Sensitivity to initial conditions

Lorenz equations used to generate plots for the y variable. The initial conditions for x and z were kept the same but those for y were changed between 1.001, 1.0001 and 1.00001. The values for ${\displaystyle \rho }$, ${\displaystyle \sigma }$ and ${\displaystyle \beta }$ were 45.92, 16 and 4 respectively. As can be seen from the graph, even the slightest difference in initial values causes significant changes after about 12 seconds of evolution in the three cases. This is an example of sensitive dependence on initial conditions.

Sensitivity to initial conditions means that each point in a chaotic system is arbitrarily closely approximated by other points that have significantly different future paths or trajectories. Thus, an arbitrarily small change or perturbation of the current trajectory may lead to significantly different future behavior.^[2]

Sensitivity to initial conditions is popularly known as the "butterfly effect", so-called because of the title of a paper given by Edward Lorenz in 1972 to the American Association for the Advancement of Science in Washington, D.C., entitled Predictability: Does the Flap of a Butterfly's Wings in Brazil set off a Tornado in Texas?.^[28] The flapping wing represents a small change in the initial condition of the system, which causes a chain of events that prevents the predictability of large-scale phenomena. Had the butterfly not flapped its wings, the trajectory of the overall system could have been vastly different.

As suggested in Lorenz's book entitled The Essence of Chaos, published in 1993,^[5] "sensitive dependence can serve as an acceptable definition of chaos". In the same book, Lorenz defined the butterfly effect as: "The phenomenon that a small alteration in the state of a dynamical system will cause subsequent states to differ greatly from the states that would have followed without the alteration." The above definition is consistent with the sensitive dependence of solutions on initial conditions (SDIC). An idealized skiing model was developed to illustrate the sensitivity of time-varying paths to initial positions.^[5] A predictability horizon can be determined before the onset of SDIC (i.e., prior to significant separations of initial nearby trajectories).^[29]

A consequence of sensitivity to initial conditions is that if we start with a limited amount of information about the system (as is usually the case in practice), then beyond a certain time, the system would no longer be predictable. This is most prevalent in the case of weather, which is generally predictable only about a week ahead.^[30] This does not mean that one cannot assert anything about events far in the future—only that some restrictions on the system are present. For example, we know that the temperature of the surface of the earth will not naturally reach 100 °C (212 °F) or fall below −130 °C (−202 °F) on earth (during the current geologic era), but we cannot predict exactly which day will have the hottest temperature of the year.

In more mathematical terms, the Lyapunov exponent measures the sensitivity to initial conditions, in the form of rate of exponential divergence from the perturbed initial conditions.^[31] More specifically, given two starting trajectories in the phase space that are infinitesimally close, with initial separation ${\displaystyle \delta \mathbf {Z} _{0}}$, the two trajectories end up diverging at a rate given by

${\displaystyle |\delta \mathbf {Z} (t)|\approx e^{\lambda t}|\delta \mathbf {Z} _{0}|,}$

where ${\displaystyle t}$ is the time and ${\displaystyle \lambda }$ is the Lyapunov exponent. The rate of separation depends on the orientation of the initial separation vector, so a whole spectrum of Lyapunov exponents can exist. The number of Lyapunov exponents is equal to the number of dimensions of the phase space, though it is common to just refer to the largest one. For example, the maximal Lyapunov exponent (MLE) is most often used, because it determines the overall predictability of the system. A positive MLE is usually taken as an indication that the system is chaotic.^[8]

In addition to the above property, other properties related to sensitivity of initial conditions also exist. These include, for example, measure-theoretical mixing (as discussed in ergodic theory) and properties of a K-system.^[11]

Non-periodicity

A chaotic system may have sequences of values for the evolving variable that exactly repeat themselves, giving periodic behavior starting from any point in that sequence. However, such periodic sequences are repelling rather than attracting, meaning that if the evolving variable is outside the sequence, however close, it will not enter the sequence and in fact, will diverge from it. Thus for almost all initial conditions, the variable evolves chaotically with non-periodic behavior.

Topological mixing

Six iterations of a set of states ${\displaystyle [x,y]}$ passed through the logistic map. The first iterate (blue) is the initial condition, which essentially forms a circle. Animation shows the first to the sixth iteration of the circular initial conditions. It can be seen that mixing occurs as we progress in iterations. The sixth iteration shows that the points are almost completely scattered in the phase space. Had we progressed further in iterations, the mixing would have been homogeneous and irreversible. The logistic map has equation ${\displaystyle x_{k+1}=4x_{k}(1-x_{k})}$. To expand the state-space of the logistic map into two dimensions, a second state, ${\displaystyle y}$, was created as ${\displaystyle y_{k+1}=x_{k}+y_{k}}$, if ${\displaystyle x_{k}+y_{k}<1}$ and ${\displaystyle y_{k+1}=x_{k}+y_{k}-1}$ otherwise.

