Strong gravitational lensing provides a natural magnifying effect for the study of the most distant galaxies. While there have been studies on the physical properties of star-forming clumps in strongly lensed galaxies, there is a critical need to automate the process of identifying these clusters, especially in scenarios where high flux density regions are to be discovered in large imaging surveys. Typical methods of clump identification rely on contrast enhancement through image smoothing and subtraction, followed by the use of visual and automatic source detection software. While generally effective, these approaches require careful parameter tuning and manual validation, limiting their efficiency and reproducibility. We present a novel software pipeline titled SUMAC (Software for Uniform Manifold Approximation of Clusters) that automatically processes FITS files of lensed galaxies, reduces the data using Uniform Manifold Approximation and Projection (UMAP), and outputs a topological map clustering together pixels with similar characteristics. