---
title: Excess EW Diagnostic Diagram
url: https://www.emergentmind.com/topics/excess-equivalent-width-diagnostic-diagram
type: topic
---

# Excess EW Diagnostic Diagram

Searching arXiv for recent and foundational papers on “excess equivalent width diagnostic diagram” and closely related EW-based diagnostics.
Search query: "excess equivalent width diagnostic diagram"
An excess equivalent width diagnostic diagram is an equivalent-width-centered diagnostic construction in which either an observed equivalent width, a luminosity-adjusted equivalent-width residual, or a wing-integrated excess relative to a reference profile is used as a classification variable. The term is not standardized across astrophysics. In cataclysmic-variable spectroscopy it denotes the red- and blue-wing excesses of an H\(\alpha\) emission profile relative to a Gaussian reference [2509.02858]. In emission-line galaxy work, analogous functions are served by diagrams that place \(\mathrm{EW}(H\alpha)\) against a line ratio or a kinematic observable, such as the EW\(\alpha\)n2, WHaN, WHaD, and WHaO families [0912.1643; 2311.10573; 2510.07256]. In quasar studies, the closest formalism is the luminosity-adjusted excess \(\Delta \log \mathrm{EW}\) defined relative to Baldwin relations [2206.11631]. A plausible implication is that the phrase operates as an umbrella label for EW-based diagnostic planes rather than as the name of a single canonical diagram.

## 1. Terminology and domain-specific meanings

The literature associates the phrase with several distinct, but structurally related, diagnostics. In each case, the central quantity is not merely a line flux but a line strength measured relative to a continuum or to a reference expectation. What changes between subfields is the reference: a Gaussian line core, a host-galaxy continuum, a luminosity-dependent mean relation, or a detection threshold.

| Context | Paper-specific construction | Primary variables |
|---|---|---|
| Accretion-disc winds in CVs | “excess equivalent width diagnostic diagram” | red-wing excess EW vs blue-wing excess EW |
| Emission-line galaxies | EW\(\alpha\)n2 / WHaN-like and WHaD / WHaO diagrams | \(\mathrm{EW}(H\alpha)\) with \([\mathrm{NII}]/H\alpha\), \(\sigma(H\alpha)\), or \([\mathrm{OIII}]/[\mathrm{OII}]\) |
| Quasars | Baldwin-residual EW diagnostics | \(\Delta \log \mathrm{EW}\), \(\Delta \alpha_{\rm ox}\), \(\log[\mathrm{EW}(\mathrm{CIV})/\mathrm{EW}(\mathrm{MgII})]\) |
| High-\(z\) emitters | EW-threshold and EW-displacement diagrams | \(\mathrm{EW}_{\mathrm{Ly}\alpha}\) or \(\mathrm{EW}_{\mathrm{H\beta+[OIII]}}\) versus metallicity-sensitive or structural observables |

This plurality of usage is explicit in the source literature. The 2025 CV study treats the excess-EW diagram as a named method for wind-formed emission lines [2509.02858]. By contrast, Cid Fernandes et al. define the EW\(\alpha\)n2 plane through \(\log([\mathrm{NII}]/H\alpha)\) and \(W_{H\alpha}\), and later work argues that \(\mathrm{EW}(H\alpha)\)-based diagrams outperform classical BPT classifications for separating star-forming galaxies, retired galaxies, and AGNs [0912.1643; 2510.07256]. Fu et al. do not use the phrase as a title, but they formalize “excess equivalent width” through luminosity-adjusted residuals in quasar UV lines [2206.11631].

## 2. Formal definitions of excess and equivalent width

The most generic equivalent-width definition appearing in the cited literature is the Ly\(\alpha\) form used for MUSE LAEs:
\[
EW \approx \frac{F^{line}_{Ly\alpha}}{f^{cont}_{Ly\alpha}},
\qquad
EW_0 = \frac{EW}{1+z}.
\]
Kerutt et al. use this to construct rest-frame Ly\(\alpha\) equivalent-width histograms and to distinguish secure measurements from lower limits when the UV continuum is undetected [2202.06642]. This formulation makes clear that “excess” can arise either from stronger line emission or from a weaker continuum.

Fu et al. introduce a formally residualized version tied to the Baldwin effect:
\[
\Delta \log \mathrm{EW}
=
\log \mathrm{EW}_{\rm obs}
-
\log \mathrm{EW}_{\rm exp}(L_{2500}),
\]
and also define a linear-space excess,
\[
\Delta \mathrm{EW}
=
\mathrm{EW}_{\rm obs}
-
\mathrm{EW}_{\rm exp}(L_{2500}).
\]
This turns equivalent width into a deviation from the mean EW–luminosity relation, so the diagnostic no longer depends on raw line strength alone [2206.11631].