The map defined by x → 4 x (1 – x) and y → (x + y) mod 1 also displays topological mixing. Here, the blue region is transformed by the dynamics first to the purple region, then to the pink and red regions, and eventually to a cloud of vertical lines scattered across the space.

Topological mixing (or the weaker condition of topological transitivity) means that the system evolves over time so that any given region or open set of its phase space eventually overlaps with any other given region. This mathematical concept of "mixing" corresponds to the standard intuition, and the mixing of colored dyes or fluids is an example of a chaotic system.

Topological mixing is often omitted from popular accounts of chaos, which equate chaos with only sensitivity to initial conditions. However, sensitive dependence on initial conditions alone does not give chaos. For example, consider the simple dynamical system produced by repeatedly doubling an initial value. This system has sensitive dependence on initial conditions everywhere, since any pair of nearby points eventually becomes widely separated. However, this example has no topological mixing, and therefore has no chaos. Indeed, it has extremely simple behavior: all points except 0 tend to positive or negative infinity.

Topological transitivity

A map ${\displaystyle f:X\to X}$ is said to be topologically transitive if for any pair of non-empty open sets ${\displaystyle U,V\subset X}$, there exists ${\displaystyle k>0}$ such that ${\displaystyle f^{k}(U)\cap V\neq \emptyset }$. Topological transitivity is a weaker version of topological mixing. Intuitively, if a map is topologically transitive then given a point x and a region V, there exists a point y near x whose orbit passes through V. This implies that it is impossible to decompose the system into two open sets.^[32]

An important related theorem is the Birkhoff Transitivity Theorem. It is easy to see that the existence of a dense orbit implies topological transitivity. The Birkhoff Transitivity Theorem states that if X is a second countable, complete metric space, then topological transitivity implies the existence of a dense set of points in X that have dense orbits.^[33]

Density of periodic orbits

For a chaotic system to have dense periodic orbits means that every point in the space is approached arbitrarily closely by periodic orbits.^[32] The one-dimensional logistic map defined by x → 4 x (1 – x) is one of the simplest systems with density of periodic orbits. For example, ${\displaystyle {\tfrac {5-{\sqrt {5}}}{8}}}$ → ${\displaystyle {\tfrac {5+{\sqrt {5}}}{8}}}$ → ${\displaystyle {\tfrac {5-{\sqrt {5}}}{8}}}$ (or approximately 0.3454915 → 0.9045085 → 0.3454915) is an (unstable) orbit of period 2, and similar orbits exist for periods 4, 8, 16, etc. (indeed, for all the periods specified by Sharkovskii's theorem).^[34]

Sharkovskii's theorem is the basis of the Li and Yorke^[35] (1975) proof that any continuous one-dimensional system that exhibits a regular cycle of period three will also display regular cycles of every other length, as well as completely chaotic orbits.

Strange attractors

The Lorenz attractor displays chaotic behavior. These two plots demonstrate sensitive dependence on initial conditions within the region of phase space occupied by the attractor.

Some dynamical systems, like the one-dimensional logistic map defined by x → 4 x (1 – x), are chaotic everywhere, but in many cases chaotic behavior is found only in a subset of phase space. The cases of most interest arise when the chaotic behavior takes place on an attractor, since then a large set of initial conditions leads to orbits that converge to this chaotic region.^[36]

An easy way to visualize a chaotic attractor is to start with a point in the basin of attraction of the attractor, and then simply plot its subsequent orbit. Because of the topological transitivity condition, this is likely to produce a picture of the entire final attractor, and indeed both orbits shown in the figure on the right give a picture of the general shape of the Lorenz attractor. This attractor results from a simple three-dimensional model of the Lorenz weather system. The Lorenz attractor is perhaps one of the best-known chaotic system diagrams, probably because it is not only one of the first, but it is also one of the most complex, and as such gives rise to a very interesting pattern that, with a little imagination, looks like the wings of a butterfly.

Unlike fixed-point attractors and limit cycles, the attractors that arise from chaotic systems, known as strange attractors, have great detail and complexity. Strange attractors occur in both continuous dynamical systems (such as the Lorenz system) and in some discrete systems (such as the Hénon map). Other discrete dynamical systems have a repelling structure called a Julia set, which forms at the boundary between basins of attraction of fixed points. Julia sets can be thought of as strange repellers. Both strange attractors and Julia sets typically have a fractal structure, and the fractal dimension can be calculated for them.

