{"id":"db6ddd3a-ea4c-4e13-a52c-2c7112bb35d7","arxiv_id":"2412.13100","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"The SepOS separatrix turbulence model, previously tested only on ASDEX Upgrade, is validated on Alcator C-Mod and extended to I-mode, ELMy-EDA transitions, and SPARC operating projections.","lead":"This paper tests a turbulence-based model of tokamak edge boundaries, called SepOS, against 14 years of Alcator C-Mod data and finds it can separate L-mode, H-mode, I-mode, and two kinds of H-mode. It then uses the same boundaries to estimate where the SPARC fusion device can operate safely, which matters for reactor design.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Concern: the empirical λ_pe regression (Eq. 4, Table 2) is trained on the same separatrix identification used to build the dimensional SepOS curves, so the favorable-drift L-H, LDL, and IBML plots are not out-of-sample predictions; a held-out or perturbation test is needed.","rationale":"The reader's weakest_assumption identifies the two-point-model power balance as fragile, which I agree is the root of the concern. However, the most load-bearing failure mode is not simply 'a few eV error shifts everything' as a sensitivity statement; it is that the λ_pe regression used to dimensionalize the boundaries is fit on the same data and the same inferred separatrix, so it can absorb systematic errors in the separatrix identification. This makes the dimensional boundary plots and the SPARC projections partially self-consistent rather than independently predictive. The dimensionless tests in Figure 4 are less affected, and the ELMy/EDA k_EM = k_RBM analysis in Figures 8-9 is a genuinely interesting independent check, though it still uses the same inferred λ_pe. I do not think this requires REJECT: the paper is a validation exercise with moderate fitting steps, and the authors acknowledge the limitations. CONDITIONAL is the right verdict, ideally with an explicit requirement for an out-of-sample or sensitivity test.","tokens_in":28601,"tokens_out":1729,"duration_ms":16305,"concrete_test":"Recompute the dimensional SepOS curves of Figures 3 and 5 using a λ_pe regression trained only on a randomly selected half of the H-mode discharges (or on the ELMy/EDA subset), then test whether held-out H-modes and L-modes are still separated by the L-H boundary with error bars. If separation degrades significantly (e.g., >20% of H-modes cross to the L-mode side), the regression is absorbing separatrix-identification bias and the claim of prediction is weakened. Additionally, perturb the two-point-model inputs (f_e,cond from 0.325 to 0.2/0.45, η_ICRF from 1 to 0.8, and allow finite T_e,div) and check whether the fitted exponents in Table 2 and the boundary positions in Figure 3 shift by more than the stated uncertainties.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The paper's central claim is that SepOS predicts the L-H transition, LDL, and IBML in separatrix (ne, Te) space. But the dimensional curves in Figures 3 and 5 are not pure predictions: they are constructed using the λ_pe(α_t, ρ_s,p) regression from Eq. 4, fit to the very same H-mode dataset that is then shown to lie on the correct side of the L-H and IBML curves. Since R_sep, n_sep, T_sep, and λ_pe all come from one iterative two-point-model fit, the regression can absorb systematic errors in the separatrix identification (e.g., Spitzer-Haerm conduction, f_e,cond = 0.325, η_ICRF = 1, T_e,div^7/2 << T_e,up^7/2, or uniform poloidal power flow). If T_e,sep from Eq. 3 is biased, the inferred λ_pe values are biased in the same direction, and the fitted C_α, a, C_ρ, r may partially compensate. The dimensionless separation in Figure 4 is more robust because it does not use the λ_pe regression, but the paper's central dimensional projections, including the SPARC extrapolation in Figure 11, rely on this regression. The reader's weakest_assumption is close, but the sharper mechanism is regression compensation: the fit can hide separatrix-identification bias. The paper itself notes the λ_pe fit coefficients change substantially when Type-I ELMy H-modes are included (Section 6 footnote), confirming sensitivity to dataset composition.