In the CV wind implementation, the excess is instead a residual relative to a symmetric line model. The 2025 SIROCCO study defines
\[
EW_{\mathrm{Excess}}
=
\sum_{i=v_{\mathrm{low}}}^{v_{\mathrm{high}}}
\frac{F_{v_i}-F_{G_i}}{F_{c_i}}\,\Delta v_i,
\]
with separate evaluations on the blue and red wings of H\(\alpha\). Positive excess means the observed profile has more flux than the Gaussian in that wing; negative excess means a deficit or absorption relative to the reference profile [2509.02858].

A fourth formalism appears in COS absorption-line significance work, where the excess is naturally the margin above a limiting detectable equivalent width. Keeney et al. define a limiting EW,
\[
\mathcal{W}_{\lim}
=
\frac{N_{\sigma}\,\Delta\lambda}{({\rm S/N})_1}\,w(x),
\]
and the corresponding optimal significance
\[
N_{\sigma}^{\rm opt}
=
({\rm S/N})_1\,\frac{\mathcal{W}_\lambda}{\Delta\lambda}\,\frac{1}{w(x_{\rm opt})}.
\]
The paper explicitly notes that this formalism directly supports constructions such as
\[
W_{\rm ex} \equiv W_\lambda - W_{\lim},
\]
or normalized versions like \((W_\lambda-W_{\lim})/W_{\lim}\) [1206.2951]. Here the reference is not a population mean or a line model, but an instrument- and line-width-dependent detection threshold.

## 3. Emission-line galaxy classification with EW-centered planes

In galaxy spectroscopy, the most widely used EW-centered diagrams replace some of the four-line demands of classical BPT space with \(\mathrm{EW}(H\alpha)\)-based observables. Cid Fernandes et al. define the EW\(\alpha\)n2 diagram through
\[
x = \log \frac{[\mathrm{NII}]\lambda 6584}{H\alpha},
\qquad
y = W_{H\alpha},
\]
motivated by the fact that weak line galaxies often lack sufficiently strong \(\mathrm{H}\beta\) and/or \([\mathrm{OIII}]\) for standard BPT classification [0912.1643]. The optimized transposed boundaries are vertical cuts in \(\log([\mathrm{NII}]/H\alpha)\) for SF/AGN separation and a horizontal cut
\[
W_{H\alpha}=6\,\AA
\]
for Seyfert/LINER separation. The paper states that these alternative diagrams allow one to classify up to \(50\%\) more emission-line galaxies than the standard BPT scheme [0912.1643].

The WHaD diagram generalizes the same logic by combining equivalent width with kinematics rather than with a line ratio. Sánchez et al. define WHaD through \(\mathrm{EW}(H\alpha)\) and the instrumental-resolution-corrected H\(\alpha\) velocity dispersion, \(\sigma_{H\alpha}\), measured from the same line. Their empirical classification cuts are
\[
\mathrm{EW}(H\alpha)<3\,\AA \quad \Rightarrow \quad \text{retired},
\]
\[
\mathrm{EW}(H\alpha)>6\,\AA
\ \text{and}\
\sigma_{H\alpha}<57~{\rm km\,s^{-1}}
\quad \Rightarrow \quad \text{star-forming},
\]
\[
\mathrm{EW}(H\alpha)>3\,\AA
\ \text{and}\
\sigma_{H\alpha}>57~{\rm km\,s^{-1}}
\quad \Rightarrow \quad \text{AGN-like},
\]
with a further split at \(10\,\AA\) between weak and strong AGN-like classes [2311.10573]. The central physical idea is that low \(H\alpha\) equivalent width identifies emission compatible with HOLMES/post-AGB ionization, whereas high velocity dispersion moves the system toward AGN- or shock-like excitation.

A 2025 reassessment extends this family by comparing BPT, WHaN, WHaD, and the new WHaO diagram, which combines \(\mathrm{EW}(H\alpha)\) with \([\mathrm{OIII}]/[\mathrm{OII}]\). That analysis concludes that classical BPT methods overestimate star-forming galaxies by about \(10\%\), misclassify up to \(45\%\) of AGNs and nearly all retired galaxies, whereas diagnostics incorporating \(\mathrm{EW}(H\alpha)\) reduce misclassifications to \(\sim 20\%\) for AGNs and \(\sim 15\%\) for retired galaxies [2510.07256]. In this usage, equivalent width functions as a proxy for nebular emission in excess of what old stellar populations can sustain.