Coexisting attractors

Coexisting chaotic and non-chaotic attractors within the generalized Lorenz model.^[37][38][39] There are 128 orbits in different colors, beginning with different initial conditions for dimensionless time between 0.625 and 5 and a heating parameter r = 680. Chaotic orbits recurrently return close to the saddle point at the origin. Nonchaotic orbits eventually approach one of two stable critical points, as shown with large blue dots. Chaotic and nonchaotic orbits occupy different regions of attraction within the phase space.

In contrast to single type chaotic solutions, recent studies using Lorenz models ^[40][41] have emphasized the importance of considering various types of solutions. For example, coexisting chaotic and non-chaotic may appear within the same model (e.g., the double pendulum system) using the same modeling configurations but different initial conditions. The findings of attractor coexistence, obtained from classical and generalized Lorenz models,^[37][38][39] suggested a revised view that "the entirety of weather possesses a dual nature of chaos and order with distinct predictability", in contrast to the conventional view of "weather is chaotic".

Minimum complexity of a chaotic system

Bifurcation diagram of the logistic map x → r x (1 – x). Each vertical slice shows the attractor for a specific value of r. The diagram displays period-doubling as r increases, eventually producing chaos. Darker points are visited more frequently.

Discrete chaotic systems, such as the logistic map, can exhibit strange attractors whatever their dimensionality. Universality of one-dimensional maps with parabolic maxima and Feigenbaum constants ${\displaystyle \delta =4.669201...}$,${\displaystyle \alpha =2.502907...}$^[42][43] is well visible with Tahn map proposed as a toy model for discrete laser dynamics: ${\displaystyle x\rightarrow Gx(1-\mathrm {tanh} (x))}$, where ${\displaystyle x}$ stands for electric field amplitude, ${\displaystyle G}$^[44] is laser gain as bifurcation parameter. The gradual increase of ${\displaystyle G}$ at interval ${\displaystyle [0,\infty )}$ changes dynamics from regular to chaotic one^[45] with qualitatively the same bifurcation diagram as those for logistic map.

In contrast, for continuous dynamical systems, the Poincaré–Bendixson theorem shows that a strange attractor can only arise in three or more dimensions. Finite-dimensional linear systems are never chaotic; for a dynamical system to display chaotic behavior, it must be either nonlinear or infinite-dimensional.

The Poincaré–Bendixson theorem states that a two-dimensional differential equation has very regular behavior. The Lorenz attractor discussed below is generated by a system of three differential equations such as:

${\displaystyle {\begin{aligned}{\frac {\mathrm {d} x}{\mathrm {d} t}}&=\sigma y-\sigma x,\\{\frac {\mathrm {d} y}{\mathrm {d} t}}&=\rho x-xz-y,\\{\frac {\mathrm {d} z}{\mathrm {d} t}}&=xy-\beta z.\end{aligned}}}$

where ${\displaystyle x}$, ${\displaystyle y}$, and ${\displaystyle z}$ make up the system state, ${\displaystyle t}$ is time, and ${\displaystyle \sigma }$, ${\displaystyle \rho }$, ${\displaystyle \beta }$ are the system parameters. Five of the terms on the right hand side are linear, while two are quadratic; a total of seven terms. Another well-known chaotic attractor is generated by the Rössler equations, which have only one nonlinear term out of seven. Sprott^[46] found a three-dimensional system with just five terms, that had only one nonlinear term, which exhibits chaos for certain parameter values. Zhang and Heidel^[47][48] showed that, at least for dissipative and conservative quadratic systems, three-dimensional quadratic systems with only three or four terms on the right-hand side cannot exhibit chaotic behavior. The reason is, simply put, that solutions to such systems are asymptotic to a two-dimensional surface and therefore solutions are well behaved.

While the Poincaré–Bendixson theorem shows that a continuous dynamical system on the Euclidean plane cannot be chaotic, two-dimensional continuous systems with non-Euclidean geometry can still exhibit some chaotic properties.^[49] Perhaps surprisingly, chaos may occur also in linear systems, provided they are infinite dimensional.^[50] A theory of linear chaos is being developed in a branch of mathematical analysis known as functional analysis.

The above elegant set of three ordinary differential equations has been referred to as the three-dimensional Lorenz model.^[51] Since 1963, higher-dimensional Lorenz models have been developed in numerous studies^{[52][53][37][38]} for examining the impact of an increased degree of nonlinearity, as well as its collective effect with heating and dissipations, on solution stability.