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper analyzes Alcator C-Mod edge Thomson scattering data to test the separatrix operational space (SepOS) model for three boundaries (L-H transition, L-mode density limit, ideal ballooning MHD limit) in both favorable and unfavorable grad-B drift directions. It constructs a regression for the electron pressure gradient scale length lambda_pe as a function of alpha_t and rho_s,p, uses it to render the SepOS boundaries in (n_e, T_e) space, extends the framework to I-mode access and the ELMy-EDA H-mode transition, and projects the resulting boundaries to the SPARC primary reference discharge. The central claim is that DALF-normalized separatrix parameters organize C-Mod data across a wide range of engineering parameters, supporting a turbulence-based description of operational boundaries.","tokens_in":28987,"tokens_out":6061,"duration_ms":56038,"significance":"This is the first validation of the SepOS model on a device other than ASDEX Upgrade, and it uses a large multi-year C-Mod database with careful edge Thomson scattering analysis and propagated uncertainties on separatrix quantities. If the observed separations are robust, the paper substantially strengthens the case that separatrix DALF parameters control L-H access, the density limit, and the MHD ballooning limit, and it provides useful guidance for SPARC. The dimensionless tests in Figure 4 are the strongest part of the paper because they use measured lambda_pe rather than the fitted regression. However, several load-bearing parameters are selected after the fact (alpha_RS, lambda_LDL_pe), and the dimensional boundary curves rely on a regression trained on the same dataset, so the predictive claims need qualification and additional robustness tests.","major_comments":[{"comment":"The lambda_pe regression is fit to the same dataset and the same separatrix-identification procedure that is then used to construct the dimensional SepOS boundaries, so Figures 3, 5, and 11 are consistency checks rather than out-of-sample predictions; any systematic bias in T_e,sep or R_sep from Eq. (3) can be partially absorbed by C_alpha, a, C_rho, and r. The authors should add a held-out test (for example, training on one configuration or year and testing on another, or leave-one-shot-out cross-validation) or a perturbation analysis that shifts the fitted coefficients within their covariances and reports how many points change side of each boundary. The footnote in Section 6 already shows that the coefficients change substantially when Type-I ELMy H-modes are added, so this sensitivity is real and should be quantified in the main text.","section":"Section 3.3, Eq. (4), Table 2; Figs. 3, 5, 11"},{"comment":"The unfavorable-drift L-H validation depends on alpha_RS: the value 0.5 is selected because it best separates the C-Mod data in Figure 5, and the alpha_RS(q_cyl) curve in Figure 6 is a manually chosen exponential fit to the same data. Even though the proximity to the AUG value of 0.4 is encouraging, the selection is still post hoc. No quantitative separation metric, uncertainty estimate, or cross-validation is provided, so the claim that the model applies to the unfavorable drift direction is supported only by tuning a free parameter. Please report misclassification rates for fixed alpha_RS values in a plausible range (e.g., 0.4-0.6) and for the empirical alpha_RS(q_cyl) curve, and test on a subset of discharges not used to choose alpha_RS.","section":"Section 5, Figs. 5-6"},{"comment":"The dimensional LDL curve in Figure 3 uses lambda_LDL_pe = 10 mm, described as the empirically observed value for the highest-density L-modes in this dataset; this is an after-the-fact choice, and the SPARC projection in Figure 11 uses the arbitrary values 2.5, 5, and 10 mm. The dimensionless LDL test in the center panel of Figure 4 is more convincing because it uses measured lambda_pe, but the dimensional claim that SepOS predicts the density limit is not supported by an independent parameter. Show how the LDL boundary and the classification of L-modes change over the plausible range of lambda_pe, or construct the curve from the measured lambda_pe of each discharge rather than from a fixed chosen value.","section":"Section 4.2, Fig. 3 caption"},{"comment":"All SepOS coordinates and lambda_pe values are determined by the two-point-model power balance, which assumes Spitzer-Harm parallel conduction, T_e,div^7/2 much less than T_e,up^7/2, uniform poloidal power flow, f_e,cond = 0.325, and eta_ICRF = 1. The paper propagates +/-20% PSOL uncertainty but does not propagate uncertainty in f_e,cond or the validity of the Spitzer-Harm assumption, which the authors note may fail at high power and low density. Add a sensitivity scan over f_e,cond (and, if possible, a kinetic-correction proxy) and demonstrate that the boundary classifications in Figures 4-6 and 8-9 