Read the chemistry, clouds and maps behind a phase curve

Begin with hydrogen changing its form

In Bell and Cowan, read sections 2.1 to 2.3 before interpreting the phase curves. Figure 2 is especially useful: at the stated pressure, the hydrogen dissociation fraction changes rapidly over an ultra-hot temperature interval. Molecular hydrogen separates into atoms on the hot side; recombination can return that stored chemical energy on the cooler side.

The relevant energy scale is about $2.14\times10^8\,\mathrm{J\,kg^{-1}}$ of hydrogen. A layer's chemical energy depends on its mass column and its dissociated fraction. In the paper's notation, that contribution has the structure

When the fraction changes steeply with temperature, a temperature change can absorb or release substantial chemical energy. Follow equations 1 to 8 together, including the temperature-dependent sensible-heat terms. The extra energy is stored and transported in matter; it is not an unreported new heating source.

The paper's comparisons in figures 1 to 4 are theoretical model predictions under specified circulation and equilibrium-chemistry assumptions. The constant-capacity layer developed in the preceding chapters omits this mechanism. The selected WASP-43b observation is not used here as a detection of ultra-hot hydrogen dissociation.

Reading: Taylor J. Bell and Nicolas B. Cowan, Increased Heat Transport in Ultra-Hot Jupiter Atmospheres Through H2 Dissociation/Recombination · Nicolas B. Cowan and Eric Agol, A Model for Thermal Phase Variations of Circular and Eccentric Exoplanets

Read the WASP-43b light before its atmospheric explanation

In Bell and colleagues, begin with the observing description and main figure 1. Locate the phase convention, the eclipses and the measured light curve. The accompanying Zenodo products retain the time series and multiple reductions. Our tutorial demonstrates why their time arrays, trimming and shared-photon origin need to be read before fitting a physical interpretation.

Next compare figure 2's observations with the atmospheric circulation predictions. Those colored model curves represent calculated atmospheres, not additional observed phase curves. Their treatment of clouds and circulation changes the brightness as a function of phase. Disagreement between a model and a nightside measurement identifies something that the atmospheric calculation must explain.

Then inspect the phase-resolved spectra in figure 3. A cloud can prevent radiation from escaping from a hotter layer below it. A weak vertical temperature gradient can also make spectral features less pronounced. The study compares several atmospheric models with its light curves and spectra to examine these possible structures.

For the spectrum, note wavelength limits and error conventions. The shadowed region beyond ten and a half micrometres is not part of the primary broadband inference used here. The product spectra arise from the same time-series measurements, so adding them as independent evidence to the same likelihood would count photons twice.

Reading: Taylor J. Bell et al., Nightside clouds and disequilibrium chemistry on the hot Jupiter WASP-43b · Taylor James Bell and the 17 other named Zenodo creators, Products and Models for Nightside clouds and disequilibrium chemistry on the hot Jupiter WASP-43b

Keep a brightness map attached to its observing wavelength

A map inferred from a particular infrared band describes where that band appears bright. Turning it into a gas temperature needs a source function, opacity, vertical structure and an instrumental response. Different wavelengths may probe different depths, so a collection of brightness maps can reveal atmospheric structure that one broad curve cannot separate.

This explains the distinct roles of the source's phase curve and its spectrum. The curve measures a change in total band light as viewing longitude changes. The spectrum measures how that light is distributed in wavelength. The quoted brightness temperatures describe the measured spectral emission with a blackbody comparison; they do not assign the same temperature to every pressure layer on each hemisphere.

Albedo introduces another wavelength distinction. Bond albedo sums reflected stellar power over wavelength and direction. MIRI primarily samples a part of the planet's thermal emission. Connecting these quantities requires an atmospheric or bolometric model that carries uncertainty through the missing spectral information.

Reading: Nicolas B. Cowan and Eric Agol, The Statistics of Albedo and Heat Recirculation on Hot Exoplanets · Taylor J. Bell et al., Nightside clouds and disequilibrium chemistry on the hot Jupiter WASP-43b

See what an occulting limb adds

In Luger and colleagues, follow the rotation in section 2.4, then the projection and occultation operators in sections 3.1 to 3.3. Equation 16 rotates the brightness representation. Equations 17 to 20 connect the representation to a projected disk and integrals over the part that remains visible. Their analytic machinery is more general than the small numerical model in our course.

Figure 5 is the useful visual comparison. Different occultor trajectories suppress or reveal different spatial patterns. A full-disk phase measurement averages many regions together; a moving stellar limb changes which subset contributes. Extra spatial weighting can expose a mode that a phase curve erases.

The independent finite-basis example alongside this guide uses constant-latitude intensity modes through longitude harmonic order six. There are thirteen coefficients: a constant and six cosine/sine pairs. In the actual retained observation window, the phase-only matrix still has four zero directions at inclination near eighty-two degrees. Tilt alone has not removed them in this basis.

Adding secondary-eclipse samples makes all thirteen singular directions nonzero. At the finer tested surface grid, the smallest normalized singular value is about $0.00647$, compared with values near floating-point zero for the missing phase-only modes. The two resolutions show how stable that numerical response is. These are geometric response strengths, not posterior detection probabilities.

A general two-dimensional temperature map has latitude structure and a brightness-to-temperature relation beyond these thirteen columns. Full rank within the selected basis therefore answers a finite linear question. It does not make a general map unique or guarantee that a weak mode survives measurement noise.

Reading: Rodrigo Luger et al., STARRY: Analytic Occultation Light Curves · Nicolas B. Cowan and Eric Agol, Inverting Phase Functions to Map Exoplanets

Compare a physical prior with an invisible mode

A transport model restricts the set of possible maps by imposing a differential equation, capacity, flow and heating. A smoothness prior restricts it by preferring neighboring brightness values to change gradually. These choices provide different information. Either can make a numerical inverse problem easier, while their conclusions remain conditional on what was supplied.

The positive third-harmonic map family in the tutorial supplies a concrete example. Its ideal edge-on phase curve is unchanged even though its surface pattern changes. The thermal model supplies another: an exact steady capacity–flow ridge. The first ambiguity comes from the observing operator; the second comes from the governing equation. Adding a more accurate optimizer does not create a new observation of either missing quantity.

When reading another paper, write down the emitting wavelength or band, the intensity law, the observed phases and eclipse track, and the map or atmosphere assumed by the inference. Then locate where chemical energy or cloud opacity enters the forward prediction. Those choices make the path from an unseen atmosphere to a measured curve explicit enough to test.

Reading: Nicolas B. Cowan and Eric Agol, Inverting Phase Functions to Map Exoplanets · Nicolas B. Cowan and Eric Agol, A Model for Thermal Phase Variations of Circular and Eccentric Exoplanets · Rodrigo Luger et al., STARRY: Analytic Occultation Light Curves