25 Years of Self-Organized Criticality: Concepts and Controversies

Nicholas Watkins; Gunnar Pruessner; Sandra Chapman; Norma Bock Crosby; Henrik Jensen

Source-linked AI summary

25 Years of Self-Organized Criticality: Concepts and Controversies

Nicholas Watkins, Gunnar Pruessner, Sandra Chapman, Norma Bock Crosby, Henrik Jensen

arXiv:1504.04991v1Apr 20, 2015cond-mat.stat-mechnlin.AO

TL;DR

The review addresses uncertainty and controversy over what SOC claims and what evidence can establish it by revisiting Bak, Tang, and Wiesenfeld’s original papers. It reconstructs BTW SOC’s mechanism and conditions, then assesses its empirical status and its broader role as a multiscale-avalanching paradigm in plasma research.

Problem
Uncertainty persists because SOC’s essential claim, evidence, and scope have been confused, while solid empirical evidence outside tuned phase transitions remains limited.
Method
The review rereads BTW’s original papers, reconstructs their reasoning, defines SOC’s necessary and sufficient conditions, and compares the core concept with contemporary plasma applications.
Results
SOC has clear numerical support and a few systems displaying it as originally envisaged, while observational evidence remains difficult and no complete theory exists.
Takeaways & Limitations
In plasma research, SOC has broadened into multiscale avalanching, which has successfully organized observations of bursty, multiscale energy release.
Takeaways & Limitations
The review remains constrained by difficult observational verification of idealized self-similarity and by examples where spatial and temporal correlations do not coincide.

Abstract

from arXiv · show

Introduced by the late Per Bak and his colleagues, self-organized criticality (SOC) has been one of the most stimulating concepts to come out of statistical mechanics and condensed matter theory in the last few decades, and has played a significant role in the development of complexity science. SOC, and more generally fractals and power laws, have attacted much comment, ranging from the very positive to the polemical. The other papers in this special issue (Aschwanden et al, 2014; McAteer et al, 2014; Sharma et al, 2015) showcase the considerable body of observations in solar, magnetospheric and fusion plasma inspired by the SOC idea, and expose the fertile role the new paradigm has played in approaches to modeling and understanding multiscale plasma instabilities. This very broad impact, and the necessary process of adapting a scientific hypothesis to the conditions of a given physical system, has meant that SOC as studied in these fields has sometimes differed significantly from the definition originally given by its creators. In Bak's own field of theoretical physics there are significant observational and theoretical open questions, even 25 years on (Pruessner, 2012). One aim of the present review is to address the dichotomy between the great reception SOC has received in some areas, and its shortcomings, as they became manifest in the controversies it triggered. Our article tries to clear up what we think are misunderstandings of SOC in fields more remote from its origins in statistical mechanics, condensed matter and dynamical systems by revisiting Bak, Tang and Wiesenfeld's original papers.

1 Introduction and synopsis

The review revisits SOC’s origins, competing interpretations, and controversies while distinguishing the original BTW formulation from broader uses in plasma physics and other fields. It clarifies SOC’s conditions, addresses misconceptions, and balances unresolved questions with successful applications.

Scope and motivation: SOC has influenced statistical mechanics, condensed matter theory, complexity science, and diverse application areas, but its parallel interpretations have divided experts.The authors draw on perspectives spanning condensed matter theory, plasma physics, and other disciplines.
Scope and motivation: The review focuses on BTW SOC, whose theoretical core has remained substantially clarified since the original 1987–88 papers.The authors distinguish BTW SOC from other perceptions and use it as the paper’s primary subject.
Review strategy: The authors re-examine BTW’s original papers and claims to define SOC precisely and separate intended meanings from later interpretations and misinterpretations.Sections 4–6 reconstruct the reasoning behind the postulate, while Section 7 identifies necessary and sufficient conditions.
Controversy and clarification: The review analyzes controversies arising from failures to distinguish necessary from sufficient conditions, including the mistaken equation of avalanching, power laws, or long-range correlations with SOC.These logical confusions contributed to divergent versions of SOC and persistent controversy.
Applications and perspective: It also documents SOC’s productive role in organizing observations and modeling multiscale plasma instabilities in solar, magnetospheric, and fusion plasmas.The authors present these applications as a counterbalance to the unresolved theoretical and empirical issues.

2 Why is multiscale avalanching a relevant paradigm for plasmas?

Multiscale avalanching is relevant to plasmas because these systems are driven, dissipative, far from equilibrium, and release energy intermittently across strongly coupled scales. When detailed event-by-event modeling becomes intractable, scaling-based approaches offer a way to characterize collective behavior.