are stable, or quantify which points change classification.","section":"Section 3.1, Eq. (3)"},{"comment":"The claim that k_EM = k_RBM describes the ELMy-EDA transition is not uniquely supported: alpha_t = 0.55 and beta_e = 10^-4 separate the phases equally well in Figure 8, and the values alpha_t = 0.53 and beta_e = 9.9 x 10^-5 obtained by fitting beta_e(alpha_t) and lambda_pe(alpha_t) and solving Eq. (12) are essentially fits to the same transition data. Because k_EM and k_RBM share beta_e and lambda_pe through omega_B, the break in slope in Figure 9 is expected from the correlations in Figure 10. The authors should state that this is an exploratory consistency check rather than a validation, and provide error bars on the fitted transition parameters.","section":"Section 6, Figs. 8-10"}],"minor_comments":[{"comment":"The text contains several typos: 'enahnced Dalpha' in Section 3.2, 'ork on AUG' in Section 6, 'T ransition' in the Section 6 heading, and 'denisty' in Reference [33].","section":"Throughout"},{"comment":"Notation for the pressure scale length is inconsistent: lambda_p, lambda_pe, and lambda_p,e are used interchangeably; choose one symbol and define it once.","section":"Throughout"},{"comment":"In Equation (8), alpha_c, tau_i, and Lambda_pi appear without definitions in the main text; define them when first introduced, even if they are defined in the cited references.","section":"Eq. (8)"},{"comment":"The captions of Figure 4 and Figure 6 should state explicitly which equation each panel implements (Eqs. (8), (9), and (10) for Figure 4) and should give the functional form of the empirical alpha_RS(q_cyl) curve shown in Figure 6.","section":"Figs. 4 and 6"},{"comment":"The regression coefficients in the footnote marked with a dagger are important because they differ strongly from the main H-mode regression; move this discussion into the main text and quantify the comparison.","section":"Section 6, footnote"},{"comment":"Figure 11 would benefit from a table listing all linestyles and colors and the lambda_pe scaling used for each boundary, since the caption currently describes solid, dashed, and dash-dotted curves without a legend.","section":"Fig. 11"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a strong candidate for publication after revision. The main risk is overclaiming prediction when several parameters are fitted post hoc; the requested sensitivity and held-out tests are feasible with existing data and would substantially strengthen the paper. I would not reject, but the revision should be substantive, not cosmetic."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The real news here is the dimensionless separation. Figure 4 shows L-, H-, and I-modes separating across a huge C-Mod database in the DALF coordinates, using measured lambda_pe and no fitted regression. That is a genuine cross-device test of the SepOS model, and it works. The new C-Mod lambda_pe regression with positive exponents on alpha_t and rho_s,p is also a useful result, and the ELMy-EDA analysis proposing k_EM = k_RBM as the separator is a reasonable, falsifiable hypothesis. The paper is honest about its assumptions; Section 3.1 lists the two-point-model simplifications, Section 7 admits the lambda_pe scaling uncertainty, and the footnote in Section 6 reports that including Type-I ELMy H-modes changes the regression coefficients substantially.\n\nThe soft spot is exactly where the stress-test note lands. The dimensional curves in Figures 3, 5, and 8 are built from the lambda_pe regression fit to the same H-mode dataset that is then shown to sit on the correct side of those curves. That is not an out-of-sample prediction. Worse, the regression can compensate for a systematic bias in the separatrix identification from the two-point model: if T_sep is off, lambda_pe is off, and the fitted coefficients can partially hide that. The dimensionless result in Figure 4 does not have this problem, so the central claim survives. But the SPARC projections in Figure 11, and any practical use of the dimensional boundaries, inherit the risk. The ad hoc choices of alpha_RS = 0.5 in the unfavorable drift direction and lambda_LDL_pe = 10 mm are minor by comparison; the paper presents them as empirical adjustments, not theory.\n\nI also wish the code and data were shipped. The fitting steps are not complicated, and a reader cannot independently reproduce the regression or the separatrix iteration without them.