Deterministic Nature of Star Formation

• The topic is defined as the deterministic nature of star formation where physical laws like gravity, turbulence, and feedback govern star formation rates and mass distributions.
• Hierarchical fragmentation produces self-similar structures and power-law mass functions, linking large-scale gravitational instabilities to local turbulent density fields.
• Observations and simulations confirm that star formation laws derived from gravitational and turbulent dynamics yield reproducible scaling relations across cosmic environments.

Star formation proceeds through the conversion of interstellar gas into stars within galaxies and molecular clouds. The deterministic nature of star formation refers to the degree to which the global and local rates, spatial configuration, and mass distribution of newly formed stars are governed by physical laws specified by initial and boundary conditions, global dynamical instabilities, and the statistical properties of turbulence and self-gravity, rather than by stochastic or environment-driven fluctuations. Contemporary galaxy- and cloud-scale observations, theoretical models, and simulations converge on a picture where gravitational instability, turbulence, and feedback combine to yield highly reproducible statistical laws—such as power-law scaling relations and mass functions—with fixed exponents, normalizations, and spatial/temporal self-similarity across a wide range of environments. However, local stochasticity at the level of individual stars or clusters, and environmental modulations of key physical parameters, superpose measurable scatter atop the underlying deterministic trends.

1. Hierarchical Fragmentation and Self-Similar Structure

Star formation follows a hierarchical fragmentation paradigm spanning spatial scales from galactic disks to parsec-scale cores. The process is initiated at the Jeans scale, where gravitational instability in the galactic midplane sets the largest cloud complexes ($\lambda_J = \sqrt{\pi c_s^2 / (G\rho)}$, $M_J \sim 10^7\,M_\odot$ for Milky Way conditions). Successive fragmentation yields structures with decreasing mass and size, ultimately forming prestellar cores of a few $M_\odot$. Each level of the hierarchy (GMCs, clumps, cores) appears phenomenologically self-similar, with observed fractal clustering (fractal dimension $D_2\simeq1.3-1.6$), power-law two-point correlations ($\xi(r)\sim r^{-\gamma},\; \gamma\sim0.4-0.7$), and size distributions ($N(>R)\sim R^{-2.5}$) over spatial scales $1$–$10^3$ pc. Temporal self-similarity manifests in the age-separation relation, $\Delta t\propto \Delta x^{0.5}$, indicating that star formation proceeds on a local crossing time at each scale (Elmegreen, 2011).

This nested hierarchy imposes robust statistical order, generating mass functions with power-law slopes $dN/dM\propto M^{-2}$ for clusters and substructures—a direct outcome of constant mass per logarithmic bin in the self-similar cascade. Despite the deterministic ensemble properties, details of fragmentation and collapse within subregions remain susceptible to local stochasticity due to the exact realization of turbulent density fields, magnetic fluctuations, and feedback events, introducing modest deviations from the large-scale scaling laws.

2. Universal Star Formation Laws and Physical Predictors

A central outcome is that global and local star formation rates obey deterministic scaling relations set by a small set of physical variables. The areal or volumetric star formation laws take the form

$$\Sigma_\mathrm{SFR} = \epsilon_\mathrm{ff}\,\frac{\Sigma_\mathrm{gas}}{t_\mathrm{ff}}$$

where $t_\mathrm{ff}$ is the local free-fall time at the mean volume density, and $\epsilon_\mathrm{ff}\simeq1\%$ is the empirically and theoretically determined efficiency per free-fall time (Federrath, 2013, Elmegreen, 2015, Krumholz et al., 2011, Lada et al., 2011). This law is applicable across scales from individual GMCs (using YSO counts) to kiloparsec galaxy patches and entire galaxies (using integrated SFR tracers).

Dimensional analysis using the Vaschy–Buckingham $\Pi$-theorem further supports a deterministic, scale-free relation

$$\dot{\Sigma}_\star = \epsilon\,\sqrt{\frac{G}{L}}\,\Sigma_\mathrm{gas}^{3/2}$$

where $L$ is the characteristic integration scale (disk height, Toomre length, Jeans length), and $\epsilon$ depends weakly on turbulent and magnetic parameters (Escala, 2014). Specializations of $L$ recover the canonical Kennicutt–Schmidt ($$\Sigma_\mathrm{SFR}\propto\Sigma_\mathrm{gas}^{1.5}$$), orbital time, and free-fall time formulations, all with small scatter.

The conversion of gas into stars is tightly regulated by the mass of gas above a threshold density ($n({\rm H}_2)\sim10^4\,\mathrm{cm}^{-3}$ or $A_V\gtrsim8$ mag), with observed linear relations between SFR and dense or gravitationally bound gas mass extending over up to nine orders of magnitude in mass (Jiao et al., 12 May 2025, Lada et al., 2011). Empirically, SFR $= \epsilon_\mathrm{bound}\,M_\mathrm{bound}$ with $$\epsilon_\mathrm{bound}\simeq4\times10^{-3}\,\mathrm{Myr}^{-1}$$, regardless of environment or turbulent intensity (Jiao et al., 12 May 2025).