## 4. The accretion-disc wind implementation

The most literal use of the phrase “excess equivalent width diagnostic diagram” appears in work on optical emission lines formed or shaped by accretion-disc winds in cataclysmic variables. The 2025 SIROCCO grid study asks how one can recognize wind-formed optical lines when blueshifted absorption or a P-Cygni profile is absent. The diagnostic therefore targets line-shape residuals rather than absolute line strength [2509.02858].

The method is operationally simple. One fits the continuum, normalizes the spectrum, fits a Gaussian to the emission line, subtracts that Gaussian, and integrates the residual over chosen wing regions on each side. Each spectrum is then represented by a point \((EW_{\rm red,excess},EW_{\rm blue,excess})\). The classic wind-like quadrant is the south-east one,
\[
EW_{\rm red,excess}>0,
\qquad
EW_{\rm blue,excess}<0,
\]
which corresponds to red-wing excess and blue-wing deficit, the residual signature most closely associated with P-Cygni-like asymmetry [2509.02858].

The paper’s main methodological conclusion is that the diagram is highly sensitive to the adopted definition of the wing. Fixed masks such as \(\pm1000\)–\(2500~{\rm km\,s^{-1}}\) can allow strong core contamination, especially for double-peaked or broad profiles. The authors therefore propose an adaptive definition tied to the full width at half maximum: the inner boundary at \(1\times{\rm FWHM}\) from line center and the outer boundary at \(5\times{\rm FWHM}\). In wavelength-space notation, the preferred blue and red integration windows are
\[
[\lambda_0-5\,{\rm FWHM},\ \lambda_0-{\rm FWHM}]
\]
and
\[
[\lambda_0+{\rm FWHM},\ \lambda_0+5\,{\rm FWHM}].
\]
Using a grid of \(3645\) synthetic H\(\alpha\) profiles, the paper finds that about \(20\%\) of all lines belong to a “Gold” sample with strengths and widths consistent with observations, and that this sample can preferentially populate the proposed wind regions, although many wind-formed lines still cluster near the \(y=x\) diagonal of symmetric residuals [2509.02858]. The implication is that the diagram is diagnostic but not definitive.

## 5. Baldwin-residual and high-redshift EW diagnostics

In quasar spectroscopy, excess equivalent width is formalized against luminosity trends rather than against line-shape symmetry. Fu et al. analyze luminous radio-quiet quasars with \(\mathrm{EW}(\mathrm{C\,IV})\ge 150\,\AA\) and define both \(\Delta \log \mathrm{EW}\) and \(\Delta \mathrm{EW}\) relative to the expected Baldwin relation at fixed \(L_{2500}\) [2206.11631]. Their mean fractional enhancements are highly asymmetric across lines:
- C IV: \(\Delta \mathrm{EW}/\mathrm{EW} = 2.33 \pm 0.14\)
- He II: \(1.23 \pm 0.22\)
- Mg II: \(0.61 \pm 0.10\)
- C III]: \(0.20 \pm 0.10\)

This pattern is the empirical basis for treating excess-EW space as physically diagnostic rather than as a mere restatement of line strength. The paper argues that a hard ionizing continuum explains the He II excess well, but that the full pattern requires an additional metallicity dependence, best captured by the combination of \(\Delta \alpha_{\rm ox}\) and \(\log[\mathrm{EW}(\mathrm{CIV})/\mathrm{EW}(\mathrm{MgII})]\) [2206.11631].

At high redshift, large rest-frame EW is often treated as a selection axis for chemically young systems. A JWST/NIRSpec study of MUSE-selected LAEs splits the sample at \(\mathrm{EW}_{\mathrm{Ly}\alpha}=90\,\AA\), finding that the high-EW stack, with median \(\mathrm{EW}_{\mathrm{Ly}\alpha}=154\,\AA\), has lower
\[
\mathrm{N2}=\log\left(\frac{[\mathrm{NII}]\,\lambda6584}{H\alpha}\right)
\]
and lower
\[
\mathrm{R3}=\log\left(\frac{[\mathrm{OIII}]\,\lambda5007}{H\beta}\right)
\]
than the lower-EW stack, whose median is \(43\,\AA\). The shifts are \(1.9\sigma\) in N2 and \(2.2\sigma\) in R3, and are interpreted as evidence for lower gas-phase metallicity in the highest-EW LAEs [2304.08511]. This construction is not named an excess-EW diagram, but it uses EW as the primary diagnostic coordinate.