\n\nThis paper is for plasma-edge and divertor physicists, and for anyone using SepOS to project ITER or SPARC operating windows. It deserves a serious referee. Send it to review, and ask for a held-out or perturbation test of the lambda_pe regression, or at minimum an explicit statement of the regression-compensation risk. With that addressed, it is a solid contribution.","headline":"First cross-device SepOS validation on C-Mod shows clean L/H/I separation in dimensionless coordinates, but the dimensional boundary curves lean on a self-trained lambda_pe regression that needs a held-out check.","tokens_in":29606,"tokens_out":1871,"would_cite":true,"duration_ms":20275,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["52.55.Fa","52.35.Ra"],"model":"deepseek-v4-flash","headline":"This paper shows that a separatrix-based turbulence model, built from interchange-drift-Alfvén balances, predicts the L-H transition, the L-mode density limit, and the ideal MHD ballooning limit across a broad range of Alcator C-Mod…","keywords":["separatrix operational space","L-H transition","L-mode density limit","ideal MHD ballooning limit","interchange-drift-Alfvén turbulence","Alcator C-Mod","I-mode","SPARC"],"falsifier":"Run a dedicated density ramp on a tokamak with directly measured divertor electron temperature: if the plasma disrupts at a separatrix density clearly off the predicted $k_{\\mathrm{EM}} = k_{\\mathrm{RBM}}$ curve, or if the measured divertor temperature makes the $T_{e,\\mathrm{div}} \\ll T_{e,\\mathrm{sep}}$ assumption fail by more than a few eV, the SepOS boundary identification would be contradicted.","tokens_in":28383,"feed_emoji":"⚛️","tokens_out":15178,"duration_ms":126007,"temperature":0.7,"pith_summary":"The paper sets out to show that the separatrix operational space (SepOS) model, a set of turbulence-balance criteria evaluated at the plasma separatrix, organizes the operational boundaries of Alcator C-Mod the same way it organizes those of the device on which it was originally developed. Using edge Thomson scattering data spanning a wide range of density, toroidal field, and poloidal field, the authors identify the separatrix by a two-point power balance and compute the dimensionless quantities $\\alpha_t$ (a collisionality-like turbulence control parameter), $k_{\\mathrm{EM}}$ (the electromagnetic wavenumber), and $k_{\\mathrm{RBM}}$ (the resistive-ballooning wavenumber). They report that the L-H transition, the L-mode density limit, and the ideal MHD ballooning limit all fall on the model's predicted curves, and that an empirical scaling $\\lambda_{p_e} = (1 + C_\\alpha \\alpha_t^a) C_\\rho \\rho_{s,p}^r$ captures turbulence-driven widening of the pressure-gradient scale length. In the unfavorable drift direction the same boundaries hold with a reduced Reynolds-stress factor $\\alpha_{\\mathrm{RS}} < 1$, while I-modes cluster at $\\alpha_t \\lesssim 0.35$; the Type-I ELMy/EDA transition is better described by $k_{\\mathrm{EM}} = k_{\\mathrm{RBM}}$ than by a fixed $\\alpha_t = 0.55$. The payoff is that dimensionless turbulence balances can be translated into the dimensional $(n_e, T_e)$ space a control room can act on, and projected to the SPARC primary reference discharge.","feed_headline":"Turbulence model predicts C-Mod's three confinement limits","feed_subtitle":"Validated on a second tokamak, the separatrix framework projects H-mode access and limit avoidance for SPARC.","key_machinery":"The load-bearing object is the SepOS model, a set of identifications between turbulence quantities normalized by the DALF (interchange-drift-Alfvén) equations and evaluated at the separatrix. The L-H criterion is the energy balance of Equation 8, equating the stabilizing Reynolds-stress term $\\alpha_{\\mathrm{RS}} k_{\\mathrm{EM}} \\tau_i \\Lambda_{p_i}/\\bigl(1 + (\\alpha_t/\\alpha_c k_{\\mathrm{EM}})^2\\bigr)$ to the destabilizing turbulent energy input $\\alpha_t/\\alpha_c (k_{\\mathrm{EM}}^2 + 1/2) + \\tfrac12 k_{\\mathrm{EM}}^2 \\sqrt{\\omega_B \\tau_i \\Lambda_{p_i}}$. The density limit and the ideal MHD ballooning limit are wavenumber equalities, $k_{\\mathrm{EM}} = k_{\\mathrm{RBM}}$ and $k_{\\mathrm{ideal}} = k_{\\mathrm{RBM}}$, where $k_{\\mathrm{EM}} = \\sqrt{\\beta_e/\\mu}$ and $k_{\\mathrm{RBM}}$ is the resistive-ballooning wavenumber. The fourth ingredient is the empirical scaling $\\lambda_{p_e} = (1 + C_\\alpha \\alpha_t^a) C_\\rho \\rho_{s,p}^r$, which converts these dimensionless balances into the dimensional $(n_e, T_e)$ space used for the projections. The parameter $\\alpha_t$ controls electron adiabaticity and thus the balance between interchange and drift-wave