Starting from zero, the Landweber iteration is
For , take the step size . Iterating the linear update gives
and its Landweber spectral filter in a singular system of a compact operator is
The strict upper step size bound makes for every positive singular value. A common more restrictive choice is , giving nonnegative damping factors. Each finite iterate is bounded and linear; early stopping of Landweber iteration controls the amplification of noise as smaller singular values are progressively inverted. If , every iterate is zero for any positive step size.
Suppose each is a bounded linear operator and for every as . A standard sufficient a priori regularization parameter choice satisfies
For every datum with , the noise-bias decomposition for linear regularization gives
Thus the parameter rule gives a convergent regularization of an inverse problem, uniformly over data in the prescribed noise ball for each fixed admissible exact datum. For Tikhonov regularization, , so and suffice. The same sufficient noise scaling holds for spectral cutoff regularization.
Fix a singular system of a compact operator, with and , . A spectral regularization method has the form
It annihilates . Standard sufficient spectral filter conditions are boundedness of for each , pointwise as for , and a uniform bound on . These give bounded linear operators and convergence on the domain of the Moore–Penrose inverse of an operator by the Picard criterion and dominated convergence theorem.
For Tikhonov regularization,
This can be computed by solving the Tikhonov normal equation, without knowing the singular value decomposition.
For spectral cutoff regularization, take
with value zero at . Only singular values are inverted. Here multiplies ; conventions calling the full coefficient the spectral filter are equivalent.
Write . Every least-squares solution is for some , while . Decompose using orthogonal projections. Then the Pythagorean identity gives
The first term is independent of , and the second vanishes at exactly one point. Hence the unique nearest least-squares solution is
Geometrically this is the orthogonal projection of onto the closed affine subspace of least-squares solutions.
The Moore–Penrose inverse of an operator is the linear map
that assigns each admissible datum its minimum-norm least-squares solution. If with and , then is the unique satisfying . In particular it vanishes on , and
The continuity criterion, with the inherited data norm on its domain, is
For a closed operator range, the restriction of from to is a bounded bijection between Banach spaces, so its inverse is bounded by the bounded inverse theorem. Conversely, if is bounded and , replace each by . This is a Cauchy sequence, hence converges to , and . Thus every limit of range elements remains in the operator range.
The map is a bounded linear operator, so the inverse image of the closed singleton is closed. That inverse image is also a convex set: if , every convex combination has the same image.
The identity gives
Consequently such a exists exactly when is the sum of an element of and one of its orthogonal complement. Thus
These are precisely the least-squares solutions. Indeed, if and , then for every the Pythagorean identity gives
Conversely, differentiating the squared residual along any real direction forces , hence the normal equation for a linear inverse problem. For any , the difference of two solutions of the normal equation for a linear inverse problem lies in , since . Therefore
a closed affine subspace; the empty case is closed and convex as well.
Use the Hilbert space interpretation of a linear inverse problem, with a bounded linear operator . A least-squares solution minimizes over . A minimum-norm least-squares solution additionally has the smallest -norm among all least-squares solutions. When the equation is consistent this is the smallest-norm exact solution.
The least-squares solutions form a nonempty closed affine subspace precisely when
The closest point theorem in a Hilbert space then gives a unique nearest point to zero, hence a unique minimum-norm least-squares solution. Equivalently, it is the unique least-squares solution in . If is closed, this exists for every datum; a nonclosed operator range can leave some data without any least-squares solution.
For example, on the l2 sequence space, take and . The equation would require for every , which is not square summable. However, setting the first coordinates of to one and the rest to zero gives
Thus the least-squares infimum is zero but is not attained, and no minimum-norm least-squares solution exists for these data.
At fixed temperature and pressure, the atmospheric carbon-to-oxygen ratio controls the division of the elemental inventory among carbon monoxide, methane, water and other molecules. For an oxygen-rich mixture, water can remain after carbon monoxide consumes its share of oxygen. Increasing the atmospheric carbon-to-oxygen ratio reduces the oxygen available for water, particularly in the hotter layers where carbon monoxide is stable. Near or above unity there, water can become strongly depleted while excess carbon feeds methane, hydrogen cyanide and acetylene. This familiar hot carbon-rich behavior should not be applied unchanged to every cool layer: at , methane formation can leave substantial water even at high atmospheric carbon-to-oxygen ratio.