Plasma phenomenology: Plasma systems exhibit driven, dissipative, far-from-equilibrium behavior with bursty energy release across multiple scales and many instability pathways.Solar flares, magnetospheric substorms, and confined fusion plasmas exemplify this phenomenology.
Modeling choices: Detailed modeling of isolated events is suited to a single spatio-temporal scale but becomes inadequate when physics on all strongly coupled scales matters.The alternative is motivated by the intractability of solving individual-event dynamics across the full multiscale system.
Scaling perspective: Renormalization-group reasoning characterizes local interactions and their coarse-graining, with self-similar scaling producing power-law event sizes and long-range correlations.These observables provide a collective description when bottom-up event modeling cannot be completed.
Why SOC is attractive: SOC attracts plasma researchers because it introduces dynamics into a criticality-based paradigm, although successful renormalization-group applications to plasma physics remain elusive.The contrast is with the established successes of RG in equilibrium critical phenomena and its extensions to nonequilibrium systems.

3 Perceptions and receptions of SOC

The literature contains nested perceptions of SOC ranging from a narrowly defined self-tuned phase transition to broad claims about fractals, power laws, and contingency. The review argues that these expansions generated confusion, while the core idea remains theoretically meaningful but empirically difficult to establish.

Nested perceptions: The core SOC claim is that self-tuned phase transitions can exist in nature and dynamically generate spatio-temporal fractals.This is the narrowest interpretation represented in the review’s schematic.
Figure 1: Figure 1 orders perceptions from minimal to visionary, but the outermost claim is not necessarily a superset of the inner claims.The schematic’s nested presentation therefore represents a range of perceptions rather than a strict logical hierarchy.
Nested perceptions: Broader interpretations claim that SOC causes all natural fractals, all power laws, or nature’s contingency, with increasing speculative scope.The review treats the latter interpretations as progressively more sweeping than the original core.
Consequences: Divergent perceptions produced confusion about SOC’s claims, proofs, evidence, and explanatory scope; computational confirmation mainly supports the core claim.The review notes that observational, numerical, and analytical work was still converging on corroboration of that core.
Authors’ scope: The core interpretation was relatively tightly defined, whereas the second and third interpretations were long recognized as wrong and the fourth was more speculative.The authors focus on the core interpretation and discuss the others mainly to explain resulting confusion.

4 BTW’s Stated Motivation: “a dynamical theory of the physics of fractals”

BTW proposed SOC to supply a dynamical theory linking spatial fractals and temporal 1/f noise through self-organized critical states. The review emphasizes that this original goal concerned dynamically evolving spatio-temporal fractals, not every power law in nature.

Motivation: BTW’s motivation combined widespread spatial fractals, unresolved 1/f noise, and the perceived absence of a general framework connecting them.Their proposal was guided by scaling observed in equilibrium and nonequilibrium critical phenomena.
Scope of the claim: BTW’s concern was power-law correlations between spatial fluctuations, not the unrestricted problem of explaining every power-law size distribution.They sometimes used power-law distributions as proxies for correlations, but the connection is not generally automatic.
1/f noise: The review notes that 1/f noise had existing explanations, but BTW considered them unsatisfying and lacking generality.Geomagnetic 1/f spectra nevertheless became early supporting evidence for SOC in magnetospheric physics.
Core postulate: The original SOC idea treated spatial fractals and temporal 1/f noise as two manifestations of a self-organized critical state.The proposed mechanism was a dynamical theory of spatio-temporal correlations.
Scope of the claim: The gap between explaining spatio-temporal fractal avalanching and explaining any fractal or power law became a major source of SOC controversy.The authors identify this broadened interpretation as a perennial misunderstanding of BTW’s aim.

5 Self-organization, the SOC postulate and BTW’s definition of SOC

BTW’s SOC hypothesis sought to unify dynamically evolving spatial and temporal fractals through slowly driven, open, dissipative systems that organize toward a critical state. The review distinguishes this core proposal from broader interpretations and stresses that spatial and temporal scaling need not always coincide.

BTW’s original aim was to unify dynamically evolving spatial and temporal fractals rather than explain every power law with one mechanism.The authors connect this program to questions about fractals, complexity, and renormalization-group ideas.
BTW argued that spatial and temporal scaling were intrinsically linked, associating temporal 1/f noise with spatial fractal scaling.This proposed connection was motivated by the difficulty of maintaining long-range coherence without interactions across scales.
Later work showed that temporal scaling can occur in nonequilibrium and equilibrium dynamics without requiring related spatial fractality.A spatiotemporal correlation function can test whether scaling is algebraic in both r and t.
The SOC postulate states that open, extended, dissipative systems driven slowly may automatically reach a critical state, with fractals appearing as snapshots of that state.The authors summarize the associated conditions as slowly driven, interaction-dominated, thresholded systems.
In BTW’s framework, openness and boundary dissipation allow conserved quantities to flow through the system and relax at its boundaries.The boundary remains relevant even as system size grows, rather than scaling away as in conventional statistical physics.
The review argues that interpreting every fractal, avalanche, long-range correlation, or power law as SOC creates logical errors and divergent versions of the concept.It therefore separates necessary from sufficient conditions and distinguishes the original theoretical core from broader disciplinary usages.