driving, which is why the same parameter appears in the scale-length widening, the I-mode ceiling, and the ELMy/EDA separation.","core_discovery":"The paper's central claim is that the separatrix operational space (SepOS) model predicts three confinement boundaries on Alcator C-Mod: the L-H transition, the L-mode density limit, and the ideal MHD ballooning limit. In the favorable drift direction the L-H transition is described by a balance between Reynolds-stress energy transfer into the shear flow and the total turbulent energy input; the L-mode density limit is identified with the wavenumber equality $k_{\\mathrm{EM}} = k_{\\mathrm{RBM}}$; and the ideal MHD ballooning limit is identified with $k_{\\mathrm{ideal}} = k_{\\mathrm{RBM}}$. In the unfavorable drift direction the same three boundaries apply once the Reynolds-stress term is reduced, $\\alpha_{\\mathrm{RS}} < 1$, and I-modes occupy the low-$\\alpha_t$ side of the L-H curve. For the EDA/Type-I ELMy transition, the data support the wavenumber balance $k_{\\mathrm{EM}} = k_{\\mathrm{RBM}}$ as the better separator. The model is then used to project H-mode access, density-limit avoidance, and ELM-free operating space for the SPARC primary reference discharge.","pith_inferences":["A natural next step is to apply the same separatrix procedure to a third tokamak with different aspect ratio and shaping; if the dimensionless curves still separate L/H modes and disruptive density limits, the claim that turbulence physics dominates atomic physics at the separatrix would be substantially strengthened.","The apparent dependence of $\\alpha_{\\mathrm{RS}}$ on $\\hat{q}_{\\mathrm{cyl}}$ suggests plasma current could be used as a deliberate actuator to favor I-mode access in the unfavorable drift direction, a consequence the paper only notes qualitatively.","Because the SPARC density-limit projection depends on an untested value of $\\lambda_{p_e}^{\\mathrm{LDL}}$, the quickest way to reduce projection uncertainty would be a multi-machine scaling of $\\lambda_{p_e}$ against $\\alpha_t$ and $\\rho_{s,p}$.","If the EDA/ELMy boundary is genuinely $k_{\\mathrm{EM}} = k_{\\mathrm{RBM}}$ rather than a fixed $\\alpha_t$, then the transition density should vary with toroidal field and shaping in a way that targeted density scans could check."],"forward_implications":["If the SepOS boundaries hold, confinement regime access is set by turbulence balances at the separatrix, so a control room could use separatrix density and temperature, rather than core or pedestal quantities, to steer away from disruptive limits.","The positive exponents in the $\\lambda_{p_e}$ regression imply that near-SOL pressure-gradient scale lengths widen as $\\alpha_t$ rises and shrink with increasing $B_p$, a trend consistent with the multi-machine power-width scaling at low $\\alpha_t$.","In the unfavorable drift direction, a reduced $\\alpha_{\\mathrm{RS}}$ raises the separatrix temperature required for H-mode and explains both the higher power threshold and the confinement of I-modes to $\\alpha_t \\lesssim 0.35$.","The crossing $k_{\\mathrm{EM}} = k_{\\mathrm{RBM}}$ provides a wavenumber-based criterion for the EDA/ELMy transition, tying the disappearance of Type-I ELMs to the equilibration of resistive-ballooning and electromagnetic turbulence scales.","Projected to the SPARC primary reference discharge, the model predicts a minimum separatrix density for H-mode near $1.5 \\times 10^{20}\\,\\mathrm{m}^{-3}$ and outlines a region where EDA-like ELM-free operation avoids the Type-I ELMy regime."],"supporting_citations":[{"why":"Introduces the SepOS model and its three separatrix boundaries that this paper validates on a new device.","marker":"[10]"},{"why":"Supplies the $\\alpha_t$ and $\\rho_{s,p}$ scaling form for $\\lambda_p$ and the AUG regression coefficients against which the C-Mod results are compared.","marker":"[13]"},{"why":"Provides the two-parameter phase-space picture of edge turbulence from which the SepOS wavenumber balances derive.","marker":"[12]"},{"why":"Introduces the $\\alpha_{\\mathrm{RS}} < 1$ correction to the L-H criterion for the unfavorable drift direction, which this paper applies to C-Mod.","marker":"[16]"},{"why":"Identifies the $\\alpha_t = 0.55$ Type-I ELMy/QCE boundary against which the C-Mod EDA/ELMy transition is compared.","marker":"[15]"},{"why":"Earlier C-Mod edge study showing electromagnetic fluid drift turbulence organizes near-SOL scale lengths and the L-H and density boundaries.","marker":"[17]"},{"why":"Establishes the connection between