The Stellar Mass and Age Distributions of Star-Forming Clumps at $0.5 < z < 5$ in JWST CANUCS: Implications for Clump Formation and Destruction

Abstract

We investigate the resolved properties of star-forming clumps and their host galaxies at $0.5<z<5$ in the JWST CANUCS fields. We find that the fraction of clumpy galaxies peaks near $z\sim2$ for galaxies with masses of $\log(M_{g,*}/M_\odot)\geq10$, while galaxies with masses of $8.5 \leq \log(M_{g,*}/M_\odot) < 10$ show lower clumpy fractions with little redshift evolution. We identify and measure individual clump masses, finding that the aggregated clump stellar mass function (cSMF) follows a power-law slope of $α= -2$ across all redshift bins, broadly consistent with \textit{in-situ} clump formation. However, when split by galaxy masses, the cSMF is found to be flatter ($α\sim-1.6$) for massive galaxies and steeper ($α\sim-2.3$) for lower mass galaxies, with little redshift evolution in both cases. We explore how different formation mechanisms and disruptive processes affect the shape of the clump mass function. In particular, we find that the cSMF slope is flatter with increasing gas fractions in younger clump populations ($<300$ Myr old), suggesting that higher gas availability leads to more massive clumps forming at the time of formation. Alternatively, many high-redshift galaxies in the sample have disturbed morphologies and simulations show that clumps of \textit{ex-situ} origins can flatten the cSMF slope. We also investigate the evolution of clump populations, where we find the cSMF slope become flatter as clumps evolve and age. We interpret this as an indication of the long-term survivability of massive clumps, with feedback mechanisms preferentially disrupting low-mass clumps. Overall, the galaxy-mass dependent cSMF and age distribution point to a complex history for clumps, involving different and competing mechanisms for their formation and destruction.

Figures (15)

• Figure 1: We selected galaxies based on their integrated SNR within a $0.5^{\prime\prime}$ aperture and stellar mass. Additionally, we used the $UVJ$ color selection to select star-forming galaxies (see text for details). Only ${\sim}20\%$ of all sources with $\texttt{USE\_PHOT = True}$ and $\texttt{FLAG\_POINTSRC = False}$ satisfies our selection criteria. It should also be noted that not all sources falling within each selection region necessarily satisfy all the selection criteria in this figure.
• Figure 2: An illustration showing how clumps are identified with starlet wavelet transform. The rest-frame U image is first decomposed into multiple spatial scales. The top row shows the resulting wavelet scales. A smooth model of the galaxy is then reconstructed by combining only the large-scale components (scales 3 and higher), while setting scales 1 and 2 to zero. The bottom row shows the composite color image, the rest-frame U image, the reconstructed smooth image, and finally the high-contrast residual image obtained by subtracting the smooth model from the original. Clumps are identified using a peak finding algorithm on the high-contrast image (see text) and are marked by red circles with a diameter of $0.2^{\prime\prime}$. The cutouts shown here are $3.6^{\prime\prime}\times3.6^{\prime\prime}$.
• Figure 3: Examples of clumpy galaxies in the CANUCS fields. These galaxies are selected based on their clumpy morphology to illustrate our clump detection algorithm, and are ordered by redshift, with less massive galaxies ($\log(M_{g,*}/M_\odot) \lesssim 10$) shown on the left and more massive galaxies ($\log(M_{g,*}/M_\odot) \gtrsim 10$) on the right. In each panel, we show the composite RGB image, the stellar mass maps, the rest-frame UV map, and the same UV map with detected clumps (shown as the circle markers) and bulges (shown as the star markers).
• Figure 4: The fractional luminosity function of clumps as a function of redshift. In general, we find that the distribution is shifted toward higher $L_\mathrm{clump}/L_\mathrm{gal}$ values at higher redshifts as fainter clumps become harder to detect. The gray regions denote the 50% and 80% completeness limit determined by simulating clumpy galaxies and determining how many clumps are retrieved back.
• Figure 5: The completeness limit as a function of host galaxy mass and redshift bins. In the top panels, we show the completeness as a function of the fractional luminosity of clumps. The bottom panels show the completeness as a function of clump stellar mass. We determine the completeness by injecting clumps into galaxy images and rerunning our SED fitting and clump detection pipelines (see text for details). In general, we find that the clump completeness is similar for both galaxy mass bins.