6 Criticality and minimal stability

The review separates three meanings of “critical”: critical correlations, local thresholds, and a global parameter at a critical value. In SOC, the defining claim concerns system-wide critical features, not universal local proximity to thresholds.

Meanings of criticality: “Critical” in SOC can mean spatiotemporal correlations, microscopic thresholds, or a global control parameter reaching a critical value.The review identifies these as distinct meanings and illustrates them with the BTW sandpile.
The BTW model: The one-dimensional BTW model topples a site when h_i − h_i+1 > 1, transfers one grain downstream, and continues until no site exceeds the threshold.Driving resumes only after the avalanche ends, while the boundary site dissipates particles.
The BTW model: In higher-dimensional sandpiles, local threshold exceedance redistributes slope units to neighboring sites, while boundary conditions remove units outside the lattice.This provides the local interaction and dissipation rules underlying avalanche dynamics.
Self-organized control: The slope density ζ acts as a pile-generated control parameter: driving increases it, dissipation decreases it, and large avalanches promote further dissipation.Because ζ is a global average, its fluctuations decrease as system size L increases.
Critical spatiotemporal correlations: In continuous phase transitions, criticality involves power-law correlations, fluctuations across all length scales, and divergent susceptibility at a special control-parameter value.Perturbations can therefore propagate across all length and time scales because no characteristic scale exists.
SOC criticality: SOC criticality is not reached by externally setting a control parameter to a critical value; it concerns the system displaying long-ranged correlations and susceptibility-like critical features.The authors distinguish this from the looser idea that every local degree of freedom is near a threshold.
Minimal stability: The average dynamical variable can remain well below the local threshold: the two-dimensional sandpile’s average height is 17/8 = 2.125 versus threshold 3.The one-dimensional model’s local near-threshold behavior is treated as a coincidence rather than the general meaning of SOC.

7 The necessary and sufficient conditions for SOC

The paper distinguishes SOC’s observable phenotype from the sufficient ingredients proposed to generate it, while questioning whether those ingredients are complete or equivalent to SOC.

SOC’s phenotype: SOC requires non-trivial scaling, spatio-temporal power-law correlations, and apparent self-tuning to a critical point.The first two features are treated as aspects of criticality, while self-tuning distinguishes SOC from traditionally tuned critical phenomena.
Boundary cases: Invasion percolation exhibits scaling and spatial correlations but lacks stationary evolution and self-tuning, illustrating that critical-like features alone do not establish SOC.It sits at the critical point by definition rather than dynamically organizing itself there.
Necessary versus sufficient conditions: SOC is defined phenomenologically by its necessary conditions, whereas sufficient conditions describe a mechanism expected to produce those characteristics.This distinction separates observable behavior from causal system ingredients and motivates the paper’s examination of SOC’s relation to SDIDT.
SOC’s genotype: The proposed sufficient ingredients are non-linear interactions, avalanching or intermittency, and separated driving and relaxation time scales.Thresholds commonly implement the non-linear interaction, while slow driving helps sustain distinct avalanches.
SOC and SDIDT: If conditions 4–6 are complete and minimal, SDIDT and SOC coincide; otherwise, only some SDIDT systems display SOC.The paper presents these alternatives through competing Venn-diagram relationships between SDIDT and SOC.

8 Why then is SOC controversial?

SOC remains controversial because its core claim has been confused with the phenomena it seeks to explain, while evidence and theoretical understanding remain limited. The authors distinguish necessary from sufficient conditions and review empirical and modeling difficulties.

Confusion between SOC’s proposed mechanism and the phenomena it explains has blurred what counts as evidence or a valid application.
Power-law distributions do not necessarily imply long-range correlations or SOC; some directed sandpile models show power-law avalanches without spatial correlations.
Long-range spatial and temporal correlations can occur independently, so their simultaneous presence requires separate justification as an SOC condition.
Observational support for natural scaling is often limited in scale range and robustness, making direct identification of SOC difficult.
Real sandpile experiments failed to detect robust scaling, while accepted SOC models remain analytically unresolved and lack a quantitative mean-field theory for finite boundaries.
Proposed SOC criteria include robust parameter-independent finite-size scaling, spatio-temporal correlations, and apparent self-tuning to a critical point.
Some systems may approach a critical transition without fine-tuning because coupling between dynamics and the order parameter prevents exact criticality.
Long temporal correlations can support probabilistic prediction of large-event sequences even though SOC systems do not signal individual large events.

