Fact-checked by Grok 2 weeks ago

Primitive equations

The primitive equations are a set of five coupled partial differential equations that describe the large-scale dynamics of the atmosphere and oceans, comprising three momentum equations for fluid velocity components, a continuity equation for mass conservation, and a thermodynamic equation for energy balance, often under the hydrostatic approximation for vertical motion.^[1]^[2]^[3] Derived from the fundamental conservation laws of mass, momentum, and energy applied to a rotating fluid system, these equations use primitive variables such as horizontal and vertical velocities (u, v, w), density (ρ), pressure (p), and potential temperature (θ), with the material derivative capturing advective changes along fluid parcels.^[3] The momentum equations balance local acceleration against forces including the Coriolis effect (2Ω × v), pressure gradients (-∇p/ρ), gravity (-∇Φ), and viscous dissipation, while the continuity equation enforces ∇ · (ρv) = 0 in compressible form, and the thermodynamic equation conserves potential temperature adiabatically as Dθ/Dt = 0 or includes diabatic heating terms.^[1]^[3] The hydrostatic approximation, ∂p/∂z = -ρg, replaces the full vertical momentum equation due to the dominance of gravity over vertical accelerations in large-scale flows, enabling efficient numerical solutions while retaining gravity wave propagation.^[2]^[1] Since their formulation in the mid-20th century, primitive equations have formed the core of global circulation models for numerical weather prediction and climate simulation, as demonstrated in early baroclinic experiments that captured atmospheric variability.^[4] In oceanography, they underpin models like NEMO and ROMS, incorporating a nonlinear equation of state to simulate density-driven circulation and tides.^[5]^[2] Their spectral and finite-difference discretizations, as in NOAA's dynamical cores, support long-term integrations essential for forecasting and understanding phenomena like cyclones and ocean currents.^[6]

Introduction

Definition and Scope

The primitive equations constitute a foundational set of five nonlinear partial differential equations in geophysical fluid dynamics, approximating the compressible Navier-Stokes equations for large-scale motions in rotating, stratified fluids such as the atmosphere and oceans. These equations encompass the continuity equation for mass conservation, three momentum equations describing horizontal and vertical velocity components, and a thermodynamic energy equation that governs the evolution of potential temperature or density perturbations. In moist atmospheric applications, an additional equation for water vapor conservation is included, while oceanic variants incorporate salinity to account for density variations due to thermohaline effects. This framework arises from key simplifications, including the hydrostatic approximation—which balances vertical pressure gradients with gravitational forces, effectively setting vertical accelerations to negligible values—and the Boussinesq approximation, which treats fluid density as constant except in buoyancy terms to filter out acoustic waves.^[7]^[8]^[9] The scope of the primitive equations is primarily within geophysical fluid dynamics, where they serve as the core model for global circulation in numerical weather prediction and climate simulations, capturing synoptic- to planetary-scale phenomena without the further filtering applied in reduced models like the quasi-geostrophic equations. Unlike those filtered approximations, which assume near-geostrophic balance for mid-latitude flows, the primitive equations retain full nonlinearity to describe a broader range of dynamics, including tropical circulations and frontal systems. They apply to both atmospheric contexts—encompassing dry, adiabatic processes or moist convection with phase changes—and oceanic regimes, where salinity gradients drive deep circulation patterns. Formulated in a spherical coordinate system on a rotating Earth, the equations neglect molecular viscosity, relying instead on parameterized subgrid-scale turbulence for realistic dissipation, and incorporate the Coriolis effect as a primary restoring force alongside pressure gradients.^[7]^[10]^[11] A brief overview of the equation structure highlights the horizontal momentum equations for zonal (u) and meridional (v) velocities, which include advective, Coriolis, and pressure gradient terms; the vertical momentum equation, simplified under the hydrostatic approximation (∂p/∂z = -ρg), neglecting vertical accelerations, with vertical velocity w determined from the continuity equation; the continuity equation for mass conservation, which under the Boussinesq approximation enforces divergence-free flow (∇ · v = 0); and the thermodynamic equation tracking changes in potential temperature (for atmospheres) or buoyancy (for oceans) due to advection and diabatic heating. These components collectively enable the simulation of balanced, three-dimensional flows while excluding fast internal gravity waves through the core approximations.^[7]^[8]