Infinite dimensional maps

The straightforward generalization of coupled discrete maps^[54] is based upon convolution integral which mediates interaction between spatially distributed maps: ${\displaystyle \psi _{n+1}({\vec {r}},t)=\int K({\vec {r}}-{\vec {r}}^{,},t)f[\psi _{n}({\vec {r}}^{,},t)]d{\vec {r}}^{,}}$,

where kernel ${\displaystyle K({\vec {r}}-{\vec {r}}^{,},t)}$ is propagator derived as Green function of a relevant physical system,^[55] ${\displaystyle f[\psi _{n}({\vec {r}},t)]}$ might be logistic map alike ${\displaystyle \psi \rightarrow G\psi [1-\tanh(\psi )]}$ or complex map. For examples of complex maps the Julia set ${\displaystyle f[\psi ]=\psi ^{2}}$ or Ikeda map ${\displaystyle \psi _{n+1}=A+B\psi _{n}e^{i(|\psi _{n}|^{2}+C)}}$ may serve. When wave propagation problems at distance ${\displaystyle L=ct}$ with wavelength ${\displaystyle \lambda =2\pi /k}$ are considered the kernel ${\displaystyle K}$ may have a form of Green function for Schrödinger equation:.^[56][57]

${\displaystyle K({\vec {r}}-{\vec {r}}^{,},L)={\frac {ik\exp[ikL]}{2\pi L}}\exp[{\frac {ik|{\vec {r}}-{\vec {r}}^{,}|^{2}}{2L}}]}$.

Jerk systems

In physics, jerk is the third derivative of position, with respect to time. As such, differential equations of the form

${\displaystyle J\left({\overset {...}{x}},{\ddot {x}},{\dot {x}},x\right)=0}$

are sometimes called jerk equations. It has been shown that a jerk equation, which is equivalent to a system of three first order, ordinary, non-linear differential equations, is in a certain sense the minimal setting for solutions showing chaotic behavior. This motivates mathematical interest in jerk systems. Systems involving a fourth or higher derivative are called accordingly hyperjerk systems.^[58]

A jerk system's behavior is described by a jerk equation, and for certain jerk equations, simple electronic circuits can model solutions. These circuits are known as jerk circuits.

One of the most interesting properties of jerk circuits is the possibility of chaotic behavior. In fact, certain well-known chaotic systems, such as the Lorenz attractor and the Rössler map, are conventionally described as a system of three first-order differential equations that can combine into a single (although rather complicated) jerk equation. Another example of a jerk equation with nonlinearity in the magnitude of ${\displaystyle x}$ is:

${\displaystyle {\frac {\mathrm {d} ^{3}x}{\mathrm {d} t^{3}}}+A{\frac {\mathrm {d} ^{2}x}{\mathrm {d} t^{2}}}+{\frac {\mathrm {d} x}{\mathrm {d} t}}-|x|+1=0.}$

Here, A is an adjustable parameter. This equation has a chaotic solution for A=3/5 and can be implemented with the following jerk circuit; the required nonlinearity is brought about by the two diodes:

In the above circuit, all resistors are of equal value, except ${\displaystyle R_{A}=R/A=5R/3}$, and all capacitors are of equal size. The dominant frequency is ${\displaystyle 1/2\pi RC}$. The output of op amp 0 will correspond to the x variable, the output of 1 corresponds to the first derivative of x and the output of 2 corresponds to the second derivative.

Similar circuits only require one diode^[59] or no diodes at all.^[60]

See also the well-known Chua's circuit, one basis for chaotic true random number generators.^[61] The ease of construction of the circuit has made it a ubiquitous real-world example of a chaotic system.

Cryptography

Chaos theory has been used for many years in cryptography. In the past few decades, chaos and nonlinear dynamics have been used in the design of hundreds of cryptographic primitives. These algorithms include image encryption algorithms, hash functions, secure pseudo-random number generators, stream ciphers, watermarking, and steganography.^[124] The majority of these algorithms are based on uni-modal chaotic maps and a big portion of these algorithms use the control parameters and the initial condition of the chaotic maps as their keys.^[125] From a wider perspective, without loss of generality, the similarities between the chaotic maps and the cryptographic systems is the main motivation for the design of chaos based cryptographic algorithms.^[124] One type of encryption, secret key or symmetric key, relies on diffusion and confusion, which is modeled well by chaos theory.^[126] Another type of computing, DNA computing, when paired with chaos theory, offers a way to encrypt images and other information.^[127] Many of the DNA-Chaos cryptographic algorithms are proven to be either not secure, or the technique applied is suggested to be not efficient.^{[128][129][130]}

Robotics

Robotics is another area that has recently benefited from chaos theory. Instead of robots acting in a trial-and-error type of refinement to interact with their environment, chaos theory has been used to build a predictive model.^[131] Chaotic dynamics have been exhibited by passive walking biped robots.^[132]

Biology

For over a hundred years, biologists have been keeping track of populations of different species with population models. Most models are continuous, but recently scientists have been able to implement chaotic models in certain populations.^[133] For example, a study on models of Canadian lynx showed there was chaotic behavior in the population growth.^[134] Chaos can also be found in ecological systems, such as hydrology. While a chaotic model for hydrology has its shortcomings, there is still much to learn from looking at the data through the lens of chaos theory.^[135] Another biological application is found in cardiotocography. Fetal surveillance is a delicate balance of obtaining accurate information while being as noninvasive as possible. Better models of warning signs of fetal hypoxia can be obtained through chaotic modeling.^[136]

