Warm-Hot Intergalactic Medium (WHIM)

• WHIM is a diffuse, shock-heated intergalactic gas (10^5–10^7 K) found in cosmic filaments and sheets, and is a prime candidate for the missing low‐redshift baryons.
• Detection relies on metal-line diagnostics like O VI, O VII, and O VIII, since nearly fully ionized hydrogen makes classical Lyα absorption weak.
• Simulations and observations indicate WHIM forms via gravitational shocks during structure formation, with its properties affected by radiative cooling and feedback from galaxies and AGN.

The warm-hot intergalactic medium (WHIM) is the low-redshift phase of diffuse baryons in the cosmic web, predominantly located in filaments and sheets and characterized by temperatures of roughly $10^{5}-10^{7}\,\mathrm{K}$, low-to-intermediate overdensities, and a high degree of ionization. In the standard $\Lambda$CDM picture, this gas is produced as baryons are shock-heated during structure formation, and it is widely treated as the leading reservoir of the low-redshift “missing baryons,” with multiple studies citing a shortfall of about $30\%$–$50\%$ in the directly observed baryon census and simulations placing a comparable fraction in the WHIM phase (Klar et al., 2010). Because hydrogen is mostly ionized, the WHIM is difficult to detect directly: classical Ly$\alpha$ absorption is weak, while the most informative diagnostics are highly ionized metal lines, especially O VI, O VII, O VIII, and Ne IX, together with increasingly sensitive soft X-ray emission measurements (Nicastro et al., 2022).

1. Definition, baryon census, and phase-space location

WHIM denotes intergalactic gas in the temperature interval $10^{5}-10^{7}\,\mathrm{K}$, generally at baryon number densities or hydrogen densities of order $10^{-6}-10^{-4}\,\mathrm{cm^{-3}}$, and at overdensities ranging from roughly $\delta \sim 5$–$500$ in absorption-oriented summaries to $\delta<1000$ in several simulation-based emission studies (Nicastro et al., 2022). It is distinguished from the high-redshift photoionized Ly$\Lambda$0 forest by being primarily shock-heated and collisionally ionized, and from dense virialized intracluster gas by its lower density and filamentary or sheet-like geometry (Roncarelli et al., 2012).

Its cosmological significance follows from the low-redshift baryon census. Several studies summarize that only about $\Lambda$1–$\Lambda$2 of baryons are directly accounted for in galaxies, stars, cold gas, the intracluster medium, and the photoionized intergalactic medium, leaving a deficit of order $\Lambda$3–$\Lambda$4 or, in some formulations, $\Lambda$5–$\Lambda$6 of baryons still undetected (Nicastro et al., 2018). Simulation-driven reviews and observational chapters consistently identify the WHIM as the dominant candidate reservoir for that deficit, especially in filaments and halo outskirts at $\Lambda$7 (Nicastro et al., 2022).

The phase is not uniform. Simulation-based analyses distinguish cooler and warmer UV-traced components, hotter O VII/O VIII-bearing components, and denser X-ray-emitting subsets. In absorption-oriented treatments, O VII is emphasized as the dominant ion for $\Lambda$8, with O VIII dominating at still higher temperatures (Yao et al., 2011). In emission-oriented analyses, the detectable subset is biased toward the hotter and denser tail of the WHIM distribution rather than toward the full mass-weighted reservoir; one eROSITA stacking study explicitly concludes that broadband X-ray emission traces the high end of the temperature and density values that characterize the entire WHIM population (Zhang et al., 2024).

2. Formation in large-scale structure and thermodynamic regulation

In the $\Lambda$9CDM framework, the WHIM forms as gas flows into the large-scale potential wells of the cosmic web and is shock-heated during collapse into sheets, filaments, halos, and nodes. The simplest controlled treatment in the provided literature is the one-dimensional “pancake” collapse model, where a sinusoidal perturbation of comoving scale $30\%$0 is evolved with hydrodynamics, gravity, chemistry, radiative cooling, a UV background, and thermal conduction (Klar et al., 2010). In that formulation, the initial density perturbation is written as

$30\%$1

with $30\%$2.

