EHVO: Extreme Quasar Outflows

• EHVOs are extremely high-velocity broad absorption outflows in quasars, typically spanning 0.1c–0.2c and identified via highly blueshifted C IV absorption features.
• They are associated with extreme C IV blueshifts and weak He II emission, indicating wind conditions favorable for efficient radiative line driving.
• Observational studies across a wide redshift range show that EHVOs offer critical insights into AGN feedback, with kinetic power scaling as the cube of velocity.

Extremely High Velocity Outflows (EHVOs) are most commonly defined in quasar studies as broad absorption outflows at speeds between $0.1c$ and $0.2c$, identified primarily through highly blueshifted C IV absorption and extending the velocity baseline beyond the conventional broad-absorption-line regime to $\sim 60{,}000\ {\rm km\,s^{-1}}$. EHVOs have been observed over $1.052<z_{\rm em}<7.641$, and their significance is amplified by the fact that, for otherwise comparable outflow properties, kinetic power scales as $v^3$ (Hidalgo et al., 2020, Hidalgo et al., 2022). In a distinct protostellar usage, the related term “extremely high-velocity (EHV) flow” denotes the compact, highest-speed molecular component of a young-stellar outflow, typically separated kinematically and morphologically from the broader lower-velocity molecular outflow (Matsushita et al., 2018).

1. Definition, scope, and terminology

In quasar work, EHVOs are an operationally selected class of very high-velocity broad C IV absorption outflows. The defining interval used in the first large SDSS survey is $0.1c$–$0.2c$, or roughly $30{,}000$–$60{,}000\ {\rm km\,s^{-1}}$, with absorption broader than $1000\ {\rm km\,s^{-1}}$ and deeper than 10% below the continuum (Hidalgo et al., 2020). This places EHVOs beyond the range usually emphasized in classical BAL surveys, which often imposed an upper velocity cutoff near $0.2c$0 to avoid confusion with neighboring features blueward of C IV emission (Hidalgo et al., 2022).

In protostellar studies, the terminology is different in detail but similar in spirit. The EHV flow is the compact, jet-like, highest-speed molecular component, contrasted with the broader molecular outflow. In MMS 5 / OMC-3, for example, the EHV flow is defined observationally by $0.2c$1–$0.2c$2, whereas the lower-velocity outflow occupies $0.2c$3–$0.2c$4 (Matsushita et al., 2018).

Domain	Velocity scale in the cited literature	Principal tracer(s)
Quasar EHVO	$0.2c$5–$0.2c$6; $0.2c$7–$0.2c$8	Broad C IV absorption
Protostellar EHV flow	$0.2c$9–$\sim 60{,}000\ {\rm km\,s^{-1}}$0	CO, SiO
Galaxy-scale high-velocity outflow	$\sim 60{,}000\ {\rm km\,s^{-1}}$1–$\sim 60{,}000\ {\rm km\,s^{-1}}$2 in one $\sim 60{,}000\ {\rm km\,s^{-1}}$3 Mg II study	Mg II absorption

Terminological slippage is common but not neutral. High-velocity narrow absorption lines at $\sim 60{,}000\ {\rm km\,s^{-1}}$4–$\sim 60{,}000\ {\rm km\,s^{-1}}$5 in quasars are not automatically assumed to be intervening systems: one study found that such high-velocity NALs are strongly correlated with AALs, BALs, and radio loudness, and that line-locked C IV doublets imply that physical line locking due to radiative forces is both common and real (Chen et al., 2021). By contrast, galaxy-scale Mg II outflows at $\sim 60{,}000\ {\rm km\,s^{-1}}$6 are described in the literature as “high-velocity outflows” rather than EHVOs in the standard quasar sense (Taylor et al., 2024).

2. Observational identification in quasars

The first systematic UV/optical quasar survey of EHVOs used the SDSS DR9 quasar catalog and defined a searchable parent sample of 6743 quasars after requiring $\sim 60{,}000\ {\rm km\,s^{-1}}$7, $\sim 60{,}000\ {\rm km\,s^{-1}}$8, and adequate spectral coverage of the C IV search region (Hidalgo et al., 2020). Candidate detection used a balnicity-like statistic over $\sim 60{,}000\ {\rm km\,s^{-1}}$9–$1.052<z_{\rm em}<7.641$0,

$1.052<z_{\rm em}<7.641$1

with $1.052<z_{\rm em}<7.641$2, $1.052<z_{\rm em}<7.641$3, and $1.052<z_{\rm em}<7.641$4 only when the absorption remained continuously deeper than 10% below continuum over at least $1.052<z_{\rm em}<7.641$5 (Hidalgo et al., 2020). Visual vetting then rejected ambiguous cases such as lower-velocity Si IV, N V, O VI, Ly$1.052<z_{\rm em}<7.641$6, or blends of narrow intervening lines.

That survey yielded 40 secure EHVO quasars containing 48 distinct EHVO troughs, because 6 of the 40 quasars show multiple detached high-velocity features (Hidalgo et al., 2020). C IV is the primary identification line, but line coincidence with other ions is central to intrinsicity arguments: 26/48 EHVO features have confirmed or likely N V at corresponding velocities, O VI is likely present in at least 6/48, strong matching Ly$1.052<z_{\rm em}<7.641$7 is rare, and only 7/40 are also traditional BALQSOs under the usual low-velocity C IV BAL definition (Hidalgo et al., 2020).

The boundary between EHVOs and other quasar absorption classes is not purely taxonomic. Extreme high-velocity NALs at $1.052<z_{\rm em}<7.641$8–$1.052<z_{\rm em}<7.641$9 complicate the older assumption that all such systems are intervening gas. Their strong correlation with AALs, BALs, and radio loudness, together with the presence of highly ionized line-locked C IV doublets, indicates that a significant fraction of high-velocity systems are either ejected from the quasars or form in material swept up by radio jets (Chen et al., 2021).

3. Population properties and connection to emission-line winds

A later study linked EHVO absorption to quasar emission-line wind signatures by cross-matching the DR9 EHVO sample to a DR14Q C IV emission-line catalog. After applying common quality cuts, the matched analysis contained 31 confirmed EHVO quasars, compared with a 41,535-object comparison set including 35,694 non-BALQSOs and 5,841 BALQSOs (Hidalgo et al., 2022). EHVO-hosting quasars cluster in the part of C IV parameter space characterized by large C IV blueshifts and low C IV equivalent widths, more extreme than the typical location of both non-BAL and BAL quasars (Hidalgo et al., 2022).

The statistical separation is strong. K–S tests comparing the EHVO C IV-blueshift distribution with those of the parent sample, non-BALQSOs, and BALQSOs each yield $v^3$0 and K–S statistics $v^3$1 (Hidalgo et al., 2022). Almost all such BALs and EHVOs have C IV blueshifts $v^3$2, and within the EHVO sample the maximum EHVO velocity shows a tentative tendency to increase with increasing C IV blueshift, extending a BALQSO trend to speeds up to $v^3$3 (Hidalgo et al., 2022).

He II is a second important discriminator. EHVO quasars occupy the large-C IV-distance, low-He II part of parameter space, and the largest He II equivalent widths seen among EHVO quasars are only about $v^3$4 Å, suggesting a tentative upper limit to He II strength for a quasar to host an EHVO (Hidalgo et al., 2022). Earlier survey work found that EHVO quasars lack He II emission and have overall larger bolometric luminosities and black hole masses than those of their parent sample and BALQSOs, while not showing significant differences in Eddington ratios (Hidalgo et al., 2020). The later emission-line analysis refined this by concluding that bolometric luminosities, black hole masses, and Eddington ratios are overall very similar to the general quasar population once their C IV emission properties are taken into account (Hidalgo et al., 2022).

The redshift distribution also points to a nontrivial demographic trend. In the SDSS DR9 survey, the median redshifts are 2.80 for EHVO quasars, 2.42 for the parent sample, and 2.39 for BALQSOs, and nearly one-third of EHVO quasars have $v^3$5 (Hidalgo et al., 2020). The same work states that EHVO quasars seem to be more predominant at large redshifts, and notes that at $v^3$6 they can be as common as or more common than BALQSOs, albeit with small-number statistics (Hidalgo et al., 2020).

4. Representative quasar case studies

PG0935+417 is a canonical ultraviolet EHVO case. It is a luminous, radio-quiet quasar at $v^3$7 with an intrinsic mini-BAL outflow spanning roughly $v^3$8 to $v^3$9, with the strongest absorption near $0.1c$0 (Hidalgo et al., 2010). O VI, N V, and C IV are all detected; the O VI doublet indicates moderate saturation and partial covering with $0.1c$1; the inferred ionization parameter is $0.1c$2, with a preferred $0.1c$3; and the total hydrogen column is $0.1c$4 for solar abundances (Hidalgo et al., 2010). The outflow emerged between 1982 and 1993, varied strongly thereafter, and showed no clear evidence for acceleration or deceleration (Hidalgo et al., 2010).

SDSS J023011.28+005913.6 provided a complementary multi-epoch variability case. It hosts two emergent C IV troughs: trough A at $0.1c$5 and trough B at $0.1c$6, with broad absorption changing on rest-frame timescales as short as $0.1c$7 days (Rogerson et al., 2015). The authors ruled out some simple bulk-motion models, proposed an augmented crossing-disk scenario for trough A with $0.1c$8, and a flow-tube model for trough B with $0.1c$9. If ionization variability drives the changes, the inferred lower limits are $0.2c$0 and $0.2c$1 for trough A, and $0.2c$2 and $0.2c$3 for trough B (Rogerson et al., 2015).

J164653.72+243942.2 pushed the physical characterization further. This $0.2c$4 quasar shows a variable EHVO with $0.2c$5 in one epoch and $0.2c$6 in another, accompanied by Si IV, N V, and Ly$0.2c$7 (Hidalgo et al., 19 Aug 2025). SimBAL modeling gives $0.2c$8 for the high-velocity component and total columns $0.2c$9 for the full 2004 trough and $30{,}000$0 for the 2011 EHVO (Hidalgo et al., 19 Aug 2025). Under adopted bounds of $30{,}000$1, the EHVO mass outflow rate is $30{,}000$2–$30{,}000$3 and the kinetic luminosity is $30{,}000$4–47.2 $30{,}000$5 (Hidalgo et al., 19 Aug 2025).

At still higher luminosity and redshift, SMSS J2157-3602 at $30{,}000$6 hosts an EHVO with $30{,}000$7, $30{,}000$8, and $30{,}000$9, together with a large C IV emission-line blueshift $60{,}000\ {\rm km\,s^{-1}}$0 (Vietri et al., 10 Sep 2025). Its SED-based luminosity is $60{,}000\ {\rm km\,s^{-1}}$1, making it the most luminous known quasar in the first 1.3 Gyr, and the paper concludes that the EHVO persists over rest-frame timescales of a few months to one year (Vietri et al., 10 Sep 2025).

5. Physical interpretation, energetics, and wind structure

Across the survey literature, the favored physical interpretation is that EHVOs arise under conditions favorable for radiative line driving. Two empirical patterns motivate this view: EHVO quasars are more luminous than comparison samples in the original SDSS survey, and they almost universally show weak or absent He II, implying a softer ionizing SED that leaves the gas at an ionization state where line opacity in ions such as C IV and N V is effective (Hidalgo et al., 2020). The later emission-line study reached a closely related conclusion: large C IV blueshifts, low C IV equivalent widths, and relatively weak He II place EHVO quasars in wind/SED states favorable for line driving rather than in a population defined primarily by unusual bulk luminosity or black-hole mass (Hidalgo et al., 2022).

The shielding problem is more nuanced than a single mechanism. PG0935+417 demonstrates that a quasar can host a $60{,}000\ {\rm km\,s^{-1}}$2 ultraviolet outflow with $60{,}000\ {\rm km\,s^{-1}}$3, typical of ordinary non-BAL quasars, and with X-ray absorption limited to $60{,}000\ {\rm km\,s^{-1}}$4 at 90% confidence (Hidalgo et al., 2010). That result challenges models in which strong radiative shielding in the X-rays/far-UV is essential for achieving extreme velocities. By contrast, SMSS J2157-3602 is X-ray weak-like, with $60{,}000\ {\rm km\,s^{-1}}$5, $60{,}000\ {\rm km\,s^{-1}}$6, and intrinsic X-ray absorption $60{,}000\ {\rm km\,s^{-1}}$7, and its paper argues that this X-ray weak nature likely prevents overionization of the innermost disk atmosphere and facilitates the efficient launch of both the EHVO and the BLR wind (Vietri et al., 10 Sep 2025). Taken together, this suggests that shielding and X-ray weakness are important in some EHVOs but not reducible to a single observational pattern.

A recent radiative-transfer study made the internal wind-structure problem explicit by modeling biconical disk winds with different acceleration lengths $60{,}000\ {\rm km\,s^{-1}}$8 (Dannen et al., 24 Jun 2026). In that framework, highly detached and blueshifted C IV absorption like that observed in EHVO quasars appears only in models with small $60{,}000\ {\rm km\,s^{-1}}$9: the gas reaches high velocity before attaining the ionization and density conditions favorable for C IV (Dannen et al., 24 Jun 2026). Models with larger $1000\ {\rm km\,s^{-1}}$0 instead produce broader, less detached troughs even when terminal velocities are very high. The paper therefore argues that highly detached and blueshifted absorption requires both a high terminal velocity and small $1000\ {\rm km\,s^{-1}}$1, and proposes an extended disk-wind view of AGN unification in which viewing angle $1000\ {\rm km\,s^{-1}}$2 selects the observed wind region while $1000\ {\rm km\,s^{-1}}$3 shapes the emergent spectrum (Dannen et al., 24 Jun 2026).

The feedback argument follows directly from velocity leverage. In standard shell-like parameterizations, $1000\ {\rm km\,s^{-1}}$4, so increasing the speed from $1000\ {\rm km\,s^{-1}}$5–$1000\ {\rm km\,s^{-1}}$6 to $1000\ {\rm km\,s^{-1}}$7 raises the mechanical power by roughly 1–2.5 orders of magnitude for otherwise similar gas properties (Hidalgo et al., 2020). Detailed studies of individual EHVO quasars show how sensitive the inferred impact remains to radius, column density, and covering factor. In J1646, the preferred radius range $1000\ {\rm km\,s^{-1}}$8 yields $1000\ {\rm km\,s^{-1}}$9–$0.2c$00 and $0.2c$01–47.2 $0.2c$02 for the EHVO alone (Hidalgo et al., 19 Aug 2025). In SMSS J2157, conservative assumptions using only the observed C IV phase give much smaller kinetic efficiencies, but the paper emphasizes that plausible larger radii and a very-high-ionization phase could move the EHVO into the efficient-feedback regime (Vietri et al., 10 Sep 2025). A plausible implication is that present UV measurements often provide lower bounds on EHVO feedback.

6. Other astrophysical usages: protostellar EHV flows and adjacent high-velocity outflows

In the star-formation literature, the analogous phenomenon is the EHV molecular jet. MMS 5 in OMC-3 provides a well-resolved example: CO $0.2c$03 traces both the EHV flow and the lower-velocity outflow, while SiO $0.2c$04 traces only the EHV flow (Matsushita et al., 2018). The EHV component is collimated, located at the root of the V-shaped outflow, and distinguished by $0.2c$05–$0.2c$06, compared with $0.2c$07–$0.2c$08 for the broader molecular outflow (Matsushita et al., 2018). Its knot spacing of about 220–280 AU implies episodic ejection every $0.2c$09–12 yr, and the preferred interpretation is a nested wind scenario, although jet entrainment cannot be completely ruled out (Matsushita et al., 2018).

Mid-$0.2c$10 CO observations sharpened that picture. In OMC-3 MMS 5, CO $0.2c$11–$0.2c$12 revealed a previously undetected EHV component located within a radius of $0.2c$13 from the source; the associated high-velocity gas has $0.2c$14–$0.2c$15 and $0.2c$16, and the paper concludes that mid-$0.2c$17 CO lines are good tracers of extremely high velocity gas in young outflows likely related to jets (Gómez-Ruiz et al., 2019). In the massive YSO / ultracompact-H II region G5.89-0.39, CO emission reaches $0.2c$18 on the blue side and shows a Hubble-like kinematic structure, while LVG analysis indicates that temperature increases with velocity; the observations qualitatively favor a jet-driven bow shock model (Su et al., 2011).

These protostellar EHV flows are physically and observationally distinct from quasar EHVOs, but the shared terminology reflects a common empirical structure: a compact, kinematically detached, highest-velocity component distinguished from broader and slower outflow material. By contrast, galaxy-scale Mg II outflows with $0.2c$19 to $0.2c$20 in recently quenched galaxies at $0.2c$21 are treated as high-velocity outflows rather than EHVOs in the standard quasar sense (Taylor et al., 2024). The term EHVO therefore denotes a family resemblance in extremeness, not a single velocity threshold or physical scale across all astrophysical subfields.