As Perry points out, modeling of chaotic time series in ecology is helped by constraint.^{[137]: 176, 177} There is always potential difficulty in distinguishing real chaos from chaos that is only in the model.^{[137]: 176, 177} Hence both constraint in the model and or duplicate time series data for comparison will be helpful in constraining the model to something close to the reality, for example Perry & Wall 1984.^{[137]: 176, 177} Gene-for-gene co-evolution sometimes shows chaotic dynamics in allele frequencies.^[138] Adding variables exaggerates this: Chaos is more common in models incorporating additional variables to reflect additional facets of real populations.^[138] Robert M. May himself did some of these foundational crop co-evolution studies, and this in turn helped shape the entire field.^[138] Even for a steady environment, merely combining one crop and one pathogen may result in quasi-periodic- or chaotic- oscillations in pathogen population.^[139]: 169

Economics

It is possible that economic models can also be improved through an application of chaos theory, but predicting the health of an economic system and what factors influence it most is an extremely complex task.^[140] Economic and financial systems are fundamentally different from those in the classical natural sciences since the former are inherently stochastic in nature, as they result from the interactions of people, and thus pure deterministic models are unlikely to provide accurate representations of the data. The empirical literature that tests for chaos in economics and finance presents very mixed results, in part due to confusion between specific tests for chaos and more general tests for non-linear relationships.^[141]

Chaos could be found in economics by the means of recurrence quantification analysis. In fact, Orlando et al.^[142] by the means of the so-called recurrence quantification correlation index were able detect hidden changes in time series. Then, the same technique was employed to detect transitions from laminar (regular) to turbulent (chaotic) phases as well as differences between macroeconomic variables and highlight hidden features of economic dynamics.^[143] Finally, chaos theory could help in modeling how an economy operates as well as in embedding shocks due to external events such as COVID-19.^[144]

Other areas

In chemistry, predicting gas solubility is essential to manufacturing polymers, but models using particle swarm optimization (PSO) tend to converge to the wrong points. An improved version of PSO has been created by introducing chaos, which keeps the simulations from getting stuck.^[145] In celestial mechanics, especially when observing asteroids, applying chaos theory leads to better predictions about when these objects will approach Earth and other planets.^[146] Four of the five moons of Pluto rotate chaotically. In quantum physics and electrical engineering, the study of large arrays of Josephson junctions benefitted greatly from chaos theory.^[147] Closer to home, coal mines have always been dangerous places where frequent natural gas leaks cause many deaths. Until recently, there was no reliable way to predict when they would occur. But these gas leaks have chaotic tendencies that, when properly modeled, can be predicted fairly accurately.^[148]

Chaos theory can be applied outside of the natural sciences, but historically nearly all such studies have suffered from lack of reproducibility; poor external validity; and/or inattention to cross-validation, resulting in poor predictive accuracy (if out-of-sample prediction has even been attempted). Glass^[149] and Mandell and Selz^[150] have found that no EEG study has as yet indicated the presence of strange attractors or other signs of chaotic behavior.

Researchers have continued to apply chaos theory to psychology. For example, in modeling group behavior in which heterogeneous members may behave as if sharing to different degrees what in Wilfred Bion's theory is a basic assumption, researchers have found that the group dynamic is the result of the individual dynamics of the members: each individual reproduces the group dynamics in a different scale, and the chaotic behavior of the group is reflected in each member.^[151]

Redington and Reidbord (1992) attempted to demonstrate that the human heart could display chaotic traits. They monitored the changes in between-heartbeat intervals for a single psychotherapy patient as she moved through periods of varying emotional intensity during a therapy session. Results were admittedly inconclusive. Not only were there ambiguities in the various plots the authors produced to purportedly show evidence of chaotic dynamics (spectral analysis, phase trajectory, and autocorrelation plots), but also when they attempted to compute a Lyapunov exponent as more definitive confirmation of chaotic behavior, the authors found they could not reliably do so.^[152]

In their 1995 paper, Metcalf and Allen^[153] maintained that they uncovered in animal behavior a pattern of period doubling leading to chaos. The authors examined a well-known response called schedule-induced polydipsia, by which an animal deprived of food for certain lengths of time will drink unusual amounts of water when the food is at last presented. The control parameter (r) operating here was the length of the interval between feedings, once resumed. The authors were careful to test a large number of animals and to include many replications, and they designed their experiment so as to rule out the likelihood that changes in response patterns were caused by different starting places for r.