The non-radiative collapse is self-similar: gas contracts adiabatically until the infall speed reaches the sound speed, after which a pair of shocks forms and confines a hot central region. The post-shock density profile is described as roughly $30\%$3, and a scaling relation is obtained in which the shocked temperature grows approximately as $30\%$4, connecting larger perturbation scales directly to hotter WHIM gas (Klar et al., 2010). This provides a physically transparent route from gravitational collapse to the characteristic $30\%$5 WHIM range.

Radiative cooling and UV background heating substantially modify this basic picture. In the one-dimensional simulations, photoheating during reionization raises the gas to about $30\%$6 and prevents catastrophic central cooling, producing a three-part structure consisting of a cold dense core, a shock-heated WHIM envelope, and a low-density outer region (Klar et al., 2010). A scale threshold appears: for $30\%$7 Mpc, the sound speed in the UV-heated gas stays above the infall speed and no true confining shock forms, while for $30\%$8 Mpc shocks arise at the edges of the pre-shock core (Klar et al., 2010).

The core properties scale strongly with perturbation length. The central hydrogen density, core temperature, and core radius are reported to follow

$30\%$9

derived from an approximate hydrostatic relation and an equilibrium cooling/heating relation over the relevant density range (Klar et al., 2010). Thermal conduction adds another scale dependence because $50\%$0. It is negligible for smaller pancakes but becomes important for $50\%$1 Mpc and can trigger partial evaporation of the cold core, with an order-of-magnitude conduction criterion

$50\%$2

The resulting inference is that, in one dimension, cold shock-confined WHIM cores exist in a perturbation-scale window of roughly $50\%$3 to $50\%$4 Mpc (Klar et al., 2010).

A broader implication drawn explicitly in later environmental studies is that WHIM properties depend on both gravitational shocks and stellar or black-hole feedback. Deep galaxy-redshift work around 1ES1553+113 states that gravitational collapse heats portions of the intergalactic medium into the WHIM, while feedback enriches and further heats gas, but not necessarily to radii far beyond halos and groups (Johnson et al., 2019). This suggests that the existence of WHIM is not itself controversial, but the metallicity and observability of its most diffuse component are tightly coupled to enrichment physics.

3. Ionization state, metallicity, and spectral diagnostics

The observational difficulty of the WHIM is a direct consequence of its ionization state. Neutral hydrogen fractions are strongly suppressed; one absorption-oriented chapter summarizes the neutral fraction as roughly $50\%$5 in the WHIM regime, making the gas nearly transparent in the far-UV compared with the Ly$50\%$6 forest (Nicastro et al., 2022). As a result, the practical tracers are metal ions, above all oxygen. O VI traces cooler WHIM phases and can arise from either photoionization or collisional ionization, whereas O VII and O VIII are favored for the hotter soft-X-ray phase and are harder to create by photoionization alone in the low-density intergalactic environment (Yao et al., 2011).

A standard absorption relation used in the literature is

$50\%$7

where $50\%$8 is the baryon-equivalent column density, $50\%$9 the ionic fraction, and $\alpha$0 the abundance relative to hydrogen (Nicastro et al., 2022). Equivalent width, line profile, Doppler parameter, and line centroid are the primary observables. In the linear curve-of-growth regime, one chapter gives

$\alpha$1

while emphasizing that saturation and line-width degeneracies quickly become important for stronger systems (Nicastro et al., 2022).

The ionization balance can also diagnose temperature. In emission, the O VIII/O VII line ratio is repeatedly treated as the main WHIM thermometer. A microcalorimeter-forecast study found that, when both lines are detected, temperature can be estimated from the line ratio with a precision of about $\alpha$2, degrading when contamination, continuum variation, and reduced energy resolution are included (Takei et al., 2010). In absorption, the ratio of O VII to O VI is used to distinguish hot collisionally ionized gas from lower-temperature photoionized systems; one reanalysis notes that in shock-heated gas $\alpha$3 is expected to be about ten times larger than $\alpha$4, whereas in photoionized gas the ratio is smaller, $\alpha$5 (Yao et al., 2011).