separatrix ballooning stability and the H-mode density limit and supplies the fitting procedure used here.","marker":"[31]"},{"why":"Defines the common evaluation procedure for separatrix electron parameters used to build the C-Mod database.","marker":"[33]"},{"why":"Provides the earlier SepOS projection to next-step devices and SPARC scenarios that this paper's SPARC section builds on.","marker":"[74]"}],"fun_headline_variants":["One model predicts C-Mod's three confinement boundaries","SepOS model maps L-H, density limit, ballooning on C-Mod","C-Mod tests separatrix model across all confinement limits","Separatrix model forecasts SPARC limits from C-Mod data","Turbulence model predicts three tokamak regime edges"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire analysis rests on locating the separatrix with a power-balance formula that assumes a standard collisional heat-conduction law and a fixed electron conduction fraction, with no direct measurement of the divertor temperature; if the inferred separatrix electron temperature is off by even a few electron-volts, the separatrix radius, density, scale lengths, and every boundary curve shift.","fun_headline_variants_meta":{"raw":{"variants":["One model predicts C-Mod's three confinement boundaries","SepOS model maps L-H, density limit, ballooning on C-Mod","C-Mod tests separatrix model across all confinement limits","Separatrix model forecasts SPARC limits from C-Mod data","Turbulence model predicts three tokamak regime edges"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000538,"raw_usage":{"total_tokens":2725,"prompt_tokens":1229,"completion_tokens":1496,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":845,"completion_tokens_details":{"reasoning_tokens":1420}},"tokens_in":845,"tokens_out":1496,"duration_ms":11479,"temperature":1.0,"reasoning_tokens":1420,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T13:26:34.593148+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run a dedicated density ramp on a tokamak with directly measured divertor electron temperature: if the plasma disrupts at a separatrix density clearly off the predicted $k_{\\mathrm{EM}} = k_{\\mathrm{RBM}}$ curve, or if the measured divertor temperature makes the $T_{e,\\mathrm{div}} \\ll T_{e,\\mathrm{sep}}$ assumption fail by more than a few eV, the SepOS boundary identification would be contradicted.","supporting_citations":[{"cited_title":"The separatrix operational space of ASDEX Upgrade due to interchange-drift-Alfve ´n turbulence","cited_arxiv_id":null,"evidence_quote":"Introduces the SepOS model and its three separatrix boundaries that this paper validates on a new device."},{"cited_title":"Turbulence driven widening of the near-SOL power width in ASDEX Upgrade H-Mode discharges","cited_arxiv_id":null,"evidence_quote":"Supplies the $\\alpha_t$ and $\\rho_{s,p}$ scaling form for $\\lambda_p$ and the AUG regression coefficients against which the C-Mod results are compared."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the two-parameter phase-space picture of edge turbulence from which the SepOS wavenumber balances derive."},{"cited_title":"Grover, T","cited_arxiv_id":null,"evidence_quote":"Introduces the $\\alpha_{\\mathrm{RS}} < 1$ correction to the L-H criterion for the unfavorable drift direction, which this paper applies to C-Mod."},{"cited_title":"Faitsch, T","cited_arxiv_id":null,"evidence_quote":"Identifies the $\\alpha_t = 0.55$ Type-I ELMy/QCE boundary against which the C-Mod EDA/ELMy transition is compared."},{"cited_title":"Evidence for electromagnetic fluid drift turbulence controlling the edge plasma state in the Alcator C-Mod tokamak","cited_arxiv_id":null,"evidence_quote":"Earlier C-Mod edge study showing electromagnetic fluid drift turbulence organizes near-SOL scale lengths and the L-H and density boundaries."},{"cited_title":"Eich, R.J","cited_arxiv_id":null,"evidence_quote":"Establishes the connection between separatrix ballooning stability and the H-mode density limit and supplies the fitting procedure used here."},{"cited_title":"Silvagni, et al","cited_arxiv_id":null,"evidence_quote":"Defines the common evaluation procedure for separatrix electron parameters used to build the C-Mod database."},{"cited_title":"The separatrix operational space of next-step fusion experiments: From ASDEX Upgrade data to SPARC scenarios","cited_arxiv_id":"2407.13539","evidence_quote":"Provides the earlier SepOS projection to next-step devices and SPARC scenarios that this paper's SPARC section builds on."}],"review_version":1}