A complementary JWST medium-band imaging study identifies \(119\) H\(\beta\)+[OIII] emitters at \(z\sim7\) from F410M excess. Their rest-frame equivalent widths span approximately \(420\)–\(6850\,\AA\), with median \(\sim1700\,\AA\), and the paper defines EELGs by \(EW_0>1000\,\AA\) and a more extreme subset by \(EW_0>3000\,\AA\) [2507.13456]. The highest-EW systems are characterized by low stellar mass, blue \(\beta_{\rm UV}\), low \(A_V\), and frequent offsets between nebular and stellar/UV centroids; thirteen emitters show spatially offset H\(\beta\)+[OIII] emission, which the paper interprets as likely feedback-driven shocks or outflows rather than simple in situ H II-region photoionization [2507.13456].

Kerutt et al. add a large MUSE-based statistical baseline for Ly\(\alpha\) EWs. Across \(1920\) LAEs, the combined sample yields \(16\%\) of objects with \(EW_0>240\,\AA\) when lower limits are included, but only \(6\%\) among secure measurements. The paper also finds a strong anti-correlation between UV effective radius and \(EW_0\), whereas line asymmetry, peak separation, and FWHM do not provide a clean high-EW separator [2202.06642]. This makes compactness a more robust secondary coordinate than line-profile shape for LAE excess-EW diagrams.

## 6. Related usages, ambiguities, and limits

Several nearby concepts clarify what the phrase does and does not mean. In mid-infrared galaxy spectroscopy, Rowan-Robinson and Efstathiou analyze the Spoon IRS diagnostic plane of \(6.2\,\mu{\rm m}\) PAH equivalent width versus \(9.7\,\mu{\rm m}\) silicate optical depth. Their main conclusion is that equivalent-width diagnostics are informative but aliased: starburst age, optical depth, cirrus contribution, and AGN torus fraction can overlap in the same 2D plane, so the diagram must be supplemented by full IRS spectral fitting and \(25\)–\(850\,\mu{\rm m}\) photometry [0907.0213]. This is a general warning against treating any EW-based diagram as uniquely identifying a physical mechanism.

A similar caution appears in AGN work on \([\mathrm{OIII}]\,\lambda5007\). Caccianiga and Severgnini show that the observed \([\mathrm{OIII}]\) EW range from a few \(\AA\) to \(\sim500\,\AA\) is explained by a combination of intrinsic NLR spread and obscuration. They argue that intrinsic variation dominates below \(40\)–\(50\,\AA\), moderate absorption can raise EW to \(\sim100\)–\(150\,\AA\), and stronger obscuration can push it to \(\sim500\,\AA\) [1104.1348]. In this context, an “excess” EW is a continuum-suppression signature rather than a line-production signature.

Transient spectroscopy supplies yet another variant. De Ugarte Postigo et al. construct “EW diagrams” for \(69\) GRB afterglow spectra and compress them into a line strength parameter,
\[
LSP=\frac{1}{N}\sum_i
\frac{\log EW_i-\langle \log EW\rangle_i}{\sigma_{\log EW,i}},
\]
which measures whether a burst has stronger- or weaker-than-average absorption features relative to the GRB population [1209.0891]. The paper also notes a larger excess in the EW of \(\mathrm{CIV}\,\lambda\lambda1549\) relative to QSO DLA systems. Here the diagnostic frame is population-referenced rather than geometry-referenced.

Finally, terminology can be genuinely misleading. In the ARES+MOOG stellar-atmosphere workflow, the queried phrase does not occur at all. The relevant diagnostic is the classic abundance-versus-Reduced-EW plot, used together with abundance versus excitation potential and the mean Fe I–Fe II offset to enforce excitation and ionization balance [1711.01839]. The paper explicitly uses “Reduced EW” or “R.W.”, not “excess equivalent width.” This contrast is important because it shows that not every “EW diagnostic diagram” is an excess-EW diagram in the residual or thresholded sense.

In aggregate, the literature supports a precise but plural definition. An excess equivalent width diagnostic diagram is best understood as any diagram that uses equivalent width after subtraction of a reference expectation—whether a Gaussian profile, a Baldwin relation, a retired-galaxy baseline, or an instrumental detection threshold—to isolate physically distinct populations. The strength of the method lies in its ability to encode continuum-normalized line information that ordinary flux or ratio diagnostics omit. Its main limitation is equally consistent across domains: excess in equivalent width is almost never unique in physical interpretation without a second axis or an external constraint.

Source: https://www.emergentmind.com/topics/excess-equivalent-width-diagnostic-diagram