NGDEEP: A New Non-Parametric Measure of Local Star-Formation and Attenuation at Cosmic Noon

TL;DR

This work introduces a non-parametric, pixel-level method to quantify how local dust attenuation shapes the spatial distribution of star formation in galaxies at cosmic noon. Using JWST/NIRISS slitless spectroscopy from NGDEEP, the authors map Balmer decrements to derive per-pixel attenuation and dust-corrected SFRs across 14 galaxies at $1.0<z<2.5$, then compute the cumulative SFR as a function of attenuation, defining $A^{SFR}_{10%}$, $A^{SFR}_{50%}$, and $A^{SFR}_{90%}$. They find that the median attenuation behind which 50% of the SFR lies is $A^{SFR}_{50%} \approx 3.41$ mag, while $A^{SFR}_{10%} \approx 1.45$ mag, indicating most star formation is heavily dust-obscured and that global attenuation underestimates typical region obscuration. The results underscore the limitations of spatially integrated attenuation metrics and provide a novel framework for comparing observations with radiative-transfer models of dust geometry, with significant implications for understanding star formation during the peak epoch of cosmic activity. The approach, validated on JWST data, sets the stage for larger, more diverse samples across redshift to refine our picture of dust-rich star formation in the universe.

Abstract

We introduce a new non-parametric technique to quantify the spatially-resolved relationship between the local star-formation rate (SFR) and dust attenuation. We then apply it to 14 star-forming galaxies at $1.0<z<2.5$ using JWST/NIRISS slitless spectroscopy from the NGDEEP survey. First, we construct spatially resolved ($\sim$1~kpc per pixel) Balmer decrement ($Hα/Hβ$) maps of these galaxies and derive their corresponding dust attenuation and intrinsic SFR maps. We then rank-order the map pixels by attenuation and construct a cumulative distribution curve of the total SFR as a function of increasing attenuation. We define $\mathrm{A}^{\mathrm{SFR}}_{10\%}$, $\mathrm{A}^{\mathrm{SFR}}_{50\%}$, and $\mathrm{A}^{\mathrm{SFR}}_{90\%}$ as the dust attenuation levels behind which 10\%, 50\%, and 90\% of the total integrated SFR is screened, respectively. These metrics quantify the probability that a given star-forming region lies behind a given level of attenuation. Across the full sample, 50\% of the local star formation occurs behind an attenuation of 3.41 mag or higher ($\mathrm{A}^{\mathrm{SFR}}_{50\%}$). This indicates that the bulk of star formation in these galaxies is significantly attenuated by dust. The value of $\mathrm{A}^{\mathrm{SFR}}_{10\%}$ equals 1.45 for the average profile, indicating that even the least attenuated star-forming regions are still highly attenuated. The globally measured attenuation more closely matches $\mathrm{A}^{\mathrm{SFR}}_{10\%}$ than $\mathrm{A}^{\mathrm{SFR}}_{50\%}$. This suggests that the global value is weighted toward the least dust-obscured star-forming regions and significantly underestimates the typical attenuation a star-forming region encounters. Our results demonstrate a new approach for understanding the extremely dusty local conditions of the star-forming interstellar medium in SF galaxies at cosmic noon.

Figures (5)

• Figure 1: The maps derived for an example galaxy, ngdeep_00503, are shown. The top row includes the H$\alpha$ flux map (left), H$\beta$ flux map (middle), and the segmented Balmer decrement ($\hbox{{\rm H}$\alpha$}/\hbox{{\rm H}$\beta$}$) map (right). The bottom row shows the uncertainty of the Balmer decrement map (left) and the Balmer decrement radial profile (right). Each point of the radial profile corresponds to an individual pixel. The orange dashed line marks the intrinsic Balmer decrement expected for case-B recombination.
• Figure 2: The H$\alpha$-based SFR versus the SED-derived SFR for the galaxy sample is shown. Galaxies are color-coded black and red if they fall above or below the blue one-to-one line, respectively.
• Figure 3: The dust corrected $\hbox{{\rm H}$\alpha$}$-based SFR versus SED-calculated stellar mass is shown for the galaxies in our sample. The solid lines denote the star formation mass sequence lines at the redshifts of the galaxies in our sample Whitaker_2014. The red galaxies and lines represent $2.0<z<2.5$. The yellow galaxies and lines represent $1.5<z<2.0$. The blue galaxies and lines represent $1.0<z<1.5$. The dashed lighter lines bound $\pm0.3$ dex from their respective relation.
• Figure 4: The cumulative distribution of the dust corrected star formation rate as a function of the observed $\hbox{{\rm H}$\alpha$}/\hbox{{\rm H}$\beta$}$ emission line ratio is shown. The top axis indicates the corresponding attenuation at H$\alpha$ (in magnitudes). Conceptually, this approach quantifies the fraction of star formation at or below a given level of attenuation. The distributions for individual galaxies are shown as dashed and solid colored lines. The average distribution is shown with the bold black line. The nine galaxies below the one-to-one line in Figure \ref{['fig:fig2']} are shown by solid lines while the other are shown as dashed lines.
• Figure 5: The left panel shows the global $A(\hbox{{\rm H}$\alpha$})$ versus the quantities $\mathrm{A}^{\mathrm{SFR}}_{10\%}$ (blue circle), $\mathrm{A}^{\mathrm{SFR}}_{50\%}$ (orange square), and $\mathrm{A}^{\mathrm{SFR}}_{90\%}$ (green triangle). The stars with black outlines represent the corresponding values for the average profile (see Figure \ref{['fig:fig4']}). The right panel shows the cumulative SFR fraction corresponding to the attenuation measured globally. In short, this distribution indicates the fraction of new stars forming at or below the global attenuation level. The red dashed line indicates the SFR fraction associated with the global $A(\hbox{{\rm H}$\alpha$})$ measured for the average profile.