3. Deterministic Control by Gravity, Turbulence, and Feedback

Simulations and analytic models demonstrate that the accelerating, nonlinear time-dependence of star formation rate ($M_*(t)\propto t^2$, hence SFR~$t$) arises not from stochastic turbulence, but from deterministic, gravity-dominated collapse and the emergence of quasi-universal density (power-law tail with $P(\rho)\propto\rho^{-1.8}$) and velocity structures in collapsing regions (Lee et al., 2014, Caldwell et al., 2017). Gravitationally regulated turbulence steepens local density PDFs, modulates the size–linewidth relation, and leads to spatial localization of star formation—the detailed spatial and temporal statistics (e.g., $t^2$ growth, attractor density profiles, slow radial expansion) are robust to variations in local turbulence and magnetization.

At the galaxy scale, deterministic Toomre instability ($Q\approx1$), the largest unstable mass-scale ($M_\mathrm{rot}\propto\Sigma^3/\kappa^4$), and a log–normal turbulent density PDF self-consistently fix the mean star formation rate, with the level of turbulence ($\sigma$) set by feedback-regulated, gravitational equilibrium (Escala, 2011). The deterministic law

$$\Sigma_{\rm SFR} = \epsilon_{\rm ff}\, \frac{\Sigma}{t_{\rm ff}(\rho_0)} \exp\!\left(\frac{3}{8}\sigma_s^2\right) \frac{1}{2} \left[1+\mathrm{erf}\left(\frac{\sigma_s^2-s_{\rm crit}}{\sqrt{8\sigma_s^2}}\right)\right]$$

directly predicts galaxy SFRs from physically observed quantities, with negligible room for intrinsic stochasticity outside of explicit input parameter variations.

4. Observational Evidence for Deterministic Star Formation

Large-sample galaxy studies from SDSS show that >90% of systems—across a range of mass, metallicity, and morphology—occupy a narrow "ageing sequence" in the SSFR–color and SSFR–H$\alpha$ plane, with low scatter (σ~0.2–0.3dex) around monotonic, deterministic relations (Casado et al., 2015). Starbursts, compacts, and field galaxies alike, when age-corrected or burst-corrected to a common evolutionary stage, display nearly linear SFR–stellar mass relations with low intrinsic scatter (σ~0.2dex), implying that apparent stochasticity is actually an artifact of phase-mixing and sampling varying evolutionary points (Izotov et al., 2016).

On molecular cloud scales, variability in SFR at fixed gas mass is entirely removed when one selects gas by a physically defined, environmental threshold—specifically, the break in the column density PDF where the power-law tail indicative of self-gravitational binding sets in (Jiao et al., 12 May 2025). This approach recovers a universal SFR–$M_\mathrm{bound}$ relation even in strongly turbulent or galaxy-center environments where classic $A_V$-based thresholds fail.

5. Physical Origin and Scope of Residual Variance

While the deterministic skeleton of star formation is now firmly established, real systems exhibit residual scatter at the level of a factor 2–3 (σ~0.3–0.5dex) in SFR at fixed gas properties (Federrath, 2013). The dominant sources of this scatter are systematic, physically interpretable variations: the Mach number of turbulence, star formation efficiency per free-fall, magnetic-to-thermal pressure, virial parameter, and forcing parameter (solenoidal versus compressive). These parameters modulate the normalization but not the exponents in all main star formation laws. Environmental factors, such as cluster versus field location, act on the incidence of quenching and rejuvenation, but only rare systems are seen to deviate radically from the deterministic sequences (notably post-merger bursts and cluster-quenched galaxies) (Casado et al., 2015, Elmegreen, 2015). On sub-kiloparsec scales and protostellar disks, secular plus stochastic accretion variability (e.g., FUor/EXor outbursts) is superposed on deterministic envelope infall but does not dominate the long-term SFR statistics (Johnstone, 2017).

6. Determinism at Multiple Scales and Cosmic Epochs

The local physical ingredients—gravity, turbulence, and MHD—fix the core mass function, collapse time, and characteristic fragmentation mass nearly independently of global cosmic epoch or galactic environment (Elmegreen, 2015). Only the rate at which dense, self-gravitating cores form (i.e., gas supply, pressure, global turbulence) and the maximum cluster mass reflect external variables that evolve with redshift, environment, and feedback conditions. Even at high redshift, the increased gas richness, ISM pressure, and turbulence lead to deterministic shifts in the normalization and cluster mass scale, but do not alter the form of the fundamental star formation laws or the deterministic nature of core formation and collapse.

7. Synthesis: Statistical Laws and Predictive Principle

The deterministic nature of star formation is encapsulated in the existence of universal, scale-free power-laws, as well as the tight coupling of SFR to local gas properties—either through physical thresholds (M$_\mathrm{bound}$), volumetric collapse rates (Σ$_{\rm gas}$/t$_{\rm ff}$), or dimensionally consistent scaling (Σ$_{\rm gas}^{3/2}$). Provided that the macroscopic boundary conditions (e.g., mean density, pressure, Mach number, magnetic support) are specified, deterministic theory predicts the statistical properties—mass functions, SFRs, spatial-power spectra, and age distributions—to high accuracy. Stochasticity enters only as higher-order variance about these scaling relations and as phase- and episodic effects on small physical or temporal scales (Elmegreen, 2011, Escala, 2014).

The deterministic framework has been validated observationally and numerically in Milky Way GMCs, external galaxies, and high-redshift systems, supporting its status as the fundamental organizing principle for star formation in the universe. Scatter and exceptions are attributable to traceable, physically interpretable sources, and do not invalidate the homogeneity of the ensemble laws (Jiao et al., 12 May 2025, Krumholz et al., 2011, Federrath, 2013).