Where does fluctuation collide?

In statistical physics a fluctuation is a spontaneous, random deviation of a measurable quantity, such as a particle's position, a system's energy, or a local density, away from its average value at thermal equilibrium. Einstein's 1905 paper on the motion of small particles suspended in a liquid gave the first quantitative account of such deviations, linking the visible jittering of a suspended particle to the unseen thermal motion of the molecules around it. Callen and Welton's 1951 Irreversibility and generalized noise generalised the idea into the fluctuation-dissipation theorem, which ties the size of a system's equilibrium fluctuations to how strongly it dissipates energy when disturbed. The referent here is a microscopic, thermally driven wandering of a physical observable, scaled by temperature, and not the rise and fall of a national economy, the changing numbers of an animal population, or the scatter of mutation counts in a laboratory.[1][2]

Unit

One instance is a single departure of an observable from its mean, quantified by the mean-square deviation (variance) of that quantity

Recovery means

The quantity relaxes back toward its mean, and the rate of that return is tied to the fluctuation size by the fluctuation-dissipation theorem

Measured by

Measured as the variance or spectral density of the quantity, for example the mean-square displacement of a particle or the thermal noise power in a resistor

In ecology a population fluctuation is the change over time in the number of individuals of a species living in an area, whether roughly regular, like the linked rise and fall of predator and prey, or irregular. Elton's 1924 Periodic fluctuations in the numbers of animals: Their causes and effects was the first to document and name these swings across many species and to ask what drives them. Nicholson's 1933 The balance of animal populations then argued that density-dependent competition tends to regulate a population around a level, so that fluctuations reflect the pull between disturbance and regulation. The referent is the abundance of living organisms tracked across seasons or generations, and it is distinct from a thermal deviation in a physical system, a swing in economic output, or a variance in bacterial mutation counts.[3][4]

Unit

One instance is the count or density of a species measured at successive times, and a fluctuation is the difference between those counts

Recovery means

A regulated population tends to return toward a characteristic level after a disturbance, driven by density-dependent birth and death rates

Measured by

Measured by censusing or estimating population size over time and describing the amplitude, timing, and regularity of the changes

In macroeconomics, economic fluctuations are the recurrent, economy-wide movements of output, employment, and investment above and below their long-run trend, the pattern also called the business cycle. Frisch's 1933 Propagation problems and impulse problems in dynamic economics framed these movements as the response of a dynamic economy to repeated random shocks, separating the impulses that hit the system from the structure that carries and stretches them out over time. Kydland and Prescott's 1982 Time to build and aggregate fluctuations built this into a formal model in which real shocks to productivity drive the cycle, and set the modern meaning of aggregate fluctuations. The referent is the coordinated rise and fall of a whole economy measured in national accounts, not the thermal jitter of particles, the changing headcount of an animal species, or the scatter of mutation counts across cultures.[5][6]

Unit

One instance is the deviation of an aggregate, such as gross domestic product, from its trend in a given period, usually expressed as a percentage

Recovery means

Output tends to revert toward trend over the course of a cycle, though how fast and by what mechanism is itself contested

Measured by

Measured by detrending aggregate series and examining the size, duration, and co-movement of the cyclical components

In genetics the fluctuation test is an experiment that reads the wide variation in the number of resistant mutants found across many parallel cultures of bacteria as evidence about how mutations arise. Luria and Delbrück's 1943 Mutations of bacteria from virus sensitivity to virus resistance reasoned that if mutations occur at random before any exposure to a virus, then a mutation early in a culture's growth seeds a large clone, so mutant counts should vary enormously from culture to culture, whereas mutations triggered by the virus would give similar counts. The large variation they observed, the fluctuation, showed that the mutations were spontaneous and already present. Here the referent is the between-culture variance in mutant numbers used as a statistical signature, not a physical deviation, an economic cycle, or a change in a wild population's size.[7]

Unit

One instance is the count of resistant mutants in a single culture, and the test statistic is the variance of those counts across replicate cultures

Recovery means

Not applicable

Measured by

Measured by plating many parallel cultures on selective medium, counting resistant colonies, and comparing the variance to the mean to estimate the mutation rate