Historical Development

The origins of the primitive equations lie in the foundational work of Vilhelm Bjerknes in the late 19th and early 20th centuries. In 1898, Bjerknes formulated circulation theorems that integrated hydrodynamics and thermodynamics to describe large-scale atmospheric motions, providing the theoretical basis for predicting weather through mathematical equations.^[12] By 1904, he expanded this into a comprehensive model comprising seven equations—governing momentum, mass conservation, energy, and state relations—that essentially constituted the primitive equations for geophysical fluid dynamics.^[13] These efforts emphasized the potential for numerical integration of atmospheric circulation, though computational limitations delayed practical implementation. In the 1930s, Carl-Gustaf Rossby advanced the framework by introducing the beta-plane approximation, which simplified the latitudinal variation of the Coriolis parameter to better model mid-latitude wave propagation and planetary-scale flows.^[14] The mid-20th century saw the formalization of primitive equations for numerical weather prediction (NWP), driven by post-World War II computational advances. In 1948, Jule Charney published a paper on the scales of atmospheric motions, demonstrating the feasibility of numerical weather prediction using filtered versions of the primitive equations, which led to his leadership of the IAS meteorology project, galvanizing efforts at the Institute for Advanced Study to develop predictive models.^[15] Meanwhile, researchers like Jerome Namias pioneered empirical extended-range forecasting techniques in the 1940s. Arnt Eliassen co-authored a 1949 paper with Charney on numerical methods for predicting mid-latitude westerlies perturbations using simplified primitive forms, bridging theoretical foundations to practical NWP in the 1940s and 1950s.^[16] A pivotal validation came in 1956 through Norman Phillips' numerical experiments, which employed a two-layer primitive equation model to simulate baroclinic instability and demonstrate realistic general circulation patterns, confirming the equations' utility for atmospheric modeling. Key milestones marked the integration of primitive equations into operational and research models. In 1963, Joseph Smagorinsky at the Geophysical Fluid Dynamics Laboratory (GFDL) implemented primitive equations in a nine-level global atmospheric model, enabling the first extended simulations of climate variability and solidifying their role in general circulation models (GCMs). Their application extended to oceanography in 1969, when Kirk Bryan developed numerical methods using primitive equations to study the world ocean's circulation, laying the groundwork for oceanic general circulation models (OGCMs). From rudimentary hand calculations in the early 20th century, primitive equations evolved into the core of digital NWP and GCMs by the 1960s, facilitated by electronic computers like ENIAC.^[17] This progression culminated in operational systems, such as the European Centre for Medium-Range Weather Forecasts' (ECMWF) Integrated Forecasting System (IFS), which has employed hydrostatic primitive equations since the organization's inception in the 1970s for global medium-range predictions.^[18] By the 1980s and 1990s, awareness of limitations in resolving small-scale or non-hydrostatic processes spurred hybrid formulations that augmented primitive equations with additional dynamical terms.^[2]

Physical Foundations

Key Forces in Geophysical Fluid Motion

In geophysical fluid dynamics, the primitive equations capture the motion of rotating planetary atmospheres and oceans through a balance of key forces in the momentum equation. The primary horizontal forces include advection, which represents the nonlinear transport of momentum by the fluid velocity itself, leading to interactions across scales. The Coriolis force, given by \mathbf{f} \times \mathbf{v} where f = 2 \Omega \sin \phi (with \Omega as Earth's angular velocity and \phi as latitude), deflects moving parcels to the right in the Northern Hemisphere and arises from the rotating reference frame, dominating large-scale flows. The pressure gradient force, -\nabla p / \rho, drives the acceleration of fluid parcels toward lower pressure regions and often balances the Coriolis force in geostrophic equilibrium. Friction, typically viscous dissipation, is neglected in the inviscid primitive equations but incorporated via parameterizations for boundary layers, such as surface wind stress in the atmosphere or bottom drag in the ocean.^[3]^[19] Vertically, gravity plays a central role through the hydrostatic approximation, where the vertical momentum balance yields \partial p / \partial z = -\rho [g](/page/Gravitational_acceleration) (with g as gravitational acceleration), assuming negligible vertical accelerations compared to this force. Buoyancy effects emerge from density variations \rho, influencing vertical stability and motion; in the atmosphere, these arise primarily from temperature differences, while in the ocean, salinity gradients contribute significantly to the equation of state, altering buoyancy via terms like \beta (S - S_0) in density perturbations (where \beta is the saline contraction coefficient and S is salinity). Thermodynamic drivers, such as radiative heating, cooling, and latent heat release from condensation or evaporation, modify the potential temperature \theta through diabatic terms, fueling convective and synoptic-scale circulations in moist atmospheres.^[3]^[19]^[20] The primitive equations are valid for synoptic scales of $10^2 to $10^4 km, where the Rossby number Ro = U / (f L) < 1 (with U as characteristic velocity and L as length scale) ensures Coriolis dominance over inertial terms, justifying the approximations. In oceanic contexts, tidal forces act as external drivers, inducing mixing and currents that are often parameterized separately from the core primitive dynamics, while salinity gradients enhance buoyancy-driven thermohaline circulation. These forces manifest in the general momentum equation as terms balancing local accelerations, providing the physical basis for modeling large-scale geophysical flows.^[3]^[19]^[20]

Underlying Conservation Laws

The primitive equations in geophysical fluid dynamics are derived from fundamental conservation principles that ensure physical consistency in modeling large-scale atmospheric and oceanic flows. These laws, including conservation of mass, momentum, and energy, form the theoretical foundation, with approximations tailored to the scales of interest such as hydrostatic balance and the neglect of viscous effects. Conservation of mass is governed by the continuity equation, which states that the divergence of the mass flux vanishes:

\nabla \cdot (\rho \mathbf{v}) = 0,

where \rho is the fluid density and \mathbf{v} is the velocity vector. This equation ensures that the total mass within any fluid volume remains constant in the absence of sources or sinks. In many applications of primitive equations, particularly for oceanic flows, the Boussinesq approximation is invoked, treating the fluid as incompressible by assuming constant reference density \rho_0, which simplifies the continuity equation to \nabla \cdot \mathbf{v} = 0. This approximation is valid when density variations are small compared to the mean density, as is typical in the ocean's interior.^[21]^[22] Conservation of momentum originates from the Navier-Stokes equations, which describe the rate of change of momentum as the sum of pressure gradients, gravitational forces, and other terms. In the primitive equations, viscosity is neglected to focus on inviscid dynamics at large scales, yielding:

\frac{D\mathbf{v}}{Dt} + f \mathbf{k} \times \mathbf{v} = -\frac{1}{\rho} \nabla p + \mathbf{g},

where f is the Coriolis parameter, p is pressure, and \mathbf{g} is gravity. The hydrostatic approximation further simplifies the vertical momentum balance by assuming vertical accelerations are negligible compared to gravitational and pressure forces, reducing it to \frac{\partial p}{\partial z} = -\rho g. This filtering preserves the essential dynamics while maintaining consistency with the full momentum conservation in horizontal directions.^[3]^[23] Energy conservation follows from the first law of thermodynamics, linking changes in internal energy to work done and heat addition. In primitive equations, this is often expressed in terms of potential temperature \theta, which accounts for compressibility effects in a stratified fluid:

\frac{D\theta}{Dt} = Q,

where Q represents diabatic heating sources such as radiation or latent heat release. Potential temperature \theta is conserved for adiabatic processes, making it a key prognostic variable that ensures the model respects thermodynamic constraints. The hydrostatic approximation, by neglecting vertical acceleration, does not alter the total mechanical energy conservation when the equations are properly formulated, as the vertical kinetic energy is small and the potential energy remains balanced by pressure work.^[3]^[24] Angular momentum and vorticity conservation are implicitly embedded in the primitive equations through the Coriolis term, which arises from Earth's rotation. While exact angular momentum is not conserved due to spherical geometry and frictionless assumptions, the equations approximately conserve potential vorticity q = (\mathbf{f} + \zeta) \cdot \nabla \theta / \rho, where \zeta is relative vorticity and \mathbf{f} is the planetary vorticity vector. This quasi-conservation is crucial for understanding large-scale wave propagation and stability in geophysical flows.^[25] In atmospheric primitive equations, conservation of water substance is handled through prognostic equations for mixing ratios of water vapor q_v, cloud water q_c, and rain q_r, each following:

\frac{D q_i}{Dt} = S_i,

where S_i denotes microphysical sources and sinks from phase changes. These ensure that total water mass is conserved globally, modulo external fluxes, and are essential for simulating moist processes like precipitation.^[26]

Mathematical Forms

General Primitive Equations

The primitive equations constitute the foundational system of partial differential equations (PDEs) describing the motion of geophysical fluids, such as the atmosphere and oceans, under the approximations of hydrostatic balance, shallow fluid depth, and the traditional Coriolis force. These equations retain the full nonlinearity of the momentum balance while filtering out acoustic waves through the hydrostatic approximation, making them suitable for large-scale flows where horizontal scales vastly exceed vertical ones. In their general, coordinate-independent form, they express conservation principles in vector notation, applicable to both Cartesian and spherical geometries, though often simplified to local tangent-plane approximations for analysis.^[27] The core equations begin with the horizontal momentum equation, which governs the evolution of the horizontal velocity vector \mathbf{u}_h = (u, v):

\frac{D \mathbf{u}_h}{Dt} + f \hat{\mathbf{k}} \times \mathbf{u}_h = -\frac{1}{\rho} \nabla_h p,

where \frac{D}{Dt} = \frac{\partial}{\partial t} + \mathbf{u} \cdot \nabla is the substantial derivative, f = 2 \Omega \sin \phi is the Coriolis parameter (with \Omega the planetary rotation rate and \phi latitude), \hat{\mathbf{k}} is the unit vector in the vertical direction, \rho is density, p is pressure, and \nabla_h denotes the horizontal gradient. This form incorporates the traditional approximation, neglecting Coriolis terms involving vertical velocity. The vertical momentum balance is hydrostatic:

\frac{\partial p}{\partial z} = -\rho g,

where g is gravitational acceleration and z is the vertical coordinate, assuming vertical accelerations are negligible compared to gravity. Mass continuity follows from incompressibility in the Boussinesq approximation or density variations:

\nabla \cdot (\rho \mathbf{u}) = 0,

or equivalently \frac{D \rho}{Dt} + \rho \nabla \cdot \mathbf{u} = 0, where \mathbf{u} = (u, v, w) is the full velocity vector. The thermodynamic equation for dry, adiabatic flow conserves potential temperature \theta:

\frac{D \theta}{Dt} = 0,

with \theta = T (p_0 / p)^{R/c_p} for an ideal gas, where T is temperature, p_0 a reference pressure, R the gas constant, and c_p specific heat at constant pressure; diabatic sources like heating Q can be added as \frac{D \theta}{Dt} = \frac{Q}{\rho c_p \Pi}, with \Pi = (p_0 / p)^{R/c_p}. These equations are often analyzed in the f-plane approximation, treating f as constant, or the beta-plane, where f = f_0 + \beta y with \beta = \frac{2 \Omega \cos \phi_0}{a} ( a Earth's radius, y meridional distance), to capture mid-latitude variations in rotation effects.^[27]^[7] Extensions to moist atmospheres incorporate conservation of water vapor mixing ratio q_v:

\frac{D q_v}{Dt} = 0,

under adiabatic conditions, with precipitation terms added when saturation occurs, such as condensation rates in the thermodynamic equation; total water conservation may include liquid and ice phases for comprehensive models. In oceanic contexts, the equations adapt to nearly incompressible flow with variable density \rho(T, S, p) depending on temperature T, salinity S, and pressure, replacing \theta with a buoyancy variable b = -g \delta \rho / \rho_0 ( \rho_0 reference density) and enforcing strict incompressibility \nabla \cdot \mathbf{u} = 0, alongside \frac{D S}{Dt} = 0 for salinity; the momentum form becomes \frac{D \mathbf{u}}{Dt} + f \hat{\mathbf{k}} \times \mathbf{u} = -\nabla_h \phi + b \hat{\mathbf{k}}, with \phi geopotential and hydrostatic \frac{\partial \phi}{\partial z} = b. Dimensionless analysis highlights key balances through nondimensional numbers: the Rossby number Ro = U / (f_0 L) ( U velocity scale, L horizontal length) measures advection versus rotation, the Froude number Fr = U / \sqrt{g H} ( H vertical scale) assesses flow versus stratification, and the Ekman number Ek = \nu / (f_0 L^2) ( \nu viscosity) quantifies frictional effects, revealing regimes like geostrophic balance when Ro \ll 1. These scalings underscore the primitive equations' applicability to rotating, stratified flows without specifying vertical coordinates.^[27]^[7]

Pressure Coordinate Form

The pressure coordinate form of the primitive equations employs pressure p as the vertical coordinate in place of geometric height z, a choice well-suited to atmospheric dynamics where isobaric surfaces approximate horizontal layers. This transformation is derived from the general primitive equations by treating p as a function of position and time, p = p(x, y, z, t), and introducing the vertical velocity \omega = \frac{Dp}{Dt}, where \frac{D}{Dt} denotes the material derivative along fluid parcels. The geometric height is recovered through hydrostatic integration, \frac{\partial z}{\partial p} = -\frac{1}{\rho g}, yielding the geopotential \Phi = gz.^[3]^[28] In this coordinate system, the horizontal momentum equations take the form

\frac{Du}{Dt} - f v + \frac{\partial \Phi}{\partial x} = 0,

\frac{Dv}{Dt} + f u + \frac{\partial \Phi}{\partial y} = 0,

where u and v are the horizontal velocity components, f is the Coriolis parameter, and the material derivative expands to \frac{D}{Dt} = \frac{\partial}{\partial t} + u \frac{\partial}{\partial x} + v \frac{\partial}{\partial y} + \omega \frac{\partial}{\partial p}. The thermodynamic equation for conserved potential temperature \theta in dry adiabatic conditions is

\frac{D\theta}{Dt} = 0.

The continuity equation, reflecting mass conservation, simplifies to

\nabla_h \cdot \mathbf{V} + \frac{\partial \omega}{\partial p} = 0,

where \mathbf{V} = (u, v) and \nabla_h is the horizontal divergence operator on constant-pressure surfaces; this form arises directly from the transformation, eliminating explicit density \rho dependence.^[29]^[3] A key advantage of pressure coordinates is the elimination of surface pressure as a prognostic variable, since pressure levels remain fixed regardless of topography, facilitating computations on nearly horizontal isobaric surfaces prevalent in the atmosphere. However, the system exhibits a singularity at the top of the atmosphere where p = 0, complicating boundary conditions there, and proves less applicable to oceanic modeling due to the dominance of density-driven flows over pressure variations.^[28]^[3]

Sigma and Terrain-Following Coordinate Forms

The sigma coordinate system, introduced by Phillips in 1957, provides a terrain-following vertical coordinate for numerical models of atmospheric circulation, defined as \sigma = \frac{p - p_t}{p_s - p_t}, where p is the pressure at a point, p_s is the surface pressure, and p_t is the pressure at the model top (often set to zero for simplicity, yielding \sigma = p / p_s).^[30] This transformation maps the surface topography to \sigma = 1 and the top to \sigma = 0, facilitating the application of boundary conditions over irregular terrain without interpolating across steep slopes.^[31] The primitive equations in sigma coordinates are derived by transforming the general form from height (z) or pure pressure coordinates, incorporating metric terms that account for the sloping coordinate surfaces. The material derivative becomes \frac{D}{Dt} = \frac{\partial}{\partial t} + \mathbf{u} \cdot \nabla + \dot{\sigma} \frac{\partial}{\partial \sigma}, where \dot{\sigma} = D\sigma / Dt is the vertical velocity in sigma space and \nabla denotes the horizontal gradient at constant \sigma. The horizontal momentum equation includes additional terms for orographic slopes, such as \frac{\partial h}{\partial x} (where h is terrain height), yielding:

\frac{D u}{Dt} - f v = -\frac{\partial \Phi}{\partial x}\big|_{\sigma} + \text{metric terms involving } \frac{\partial h}{\partial x},

with a similar form for the v-component, where \Phi = gz is the geopotential and f is the Coriolis parameter; the geopotential gradient at constant \sigma incorporates effects from surface pressure and terrain variations.^[31] The continuity equation, ensuring mass conservation, is:

\frac{\partial p_s}{\partial t} + \nabla \cdot (p_s \mathbf{u}) + \frac{\partial (p_s \dot{\sigma})}{\partial \sigma} = 0,

and the hydrostatic relation adapts to \frac{\partial \Phi}{\partial \sigma} = -\frac{RT}{p_s} \frac{\partial p}{\partial \sigma}, with R the gas constant and T temperature.^[28] These forms resemble the pressure-coordinate equations but include factors of p_s and slope-induced corrections to handle variable surface pressure and topography accurately.^[32] Hybrid sigma-pressure coordinates extend the pure sigma system by blending terrain-following layers near the surface with pressure-like levels in the upper atmosphere, defined generally as \eta = A(p) p + B(\sigma) p_s, where A(p) and B(\sigma) are smoothing functions ensuring a smooth transition.^[33] This approach, developed by Simmons and Burridge in 1981, was adopted operationally by the European Centre for Medium-Range Weather Forecasts (ECMWF) in the 1980s to improve resolution and reduce errors in the stratosphere while maintaining sigma's benefits at lower levels.^[33] The equations follow a similar structure to sigma coordinates, with the vertical velocity \dot{\eta} replacing \dot{\sigma}, and metric terms adjusted for the hybrid definition to conserve energy and angular momentum in finite-difference schemes.^[33] In oceanic applications, sigma coordinates are adapted for terrain-following purposes using \sigma = \frac{z - \eta}{H + \eta}, where z is height, \eta is the free surface elevation, and H is the undisturbed depth, transforming the primitive equations to follow bathymetry in coastal and regional models.^[28] This form, applied in models like those for the South Atlantic circulation, includes contravariant velocities to account for curvilinear coordinate aspects and slope terms analogous to atmospheric orography, enhancing resolution near the sea floor.^[34] These coordinate systems offer key advantages for numerical simulations, including improved vertical resolution near the ground or sea floor to capture boundary layers, simplified handling of complex mountains and bathymetry without grid distortion, and a fixed computational domain that supports efficient differencing and modal representations.^[35] However, they introduce challenges like pressure gradient errors over steep terrain, which can be mitigated in hybrid forms.^[28]

Analysis and Solutions

Linearization of Primitive Equations

The linearization of the primitive equations is a standard approximation technique used to analyze small-amplitude perturbations, such as atmospheric or oceanic waves, by decomposing the governing variables into a basic state and deviations therefrom. This process begins by expressing each variable as the sum of a time-independent basic state and a small perturbation, denoted with primes: for horizontal velocity components, \mathbf{u} = \mathbf{U} + \mathbf{u}'; for geopotential height, \Phi = \Phi_0 + \Phi'; and for potential temperature, \theta = \theta_0 + \theta', where the basic state quantities like \mathbf{U}(z) (often a zonal flow) and \theta_0(z) satisfy the unperturbed equations. Quadratic and higher-order terms involving perturbations, such as \mathbf{u}' \cdot \nabla \mathbf{U} and \mathbf{u}' \cdot \nabla \mathbf{u}', are neglected under the assumption that perturbations are infinitesimal, leading to a linear system \mathcal{L}(\mathbf{u}') = 0, where \mathcal{L} is the linearized operator. This approximation simplifies the nonlinear primitive equations into a tractable form for studying wave propagation and stability without solving the full system.^[36] Common basic states include a resting fluid with stable stratification, characterized by the squared buoyancy frequency N^2 = \frac{g}{\theta_0} \frac{\partial \theta_0}{\partial z} > 0, or a sheared zonal flow U(z) in geostrophic and hydrostatic balance, often on an f-plane where the Coriolis parameter f is constant. These states represent idealized backgrounds like the mean zonal winds in the midlatitudes or stratified resting conditions in the ocean interior. Key assumptions underpinning the linearization are that perturbations are small (e.g., |\mathbf{u}'| \ll |\mathbf{U}|), the flow is adiabatic (no diabatic heating), moisture effects are absent, and the Boussinesq approximation holds to filter acoustic waves while retaining buoyancy effects. The f-plane further simplifies the Coriolis force to a constant, avoiding beta-effect complications initially.^[36]^[37] The resulting linearized system consists of the momentum equations, hydrostatic balance, continuity, and thermodynamic equation. In the horizontal momentum equations, for example, the zonal component is

\frac{\partial u'}{\partial t} + U \frac{\partial u'}{\partial x} + w' \frac{dU}{dz} - f v' = -\frac{\partial \Phi'}{\partial x},

with a similar form for the meridional component, incorporating advection by the basic flow and the shear term w' \frac{dU}{dz} if the basic state is sheared. The hydrostatic equation relates vertical structure to buoyancy:

\frac{\partial \Phi'}{\partial z} = \frac{g}{\theta_0} \theta'.