Increasing the atmospheric metallicity of a giant planet at fixed elemental ratios increases the heavy-element inventory relative to molecular hydrogen and helium. Abundances of water and the major carbon-bearing molecules generally rise while molecular hydrogen remains dominant. In that regime, the law of mass action for gives
Where carbon monoxide and water each scale roughly linearly with enrichment , carbon dioxide consequently scales approximately as . This scaling changes when molecular hydrogen ceases to dominate or chemical partitioning changes. Higher enrichment can also favor carbon monoxide over methane at fixed conditions, increase mean molecular weight, and reduce the atmospheric scale height and transmission-feature amplitudes. Cloud formation and atmospheric condensate rainout alter the observable elemental ratios relative to the bulk ones.
The first two signatures can be consequences of the same process and are not independent evidence for two mechanisms. Changes in atmospheric carbon-to-oxygen ratio, atmospheric metallicity of a giant planet, exoplanet cloud coverage, and the atmospheric pressure-temperature profile can mimic abundance changes. A convincing inference compares several molecular bands with a chemically consistent equilibrium model rather than identifying one unusual band alone.
For solar elemental abundances, take a molecular hydrogen- and helium-dominated atmosphere with atmospheric carbon-to-oxygen ratio near . The principal oxygen-bearing molecule is generally water; the carbon and nitrogen carriers depend on both temperature and pressure.
At the cool observable upper layers, methane is the expected main carbon reservoir under chemical equilibrium, with abundant water and much less carbon monoxide. At the hotter layers approaching , carbon monoxide becomes the main carbon carrier and water contains much of the oxygen not bound in it. The relevant law of mass action follows from
Lower temperature favors the exothermic methane-forming direction, while increasing pressure favors the side with fewer molecules. This explains why the change of dominant carrier cannot be specified by temperature alone.
Nitrogen is distributed between molecular nitrogen and Ammonia. Cooling favors Ammonia, but low pressure favors molecular nitrogen; it is therefore unsafe to call Ammonia dominant throughout the low-pressure region. The exact partition requires the equilibrium constant for .
Carbon dioxide is usually a minor constituent at solar composition; hydrogen cyanide and acetylene are much less abundant than the principal carbon carriers in this oxygen-rich equilibrium case. Condensation can remove refractory species where a condensation curve is crossed. The robust cool-atmosphere expectation is a molecular hydrogen–helium background with methane and water, changing toward carbon monoxide in hotter layers. Precise mixing ratios require thermodynamic data and an explicitly specified elemental inventory after rainout.
Take the logarithm to base ten and write pressure in units of , so its argument is dimensionless. Continuity of the atmospheric pressure-temperature profile gives
Thus
If the logarithm means , the equivalent constant is . A plot of temperature against logarithmic pressure is vertical in each isothermal region and straight between the two endpoints.
The processes can be organized by the supplied pressure ranges, although the exact boundaries require reaction rates, irradiation and mixing information:
The profile identifies plausible chemical regimes, but does not fix their transition pressures by itself. In particular, cloud formation depends on the species-specific condensation curve, and ultraviolet processing depends on shielding.
For uniform composition, the Schwarzschild criterion gives stability against convection when
A parcel displaced upward expands approximately adiabatically. If the ambient logarithmic temperature gradient is shallower than the adiabatic temperature gradient, the parcel cools more than its surroundings, becomes denser, and sinks back. Equality is neutral stability; a steeper gradient drives buoyant growth.
For a diatomic ideal gas with active translational and rotational degrees of freedom, and . If the infrared opacity is constant, , so the profile's radiative temperature gradient is
An atmospheric thermal inversion has negative and is stable by this criterion. Composition gradients require the Ledoux criterion; molecular dissociation and changing heat capacities can also change the numerical adiabatic temperature gradient.
The thermal optical depth increases inward, so an atmospheric thermal inversion requires . Differentiating the semi-grey irradiated atmosphere profile gives
Therefore the inversion criterion with intrinsic planetary flux is
An inverted layer exists at the top precisely when
For , it extends over
provided the logarithm is positive. Otherwise the profile has no inverted interval. For a strongly irradiated hot Jupiter, and the threshold is approximately ; the intrinsic flux eventually restores an inward-increasing temperature at greater depth.
Large visible absorption opacity deposits stellar heat high in the exoplanet atmosphere. Gas-phase titanium monoxide and vanadium monoxide are candidate absorbers in sufficiently hot layers. An atmospheric cold trap or other condensate loss can remove them, while sufficiently high temperature and vigorous mixing can help keep them in the gas. The condition concerns absorbing opacity: highly reflective scattering alone does not deposit the required heat.
The model assumes a static plane-parallel atmosphere, separate grey stellar and thermal bands with constant opacities, local thermodynamic equilibrium for thermal emission, negligible scattering, and transport by radiative transfer rather than convection. The Eddington closure approximation sets the thermal angular-moment ratio to , and the upper boundary supplies the usual term. Treating the incoming radiation with representative direction cosine gives its attenuation . Intrinsic flux enters from below, while the imposed external stellar flux is absorbed from above.
With constant gravity and hydrostatic equilibrium, , allowing conversion to an atmospheric pressure-temperature profile. This last relation additionally assumes constant thermal opacity. The grey treatment describes the energy balance approximately; it does not resolve individual molecular absorption lines.
The original PDF supplies the intended profiles , and ; the TeX has corrupted the first two. At fixed , , and at fixed , . The spherical continuity equation is therefore
Since is nonzero, the continuity equation for a superbubble similarity solution is
The ambient gas is at rest. In the locally stationary shock frame, its velocity is ; the downstream velocity is . Using the shock compression ratio from part (a) converts the downstream speed back to the laboratory frame and gives the inner shock boundary conditions
Immediately outside the shock, the corresponding unperturbed profiles are , and , with understood in the negligible-ambient-pressure limit.
The inner boundary must supply the injected power. If the source is idealized as a point and the profiles extend all the way inward, central energy input in a similarity solution requires
or equivalently . A finite source region or an inner wind/contact region supplies a different matching description. One cannot impose a regular zero-flux centre on a source-free interior and still inject nonzero power.

Pinned article: Introduction to the OurBigBook Project

Welcome to the OurBigBook Project! Our goal is to create the perfect publishing platform for STEM subjects, and get university-level students to write the best free STEM tutorials ever.
Everyone is welcome to create an account and play with the site: ourbigbook.com/go/register. We belive that students themselves can write amazing tutorials, but teachers are welcome too. You can write about anything you want, it doesn't have to be STEM or even educational. Silly test content is very welcome and you won't be penalized in any way. Just keep it legal!
We have two killer features:
  1. topics: topics group articles by different users with the same title, e.g. here is the topic for the "Fundamental Theorem of Calculus" ourbigbook.com/go/topic/fundamental-theorem-of-calculus
    Articles of different users are sorted by upvote within each article page. This feature is a bit like:
    • a Wikipedia where each user can have their own version of each article
    • a Q&A website like Stack Overflow, where multiple people can give their views on a given topic, and the best ones are sorted by upvote. Except you don't need to wait for someone to ask first, and any topic goes, no matter how narrow or broad
    This feature makes it possible for readers to find better explanations of any topic created by other writers. And it allows writers to create an explanation in a place that readers might actually find it.
    Figure 1.
    Screenshot of the "Derivative" topic page
    . View it live at: ourbigbook.com/go/topic/derivative
  2. local editing: you can store all your personal knowledge base content locally in a plaintext markup format that can be edited locally and published either:
    This way you can be sure that even if OurBigBook.com were to go down one day (which we have no plans to do as it is quite cheap to host!), your content will still be perfectly readable as a static site.
    Figure 5. . You can also edit articles on the Web editor without installing anything locally.
    Video 3.
    Edit locally and publish demo
    . Source. This shows editing OurBigBook Markup and publishing it using the Visual Studio Code extension.
  3. https://raw.githubusercontent.com/ourbigbook/ourbigbook-media/master/feature/x/hilbert-space-arrow.png
  4. Infinitely deep tables of contents:
    Figure 6.
    Dynamic article tree with infinitely deep table of contents
    .
    Descendant pages can also show up as toplevel e.g.: ourbigbook.com/cirosantilli/chordate-subclade
All our software is open source and hosted at: github.com/ourbigbook/ourbigbook
Further documentation can be found at: docs.ourbigbook.com
Feel free to reach our to us for any help or suggestions: docs.ourbigbook.com/#contact