Metallicity is a major systematic. Soft X-ray emission predictions depend strongly on how metals are mixed and distributed. A large SPH light-cone study tested four metallicity prescriptions—Borgani, Croft, Scatter, and Cen—and found that the predicted WHIM brightness can vary by almost an order of magnitude between them (Ursino et al., 2010). In the $\alpha$6–$\alpha$7 keV band, the mean WHIM surface brightnesses differ from $\alpha$8 to $\alpha$9 in the table units quoted there, with Scatter and Cen about $10^{5}-10^{7}\,\mathrm{K}$0–$10^{5}-10^{7}\,\mathrm{K}$1 times brighter than Borgani (Ursino et al., 2010). This dependence is physically important because line emission scales strongly with metal abundance, while the underlying continuum is much less sensitive. A plausible implication is that metallicity uncertainty does not merely perturb detectability; it materially changes which fraction of the WHIM is visible in emission at all.

4. Observational methods: absorption, emission, and signposts

The two principal observational strategies are absorption against bright background sources and soft X-ray emission from diffuse structures. Absorption is column-density weighted and thus sensitive to low-density material along individual sightlines; emission scales approximately as density squared and is therefore biased toward denser, hotter regions (Yao et al., 2011). This division underlies much of the methodological landscape of WHIM work.

In X-ray absorption, the most common background targets are blazars and quasars with bright, relatively featureless continua. The line of interest is often O VII K$10^{5}-10^{7}\,\mathrm{K}$2 at rest wavelength $10^{5}-10^{7}\,\mathrm{K}$3, or O VII He$10^{5}-10^{7}\,\mathrm{K}$4 near $10^{5}-10^{7}\,\mathrm{K}$5 keV in energy units (Yao et al., 2011). A deep XMM-Newton RGS observation toward Mkn 501, located in the Hercules Supercluster, detected an O VII K$10^{5}-10^{7}\,\mathrm{K}$6 absorption line at $10^{5}-10^{7}\,\mathrm{K}$7 with $10^{5}-10^{7}\,\mathrm{K}$8, $10^{5}-10^{7}\,\mathrm{K}$9, and a Monte Carlo significance of $10^{-6}-10^{-4}\,\mathrm{cm^{-3}}$0, consistent with WHIM conditions of $10^{-6}-10^{-4}\,\mathrm{cm^{-3}}$1, $10^{-6}-10^{-4}\,\mathrm{cm^{-3}}$2, and $10^{-6}-10^{-4}\,\mathrm{cm^{-3}}$3 under the paper’s stated assumptions (Ren et al., 2014). The same study explicitly argues that superclusters are effective signposts for WHIM searches.

A much deeper XMM-Newton RGS spectrum of 1ES 1553+113, with cleaned exposure $10^{-6}-10^{-4}\,\mathrm{cm^{-3}}$4 Ms, reported two intervening O VII absorbers at $10^{-6}-10^{-4}\,\mathrm{cm^{-3}}$5 and $10^{-6}-10^{-4}\,\mathrm{cm^{-3}}$6, with observed equivalent widths $10^{-6}-10^{-4}\,\mathrm{cm^{-3}}$7 mÅ and $10^{-6}-10^{-4}\,\mathrm{cm^{-3}}$8 mÅ, respectively (Nicastro et al., 2018). That study interpreted the absorbers as metal-rich WHIM in galaxy-overdense environments and estimated

$10^{-6}-10^{-4}\,\mathrm{cm^{-3}}$9

corresponding to roughly $\delta \sim 5$0–$\delta \sim 5$1 of the total baryon density measured by Planck (Nicastro et al., 2018).

In emission, the strategy has shifted from idealized forecasts to both stacking analyses and individual filament detections. Forecast studies for microcalorimeter missions showed that paired O VII and O VIII line searches could detect hundreds of WHIM systems per square degree in $\delta \sim 5$2 Ms observations with $\delta \sim 5$3 eV, and still about $\delta \sim 5$4 systems per square degree at $\delta \sim 5$5 eV resolution (Takei et al., 2010). More recent simulation work with CAMELS emphasized that detectable WHIM emission is largely associated with galaxy-size halos, halo outskirts, group environments, and overlapping structures rather than with the full diffuse filament population (Parimbelli et al., 2022).