The continuity equation, under Boussinesq, is

\frac{\partial u'}{\partial x} + \frac{\partial v'}{\partial y} + \frac{\partial w'}{\partial z} = 0.

For the thermodynamic equation in a resting basic state, it simplifies to

\frac{\partial \theta'}{\partial t} + w' \frac{\partial \theta_0}{\partial z} = 0;

with shear in U(z), advection by U is added, but the vertical term remains tied to the basic stratification.^[36]^[38] In oceanic applications, the linearization differs primarily in the thermodynamic component, where density perturbations arise from both temperature and salinity fluctuations via the equation of state: \rho' = -\alpha T' + \beta S', with \alpha > 0 the thermal expansion coefficient and \beta > 0 the saline contraction coefficient. The buoyancy equation then becomes

\frac{\partial b'}{\partial t} + U \frac{\partial b'}{\partial x} + w' N^2 = 0,

where b' = -g \rho' / \rho_0 incorporates salinity advection \frac{\partial S'}{\partial t} + \mathbf{u} \cdot \nabla S_0 + \mathbf{u}' \cdot \nabla S' + w' \frac{\partial S_0}{\partial z} = 0 (neglecting diffusion), reflecting the active role of salinity in oceanic stratification and double-diffusive processes, unlike the atmosphere where only thermal effects dominate. This adjustment is crucial for modeling phenomena like thermohaline instabilities.^[37]^[39]

Solutions to Linearized Equations

Solutions to the linearized primitive equations are obtained through normal mode analysis, which assumes plane-wave solutions of the form \exp[i(kx + ly - \omega t)], where k and l are the zonal and meridional wavenumbers, respectively, and \omega is the angular frequency. Substituting this ansatz into the linearized system yields an eigenvalue problem, resulting in a dispersion relation \omega = \omega(k, l, m) that relates the frequency to the horizontal wavenumbers and the vertical wavenumber m.^[7] The dispersion relation reveals several key physical modes. Inertia-gravity waves dominate at high frequencies, with the approximate relation \omega \approx \pm \frac{[N](/page/N+) k_h}{m} in the hydrostatic and Boussinesq approximations, where N is the Brunt-Väisälä frequency and k_h = \sqrt{k^2 + l^2} is the horizontal wavenumber; these waves propagate energy both horizontally and vertically in stratified fluids. Rossby waves appear at low frequencies, approximated by \omega \approx -\frac{\beta k}{K^2}, where \beta = \frac{\partial f}{\partial y} is the meridional gradient of the Coriolis parameter f and K^2 = k^2 + l^2 + \frac{f^2}{gh} for shallow-water-like extensions; these waves are dispersive and propagate westward relative to the mean flow. Acoustic waves, which would otherwise contribute high-frequency sound-like modes, are effectively filtered out by the hydrostatic approximation inherent in the primitive equations.^[40]^[7]^[7] The vertical structure of these modes arises from separating the linearized equations into horizontal and vertical components, leading to a Sturm-Liouville eigenvalue problem for the vertical dependence. In vertically bounded domains with rigid lids, the solutions are oscillatory functions satisfying boundary conditions, such as \sin(m z) or more general eigenfunctions determined by the stratification profile. For unbounded or semi-infinite domains, evanescent modes decay exponentially away from forcing regions, with structures like \exp(-m z). In equatorial dynamics on a \beta-plane, the meridional structure often involves Hermite functions as solutions to the parabolic cylinder equation, capturing trapped wave behavior near the equator.^[7]^[7]^[41] Instability analysis of the linearized system identifies regions of exponential growth when the imaginary part of \omega is positive, particularly for baroclinic instabilities driven by vertical shear. The Charney-Stern-Pedlosky criteria provide necessary conditions for such instability: the meridional gradient of quasigeostrophic potential vorticity must change sign within the domain, or it must oppose the surface potential vorticity gradient at least at one boundary. These criteria, derived from energy considerations in the linearized framework, explain the development of synoptic-scale disturbances in midlatitude jets.^[42]^[43]^[44] In oceanic contexts, the linearized primitive equations support internal waves following the same inertia-gravity dispersion relation, enabling energy propagation along characteristics in density-stratified interiors. Boundary-trapped Kelvin waves emerge as non-dispersive solutions along coastlines or the equator, with phase speed c = \sqrt{g' H} for reduced gravity g' and equivalent depth H, propagating without transverse velocity and decaying offshore due to rotation.^[41]^[40] These analytical solutions hold under the assumptions of small-amplitude perturbations and linearization around a basic state, limiting their applicability to weakly nonlinear or transitional regimes where higher-order effects become significant.^[7]