Time series and first delay plots provide the best support for the claims made, showing a fairly clear march from periodicity to irregularity as the feeding times were increased. The various phase trajectory plots and spectral analyses, on the other hand, do not match up well enough with the other graphs or with the overall theory to lead inexorably to a chaotic diagnosis. For example, the phase trajectories do not show a definite progression towards greater and greater complexity (and away from periodicity); the process seems quite muddied. Also, where Metcalf and Allen saw periods of two and six in their spectral plots, there is room for alternative interpretations. All of this ambiguity necessitate some serpentine, post-hoc explanation to show that results fit a chaotic model.

By adapting a model of career counseling to include a chaotic interpretation of the relationship between employees and the job market, Amundson and Bright found that better suggestions can be made to people struggling with career decisions.^[154] Modern organizations are increasingly seen as open complex adaptive systems with fundamental natural nonlinear structures, subject to internal and external forces that may contribute chaos. For instance, team building and group development is increasingly being researched as an inherently unpredictable system, as the uncertainty of different individuals meeting for the first time makes the trajectory of the team unknowable.^[155]

Some say the chaos metaphor—used in verbal theories—grounded on mathematical models and psychological aspects of human behavior provides helpful insights to describing the complexity of small work groups, that go beyond the metaphor itself.^[156]

The red cars and blue cars take turns to move; the red ones only move upwards, and the blue ones move rightwards. Every time, all the cars of the same colour try to move one step if there is no car in front of it. Here, the model has self-organized in a somewhat geometric pattern where there are some traffic jams and some areas where cars can move at top speed.

Traffic forecasting may benefit from applications of chaos theory. Better predictions of when a congestion will occur would allow measures to be taken to disperse it before it would have occurred. Combining chaos theory principles with a few other methods has led to a more accurate short-term prediction model (see the plot of the BML traffic model at right).^[157]

Chaos theory has been applied to environmental water cycle data (also hydrological data), such as rainfall and streamflow.^[158] These studies have yielded controversial results, because the methods for detecting a chaotic signature are often relatively subjective. Early studies tended to "succeed" in finding chaos, whereas subsequent studies and meta-analyses called those studies into question and provided explanations for why these datasets are not likely to have low-dimension chaotic dynamics.^[159]

In art (predominately art theory) a possible postpostmodern era has been outlined with emphasis on multiple narratives and the notion that every fictional angle is a possibility. In part this is therefor of a bisociate (trissociative) discourse and can be explained within emphasis on an institutional interchange of subjectivistic agents.^[160]

Articles

• Sharkovskii, A.N. (1964). "Co-existence of cycles of a continuous mapping of the line into itself". Ukrainian Math. J. 16: 61–71.
• Li, T.Y.; Yorke, J.A. (1975). "Period Three Implies Chaos" (PDF). American Mathematical Monthly. 82 (10): 985–92. Bibcode:1975AmMM...82..985L. CiteSeerX 10.1.1.329.5038. doi:10.2307/2318254. JSTOR 2318254. Archived from the original (PDF) on 2009-12-29. Retrieved 2009-08-12.
• Alemansour, Hamed; Miandoab, Ehsan Maani; Pishkenari, Hossein Nejat (March 2017). "Effect of size on the chaotic behavior of nano resonators". Communications in Nonlinear Science and Numerical Simulation. 44: 495–505. Bibcode:2017CNSNS..44..495A. doi:10.1016/j.cnsns.2016.09.010.
• Crutchfield; Tucker; Morrison; J.D. Farmer; Packard; N.H.; Shaw; R.S (December 1986). "Chaos". Scientific American. 255 (6): 38–49 (bibliography p.136). Bibcode:1986SciAm.255d..38T. doi:10.1038/scientificamerican1286-46. Online version (Note: the volume and page citation cited for the online text differ from that cited here. The citation here is from a photocopy, which is consistent with other citations found online that don't provide article views. The online content is identical to the hardcopy text. Citation variations are related to country of publication).
• Kolyada, S.F. (2004). "Li-Yorke sensitivity and other concepts of chaos". Ukrainian Math. J. 56 (8): 1242–57. doi:10.1007/s11253-005-0055-4. S2CID 207251437.
• Day, R.H.; Pavlov, O.V. (2004). "Computing Economic Chaos". Computational Economics. 23 (4): 289–301. arXiv:2211.02441. doi:10.1023/B:CSEM.0000026787.81469.1f. S2CID 119972392. SSRN 806124.
• Strelioff, C.; Hübler, A. (2006). "Medium-Term Prediction of Chaos" (PDF). Phys. Rev. Lett. 96 (4): 044101. Bibcode:2006PhRvL..96d4101S. doi:10.1103/PhysRevLett.96.044101. PMID 16486826. 044101. Archived from the original (PDF) on 2013-04-26.
• Hübler, A.; Foster, G.; Phelps, K. (2007). "Managing Chaos: Thinking out of the Box" (PDF). Complexity. 12 (3): 10–13. Bibcode:2007Cmplx..12c..10H. doi:10.1002/cplx.20159. Archived from the original (PDF) on 2012-10-30. Retrieved 2011-07-17.
• Motter, Adilson E.; Campbell, David K. (2013). "Chaos at 50". Physics Today. 66 (5): 27. arXiv:1306.5777. Bibcode:2013PhT....66e..27M. doi:10.1063/PT.3.1977. S2CID 54005470.