Observationally, stacking has become a major tool. Using the first four eROSITA all-sky scans, one study stacked 7817 SDSS filaments and obtained a $\delta \sim 5$6 detection of total $\delta \sim 5$7–$\delta \sim 5$8 keV filament-associated emission (Zhang et al., 2024). After estimating a contamination fraction of approximately $\delta \sim 5$9 from unmasked halos, AGN, and X-ray binaries, it attributed the remaining $500$0 to WHIM, corresponding to a $500$1 WHIM detection, a best-fit single-temperature value $500$2, and $500$3 for the X-ray-bright phase (Zhang et al., 2024).

Targeted observations of individual inter-cluster structures remain especially valuable because they reduce line-of-sight confusion. A pilot HST/COS experiment deliberately selected a QSO sightline passing within $500$4 Mpc of seven inter-cluster axes and found excesses of total H I, narrow H I, broad Ly$500$5, and O VI absorbers near the cluster-pair redshifts, with the largest relative excess for broad Ly$500$6 absorbers: approximately $500$7 the field expectation (Tejos et al., 2015). The authors argued that the stronger excess of BLAs compared with NLAs may be a signature of the WHIM in inter-cluster filaments.

5. Observational ambiguities, contamination, and re-interpretation

A persistent feature of WHIM research is that many detections are marginal, model-dependent, or vulnerable to environmental re-interpretation. A systematic reanalysis of all available Chandra and XMM grating observations of Mrk 421 concluded that even spectra with signal-to-noise ratios of about $500$8 and $500$9 per $\delta<1000$0 mÅ spectral bin could not confirm the previously reported systems at $\delta<1000$1 and $\delta<1000$2, and that spectra of such quality cannot constrain WHIM absorbers with $\delta<1000$3 at $\delta<1000$4 significance (Yao et al., 2011). The same work estimated that, for a $\delta<1000$5 mCrab background QSO, detecting representative $\delta<1000$6 systems at $\delta<1000$7 would require about $\delta<1000$8 Ms with Chandra or $\delta<1000$9 Ms with XMM (Yao et al., 2011).

Local multiphase absorption adds another severe complication. A 2023 RGS analysis of six extragalactic sources modeled the Galactic ISM with a physically motivated three-phase IONeq treatment and found no statistical improvement when an additional WHIM component was included for any source (Gatuzz et al., 2023). That study argues that multiphase ISM absorption can mimic or hide low-redshift WHIM features, especially in the $\Lambda$00 Å region, and that previously claimed WHIM lines can often be explained by local O II, O VI, O VII, N II, or N VII absorption (Gatuzz et al., 2023).

Environmental ambiguity is equally important. The 2018 1ES 1553+113 result was explicitly revisited by a deep galaxy-redshift survey that measured redshifts for 921 galaxies within $\Lambda$01 of the blazar and concluded that the $\Lambda$02 O VII absorber is most likely associated with the blazar’s own group or intragroup medium, while the $\Lambda$03 system lies in an unusually isolated environment where such strong O VII absorbers are statistically rare (Johnson et al., 2019). That work therefore argued that the two systems do not provide secure evidence that the WHIM closes the missing baryon budget and placed a $\Lambda$04 upper limit

$\Lambda$05

for O VII absorbers with $\Lambda$06 (Johnson et al., 2019).

This distinction between diffuse WHIM and denser CGM or intragroup gas is emphasized in the review literature. One chapter notes a break in the column-density distribution around $\Lambda$07, above which absorbers increasingly correspond to high-density halo or circumgalactic environments rather than truly diffuse intergalactic WHIM (Nicastro et al., 2022). This suggests that the strongest observed metal-line systems are not automatically the most representative baryon-census tracers. It also explains why broad H I Ly$\Lambda$08 absorbers remain important: they provide a metallicity-independent probe of warm gas and, in the 1ES1553+113 field, were found on average to lie about $\Lambda$09 pkpc from the nearest luminous galaxy, intermediate between O VI absorbers at $\Lambda$10 pkpc and narrow Ly$\Lambda$11 absorbers at $\Lambda$12 pkpc (Johnson et al., 2019).