References

[1]
The Primitive Equations
The primitive eqs. are comprised of five coupled eqs.: three momentum eq., the continuity eq. and the thermodynamic eq.
[2]
Primitive Equation - an overview | ScienceDirect Topics
Primitive equations refer to a set of theoretical or numerical formulations in meteorological dynamics that use velocity components as flow variables, ...
[3]
[PDF] 2 The equations governing atmospheric flow. - Staff
2.5 Summary of the primitive equations. The above equations that we have derived for atmospheric flow are often known as the primitive equations. The ...
[4]
GENERAL CIRCULATION EXPERIMENTS WITH THE PRIMITIVE ...
An extended period numerical integration of a baroclinic primitive equation model has been made for the simulation and the study of the dynamics of the ...
[5]
Primitive Equations - NEMO ocean model
The ocean is a fluid that can be described to a good approximation by the primitive equations, $ ie$ the Navier-Stokes equations along with a nonlinear ...
[6]
[PDF] The Spectral Dynamical Core | NOAA
These are referred to as the primitive equations in the traditional approxi- mation. Our equation set now includes 4 prognostic equations, for (u, v, ρ, T),.
[7]
[PDF] Fundamentals of Geophysical Fluid Dynamics - UCLA
Mar 14, 2005 · 3D stably stratified fluid dynamics in, say, the Boussinesq or Primitive ... Equations (as well as the 3D Primitive and Boussinesq Equations).
[8]
Primitive Equation - an overview | ScienceDirect Topics
The primitive equations are based on the so-called hydrostatic approximation, in which the conservation of momentum in the vertical direction is replaced by the ...
[9]
Global Well-Posedness of the Ocean Primitive Equations with ...
Jun 11, 2021 · 2 The Primitive Equations. The primitive equations are a central set of model equations for atmosphere and ocean dynamics. These equations ...<|control11|><|separator|>
[10]
[PDF] arXiv:2211.09573v1 [math.AP] 17 Nov 2022
Nov 17, 2022 · The 3D primitive equations are used in most geophysical fluid models to approximate the large scale oceanic and atmospheric dynamics. We prove ...
[11]
https://www.sciencedirect.com/science/article/pii/S1874579205800096
[12]
[PDF] Weather Forecasting: - University College Dublin
Vilhelm Bjerknes (1862–1951). 40. Page 70. Vilhelm Bjerknes (1862 ... – 1898: Circulation theorems. – 1904 ... • Plan A: Integrate the Primitive Equations.
[13]
https://oxfordre.com/climatescience/display/10.1093/acrefore/9780190228620.001.0001/acrefore-9780190228620-e-52
[14]
[PDF] ROSSBY WAVES - Geophysical Fluid Dynamics Laboratory
Carl-Gustav Rossby, working at MIT in 1939, introduced the useful approximation for middle latitudes, known as the 'beta-plane'. It approximates the.Missing: Gustaf | Show results with:Gustaf
[15]
The birth of numerical weather prediction - Tellus B
The further developments resulting in the use of the primitive equations and the formulation of medium-range forecasting models are not included in the paper.
[16]
operational numerical weather prediction - AMS Journals
In the summer of 1948 help arrived in the form of. Jule Charney, fresh from a fellowship year in Norway where he developed a set of predictive equations ...Missing: seminar | Show results with:seminar
[17]
The History of Numerical Weather Prediction - NOAA
Oct 31, 2023 · A three-layer hemispheric model was introduced in 1962, and a six-layer primitive equation model appeared in 1966. Additional atmospheric ...
[18]
[PDF] PART III: DYNAMICS AND NUMERICAL PROCEDURES - ECMWF
In IFS, the horizontal wind, the virtual temperature and the surface ... A stable numerical integration scheme for the primitive meteorological equations.
[19]
[PDF] Oceanic Dynamical and Material Equations - ucla.edu
Primitive Equations with “Traditional” Coriolis Approximation: These are based on a hydrostatic vertical momentum balance and Ω ≈ f(ϕ)z. They comprise the fluid ...
[20]
Geophysical and astrophysical fluid dynamics beyond the traditional ...
May 16, 2008 · This lends a key role to the three parameters associated with these forces ... primitive equations, Tellus, Ser. A, 41, 175–178. 10.1111/j.1600 ...
[21]
[PDF] Chapter 3 Basic Dynamics
Feb 3, 2016 · Coriolis force associated with the horizontal component of the earth's ... The resulting approximate set of equations is called the primitive equa ...
[22]
[PDF] Lecture 2: Basic Conservation Laws - UCI ESS
These sets of equations are called the primitive equations, which are very close to the original equations are used for numerical weather prediction. • The ...
[23]
Reviewing and clarifying the derivation of the hydrostatic primitive ...
Jul 28, 2023 · The “deep” terms in Equations 27 and 28 are each responsible for breaking the conservation of energy. To retrieve the situation with this ...INTRODUCTION · THE USUAL DERIVATION OF... · A SIMPLER DERIVATION OF...