Textbooks

• Alligood, K.T.; Sauer, T.; Yorke, J.A. (1997). Chaos: an introduction to dynamical systems. Springer-Verlag. ISBN 978-0-387-94677-1.
• Baker, G. L. (1996). Chaos, Scattering and Statistical Mechanics. Cambridge University Press. ISBN 978-0-521-39511-3.
• Badii, R.; Politi A. (1997). Complexity: hierarchical structures and scaling in physics. Cambridge University Press. ISBN 978-0-521-66385-4.
• Collet, Pierre; Eckmann, Jean-Pierre (1980). Iterated Maps on the Interval as Dynamical Systems. Birkhauser. ISBN 978-0-8176-4926-5.
• Devaney, Robert L. (2003). An Introduction to Chaotic Dynamical Systems (2nd ed.). Westview Press. ISBN 978-0-8133-4085-2.
• Robinson, Clark (1995). Dynamical systems: Stability, symbolic dynamics, and chaos. CRC Press. ISBN 0-8493-8493-1.
• Feldman, D. P. (2012). Chaos and Fractals: An Elementary Introduction. Oxford University Press. ISBN 978-0-19-956644-0. Archived from the original on 2019-12-31. Retrieved 2016-12-29.
• Gollub, J. P.; Baker, G. L. (1996). Chaotic dynamics. Cambridge University Press. ISBN 978-0-521-47685-0.
• Guckenheimer, John; Holmes, Philip (1983). Nonlinear Oscillations, Dynamical Systems, and Bifurcations of Vector Fields. Springer-Verlag. ISBN 978-0-387-90819-9.
• Gulick, Denny (1992). Encounters with Chaos. McGraw-Hill. ISBN 978-0-07-025203-5.
• Gutzwiller, Martin (1990). Chaos in Classical and Quantum Mechanics. Springer-Verlag. ISBN 978-0-387-97173-5.
• Hoover, William Graham (2001) [1999]. Time Reversibility, Computer Simulation, and Chaos. World Scientific. ISBN 978-981-02-4073-8.
• Kautz, Richard (2011). Chaos: The Science of Predictable Random Motion. Oxford University Press. ISBN 978-0-19-959458-0.
• Kiel, L. Douglas; Elliott, Euel W. (1997). Chaos Theory in the Social Sciences. Perseus Publishing. ISBN 978-0-472-08472-2.
• Moon, Francis (1990). Chaotic and Fractal Dynamics. Springer-Verlag. ISBN 978-0-471-54571-2.
• Orlando, Giuseppe; Pisarchick, Alexander; Stoop, Ruedi (2021). Nonlinearities in Economics. Dynamic Modeling and Econometrics in Economics and Finance. Vol. 29. doi:10.1007/978-3-030-70982-2. ISBN 978-3-030-70981-5. S2CID 239756912.
• Ott, Edward (2002). Chaos in Dynamical Systems. Cambridge University Press. ISBN 978-0-521-01084-9.
• Strogatz, Steven (2000). Nonlinear Dynamics and Chaos. Perseus Publishing. ISBN 978-0-7382-0453-6.
• Sprott, Julien Clinton (2003). Chaos and Time-Series Analysis. Oxford University Press. ISBN 978-0-19-850840-3.
• Tél, Tamás; Gruiz, Márton (2006). Chaotic dynamics: An introduction based on classical mechanics. Cambridge University Press. ISBN 978-0-521-83912-9.
• Teschl, Gerald (2012). Ordinary Differential Equations and Dynamical Systems. Providence: American Mathematical Society. ISBN 978-0-8218-8328-0.
• Thompson JM, Stewart HB (2001). Nonlinear Dynamics And Chaos. John Wiley and Sons Ltd. ISBN 978-0-471-87645-8.
• Tufillaro; Reilly (1992). An experimental approach to nonlinear dynamics and chaos. Vol. 61. Addison-Wesley. p. 958. Bibcode:1993AmJPh..61..958T. doi:10.1119/1.17380. ISBN 978-0-201-55441-0. {{cite book}}: |journal= ignored (help)
• Wiggins, Stephen (2003). Introduction to Applied Dynamical Systems and Chaos. Springer. ISBN 978-0-387-00177-7.
• Zaslavsky, George M. (2005). Hamiltonian Chaos and Fractional Dynamics. Oxford University Press. ISBN 978-0-19-852604-9.

