networks, which surface temperature (SST) and precipitation (P) as two key variables allows studying the spatiotemporal interdependence patterns among associated with ocean and atmosphere, respectively. There is a large two or more variables from a complex network perspective.23–25 How- body of existing studies on the interrelation between these two vari- ever, a detailed analysis of the multiscale spatial characteristics of two ables and the possible causes and associated impacts of correspond- interdependent climate variables has not yet been performed. In turn, ing extremes. In this context, one key physical mechanism is the given the multiplicity of known (almost periodic to broad-banded exchange of surface uxes.5,6 Recent studies suggest that P tends to irregular) oscillatory variability modes in dierent atmospheric and increase in some approximately linear fashion with increasing SST ocean variables from intraseasonal [e.g., the Madden-Julian Oscilla- over the tropical monsoon basins, suggesting a noticeable impact tion (MJO)] over interannual [e.g., North Atlantic Oscillation (NAO) of SST on tropical precipitation.7 By contrast, in the western North and ENSO] to multidecadal scales [e.g., Pacic Decadal Oscillation Pacic region, the correlation between local rainfall and SST in the (PDO) or Atlantic Multidecadal Oscillation (AMO)], it is pivotal boreal summer is negative, indicating that atmospheric dynamics to explicitly account for the associated distinct scales in order to dominates over oceanic preconditioning in determining the rainfall attribute the resulting spatial interdependence patterns to specic cli- variability of this region.8,9 This observation may be explained by the mate mechanisms. This is precisely what we attempt in the present monsoon activity during this season. The summer monsoon brings study. rainfall, but also leads to SST cooling and, hence, the negative corre- The remainder of this paper is organized as follows: Sec. II lation between local rainfall and SST. Trenberth and Shea8 studied the describes the utilized datasets and methods. The results are pre- covariability between surface temperature and P globally and found sented and discussed in Sec. III. Specically, we will focus on two key negative correlations over land but positive correlations over high local characteristics of spatially embedded coupled networks—the latitudes. They also concluded that ocean conditions drive the atmo- cross-degree and average cross-link distance—between sea-surface sphere. In vast parts of the tropics, higher surface air temperatures temperatures and precipitation. Both measures are presented as maps are associated