[24]
On energy conservation for the hydrostatic Euler equations
Aug 25, 2023 · We show that if the horizontal velocity (u, v) is sufficiently regular then the horizontal kinetic energy is conserved.
[25]
[PDF] The Potential Vorticity Theorem - the NOAA Institutional Repository
to stress that the theorem is still valid for the primitive equation formu. lation of adiabatic, inviscid and hydrostatic motion. Indeed, the potential.
[26]
[PDF] Dry mass versus total mass conservation in the IFS - ECMWF
The primitive equations of the IFS model describe the evolution of multi-phase air parcels constituted of “dry air”, water vapour, water droplets, rain ...
[27]
[PDF] ATMOSPHERIC AND OCEANIC FLUID DYNAMICS
The so-called primitive equations of motion are simplifications of the above equations frequently used in atmospheric and oceanic modelling.3 Three related ...
[28]
[PDF] Lecture 23 Vertical coordinates
10.5 Primitive equations in pressure coordi nates. The other natural choice of vertical coordinate is pressure. Here, however, the non-Boussinesq equations ...
[29]
[PDF] Chapter 3
becomes 0 = Ce +Co +P GFn, where the right hand side is the sum of three forces: centrifugal, Coriolis and pressure gradient normal to the flow. By ...
[30]
A COORDINATE SYSTEM HAVING SOME SPECIAL ADVANTAGES ...
A COORDINATE SYSTEM HAVING SOME SPECIAL ADVANTAGES FOR NUMERICAL FORECASTING. N. A. Phillips. N. A. Phillips Massachusetts Institute of Technology Search for ...
[31]
[PDF] ATMOSPHERIC AND OCEANIC FLUID DYNAMICS
2.2.4 The primitive equations. The so-called primitive equations of motion are simplifications of the above equations fre- quently used in atmospheric and ...
[32]
[PDF] VERTICAL DIFFERENCING OF THE PRIMITIVE EQUATIONS
In this type of coordinate, the pressure gradient force appears as the sum of the two terms that are of comparable magnitude and opposite sign near steep ...
[33]
[PDF] an energy and angular momentum conserving - finite-difference ...
Sigma- and hybrid- coordinate results for steady, forced, planetary- wave structures can be significantly different when the upper-level resolution is ...
[34]
A sigma-coordinate primitive equation model for studying the ...
This paper describes the configuration of a topography-following (sigma) coordinate, numerical ocean model for studying the circulation in the South Atlantic.Missing: derivation | Show results with:derivation
[35]
Vertical Differencing of the Primitive Equations in Sigma ...
A vertical finite-difference scheme for the primitive equations in sigma coordinates is obtained by requiring that the discrete equations retain some important ...Missing: derivation | Show results with:derivation
[36]
[PDF] An Introduction to Dynamic Meteorology
In this fourth edition of An Introduction to Dynamic Meteorology I have retained the basic structure of all chapters of the previous edition. A number of ...
[37]
[PDF] Atmospheric and Oceanic Fluid Dynamics
... equation is. Atmospheric and oceanic fluid dynamics (aofd) is both pure and applied. It is a pure because it involves some of the most fundamental and ...
[38]
[PDF] Atmospheric Dynamics II Lesson 5 – Linear Waves Reference
The governing equations are nonlinear. In order to study the properties of atmospheric waves we “linearize” the governing equations, and then study the linear.
[39]
[PDF] Ocean primitive equations and sea level equations
... (Vallis, 2006). The. 15 hydrostatic ocean primitive equations form the starting point from which OGCM equations are. 16 developed, and we write them in the ...
[40]
Planetary, inertia–gravity and Kelvin waves on the f‐plane and β ...
Wave solutions of the linearized primitive equations form the basis of our understanding of atmospheric and oceanic dynamics, as they provide a simple ...
[41]
[PDF] Waves in the Ocean: Linear Treatment - Falk Feddersen
Mar 4, 2019 · 16.5 Inertia-Gravity Waves: Plane wave solutions and dispersion relationship . ... linear surface gravity wave dispersion relationship ω2 ...
[42]
[PDF] The dynamics of long-waves in a baroclinic westerly current
THE DYNAMICS OF LONG WAVES IN A BAROCLINIC. WESTERLY CURRENT. By J. G. ... OCTOBER 1947. J. G. CHARNEY and is found to be between 350 mb and 400 mb. It ...
[43]
On the Stability of Internal Baroclinic Jets in a Rotating Atmosphere in
The internal jet is stable if the gradient of potential vorticity in isentropic surfaces does not vanish. If it vanishes at a closed isopleth of constant mean ...
[44]
journal of the atmospheric sciences
Pedlosky, J., 1963a: Baroclinic instability in two layer systems. Tellus, 15, 20-25. 1963b: The stability of currents in the atmosphere and the ocean ...