Semitechnical and popular works

• Christophe Letellier, Chaos in Nature, World Scientific Publishing Company, 2012, ISBN 978-981-4374-42-2.
• Abraham, Ralph H.; Ueda, Yoshisuke, eds. (2000). The Chaos Avant-Garde: Memoirs of the Early Days of Chaos Theory. World Scientific Series on Nonlinear Science Series A. Vol. 39. World Scientific. Bibcode:2000cagm.book.....A. doi:10.1142/4510. ISBN 978-981-238-647-2.
• Barnsley, Michael F. (2000). Fractals Everywhere. Morgan Kaufmann. ISBN 978-0-12-079069-2.
• Bird, Richard J. (2003). Chaos and Life: Complexity and Order in Evolution and Thought. Columbia University Press. ISBN 978-0-231-12662-5.
• John Briggs and David Peat, Turbulent Mirror: : An Illustrated Guide to Chaos Theory and the Science of Wholeness, Harper Perennial 1990, 224 pp.
• John Briggs and David Peat, Seven Life Lessons of Chaos: Spiritual Wisdom from the Science of Change, Harper Perennial 2000, 224 pp.
• Cunningham, Lawrence A. (1994). "From Random Walks to Chaotic Crashes: The Linear Genealogy of the Efficient Capital Market Hypothesis". George Washington Law Review. 62: 546.
• Predrag Cvitanović, Universality in Chaos, Adam Hilger 1989, 648 pp.
• Leon Glass and Michael C. Mackey, From Clocks to Chaos: The Rhythms of Life, Princeton University Press 1988, 272 pp.
• James Gleick, Chaos: Making a New Science, New York: Penguin, 1988. 368 pp.
• John Gribbin. Deep Simplicity. Penguin Press Science. Penguin Books.
• L Douglas Kiel, Euel W Elliott (ed.), Chaos Theory in the Social Sciences: Foundations and Applications, University of Michigan Press, 1997, 360 pp.
• Arvind Kumar, Chaos, Fractals and Self-Organisation; New Perspectives on Complexity in Nature , National Book Trust, 2003.
• Hans Lauwerier, Fractals, Princeton University Press, 1991.
• Edward Lorenz, The Essence of Chaos, University of Washington Press, 1996.
• Marshall, Alan (2002). The Unity of Nature - Wholeness and Disintegration in Ecology and Science. doi:10.1142/9781860949548. ISBN 9781860949548.
• David Peak and Michael Frame, Chaos Under Control: The Art and Science of Complexity, Freeman, 1994.
• Heinz-Otto Peitgen and Dietmar Saupe (Eds.), The Science of Fractal Images, Springer 1988, 312 pp.
• Nuria Perpinya, Caos, virus, calma. La Teoría del Caos aplicada al desórden artístico, social y político, Páginas de Espuma, 2021.
• Clifford A. Pickover, Computers, Pattern, Chaos, and Beauty: Graphics from an Unseen World , St Martins Pr 1991.
• Clifford A. Pickover, Chaos in Wonderland: Visual Adventures in a Fractal World, St Martins Pr 1994.
• Ilya Prigogine and Isabelle Stengers, Order Out of Chaos, Bantam 1984.
• Peitgen, Heinz-Otto; Richter, Peter H. (1986). The Beauty of Fractals. doi:10.1007/978-3-642-61717-1. ISBN 978-3-642-61719-5.
• David Ruelle, Chance and Chaos, Princeton University Press 1993.
• Ivars Peterson, Newton's Clock: Chaos in the Solar System, Freeman, 1993.
• Ian Roulstone; John Norbury (2013). Invisible in the Storm: the role of mathematics in understanding weather. Princeton University Press. ISBN 978-0691152721.
• Ruelle, D. (1989). Chaotic Evolution and Strange Attractors. doi:10.1017/CBO9780511608773. ISBN 9780521362726.
• Manfred Schroeder, Fractals, Chaos, and Power Laws, Freeman, 1991.
• Smith, Peter (1998). Explaining Chaos. doi:10.1017/CBO9780511554544. ISBN 9780511554544.
• Ian Stewart, Does God Play Dice?: The Mathematics of Chaos , Blackwell Publishers, 1990.
• Steven Strogatz, Sync: The emerging science of spontaneous order, Hyperion, 2003.
• Yoshisuke Ueda, The Road To Chaos, Aerial Pr, 1993.
• M. Mitchell Waldrop, Complexity : The Emerging Science at the Edge of Order and Chaos, Simon & Schuster, 1992.
• Antonio Sawaya, Financial Time Series Analysis : Chaos and Neurodynamics Approach, Lambert, 2012.