9 SOC in the wild: how has SOC inspired research on space and fusion plasmas?

SOC inspired researchers to organize multiscale, bursty observations in solar, magnetospheric, and fusion plasmas. Applications produced useful statistical and structural analyses, while also exposing complications from projection, variable driving, and competing origins.

Solar plasmas: Solar flares exhibit energy distributions spanning more than eight orders of magnitude, motivating BTW-inspired models and extensive SOC analysis.
Magnetospheric plasmas: Auroral research found scaling in AE time series between about 1 minute and 2 hours, with a scale break attributed to global substorms.
Fusion plasmas: Fusion tokamak plasmas motivated SOC-inspired work because they are driven, dissipative systems with multiple steady states, anomalous transport, and bursty energy and material release.
Magnetospheric plasmas: Threshold analyses of Polar auroral images found fat-tailed distributions of blob counts and areas, plus a fixed-scale population associated with global substorms.
Magnetospheric plasmas: Spatiotemporal tracking instead reported power-law distributions for blob areas and lifetimes across 3–5 orders of magnitude.
Magnetospheric plasmas: Auroral scaling is difficult to interpret because the observed emission projects dynamic magnetospheric structures and may reflect correlated bursty particle flows.
Magnetospheric plasmas: Solar-wind turbulence may contribute to AE scale-free behavior, although analyses also indicate an internally generated component may coexist with it.
Applications: Power-law distributions remain practically useful for estimating natural-hazard strength and informing spacecraft shielding against extreme space-weather events.

10 Summary and conclusion

The review untangles SOC’s origins, meanings, evidence, and applications across diverse fields. It concludes that strict SOC is uncommon and difficult to verify, while the broader multiscale-avalanching paradigm has substantially organized research on complex plasma phenomena.

SOC originated as a hypothesis that self-organizing systems can produce spatio-temporal fractals through continuous phase transitions, but its status and scope became confused.
Few systems display SOC in its original form, although some provide evidence consistent with the theory; SOC is almost certainly not ubiquitous.
SOC connected previously specialized research on heavy tails, spatio-temporal fractals, and 1/f noise across disciplinary communities.
The broader SOC-inspired perspective seeks microscopic interactions that persist across scales in rescaled form, providing a sharper quantitative emphasis than generic complexity language.
Heavy-tailed distributions and correlations challenge single-scale explanations and can support more quantitative characterization of fluctuations and risks.
Testing SOC: Testing SOC in finite plasma systems is difficult because scaling alone cannot reliably distinguish SOC from turbulence.
Testing SOC: Power-law event-size statistics are necessary but insufficient evidence for SOC because trivial linear similarity and nonlinear plasma dynamics can also produce scaling.
Testing SOC: In plasma applications, strict SOC has broadened into multiscale avalanching because observations show coupled, bursty processes across many spatiotemporal scales.

Appendix A: Dimensional analysis, scaling and self-similarity

Scaling relates observables across parameter changes; non-trivial scaling expresses self-similarity and produces power-law behavior without an intrinsic scale. Dimensional analysis gives only unit-driven, trivial scaling, whereas critical phenomena connect non-trivial exponents to interactions and symmetries.

Scaling relates the value of a physical observable at one parameter set to its value at another.It is a continuous symmetry of certain physical observables.
Dimensional analysis: For a pendulum, dimensional analysis requires the frequency to scale as a constant multiple of (g/ℓ)^1/2, with amplitude dependence represented by an unknown function.Without the small-angle approximation, the frequency can depend on the initial amplitude through f(φ0).
Dimensional analysis: Dimensional analysis constrains observables through their units, producing usually integer or simple-fraction exponents called trivial scaling.The resulting scaling is what would remain without nonlinear effects.
Self-similarity: Non-trivial scaling arises from self-similarity and can generate rich power laws, such as the percolation fractal dimension 91/48.Dimensionless finite ratios involving characteristic lengths or times permit exponents beyond dimensional-analysis values.
Self-similarity: Power-law correlations indicate competing scales that balance rather than dominate, leaving no characteristic scale within the system.For C(r) = C0(r/ξ)^−µ, the ratio C(r)/C(2r) = 2^µ cannot determine ξ without knowing C0.
Critical scaling: At continuous transitions, power-law correlations and fluctuations across length scales are associated with criticality, while finite-size scaling studies dependence on system size.Critical-point observables can scale with distance or with a system’s linear extent L.