6. Current detections and next-generation prospects

Despite the ambiguities, the recent literature contains both stronger stacked detections and more carefully isolated individual-emission detections. A 2025 study of a $\Lambda$13 Mpc filament in the Shapley supercluster combined four Suzaku pointings with XMM-Newton source characterization and reported direct imaging and spectroscopic detection of extended thermal WHIM emission from a single filament (Migkas et al., 17 Jun 2025). The imaging analysis found $\Lambda$14 excess X-ray emission relative to the sky background at $\Lambda$15, while spectroscopy of regions outside $\Lambda$16 of all nearby clusters yielded a thermal component with

$\Lambda$17

at $\Lambda$18 significance (Migkas et al., 17 Jun 2025). The paper presents this as the first X-ray spectroscopic detection of pure WHIM emission from an individual pristine filament without significant contamination from unresolved point sources and gas clumps.

Future capabilities are framed around spectral resolution, collecting area, field of view, and target selection. For absorption, a Chandra/XMM reanalysis argued that future X-ray spectrographs need approximately $\Lambda$19 and $\Lambda$20 to survey WHIM absorbers efficiently, with $\Lambda$21 singled out as the threshold needed to resolve expected O VII widths for $\Lambda$22 (Yao et al., 2011). A broader review formalized a detectability figure of merit, $\Lambda$23, and concluded that secure detections of the most diffuse gas in the low-redshift large-scale structure may have to await next-generation X-ray telescopes (Nicastro et al., 2022).

Athena/X-IFU forecasts extend this logic to transient and persistent backlights. Simulations using GRB afterglows found that, for 50 ks observations, Athena will be able to detect O VII–O VIII absorption pairs with $\Lambda$24 eV and $\Lambda$25 eV for afterglows with $\Lambda$26, implying roughly $\Lambda$27–$\Lambda$28 detectable absorbers over a four-year mission lifetime, provided the Galactic column satisfies $\Lambda$29 (Walsh et al., 2020). A 2026 NewAthena study of bright power-law sources similarly found that, in a favorable redshift interval $\Lambda$30–0.125 for a $\Lambda$31 eV O VII line, about $\Lambda$32 counts are needed for $\Lambda$33 and $\Lambda$34 counts for $\Lambda$35, underscoring that counts rather than exposure time alone are the controlling variable (Fisher et al., 26 May 2026).

On the emission side, mission concepts emphasize wide-field microcalorimetry. HUBS forecasts based on 1577 inter-cluster filaments selected from the eRASS1 supercluster catalog found that four candidate filaments are especially promising and that, with 200 ks per target, individual filament temperatures could be constrained to $\Lambda$36 keV, metallicities to $\Lambda$37 solar, and densities to $\Lambda$38, while elemental abundances of O, Ne, Mg, and Fe could be measured separately (Zhao et al., 2024). The same study shows that narrowband imaging around the O VIII line can directly map filament morphology (Zhao et al., 2024). In parallel, simulation-based emission work with CAMELS argues that the most robust observational strategy will combine O VII and O VIII line statistics, angular and spatial clustering, and cross-correlation with galaxy or halo catalogs, because detectable emission is overwhelmingly halo-associated and line counts and clustering respond differently to feedback and cosmology (Parimbelli et al., 2022).

A complementary non-X-ray direction has also emerged. A giant radio galaxy associated with NGC 6185 has been used as an “intergalactic barometer,” yielding an inferred ambient temperature

$\Lambda$39

at approximately the group virial radius and demonstrating a method that could extend from group outskirts to the WHIM (Oei et al., 2022). This suggests that future WHIM thermodynamics may be constrained not only by direct line spectroscopy but also by dynamical probes tied to cosmic-web environments.

Taken together, these results define the present state of the subject. The WHIM is a theoretically robust phase of low-redshift baryons and is increasingly supported by both stacked and targeted observations. Yet the decisive issues are not merely detectability, but whether a given signal traces diffuse intergalactic gas rather than halo, group, or local Galactic structures; whether metallicity and ionization modeling are adequate; and whether the observed subset is representative of the full missing-baryon reservoir. The current literature suggests that the answer is progressively becoming empirical, but only with instrumentation capable of high-throughput, high-resolution spectroscopy and with analysis strategies that explicitly control for environmental and foreground contamination.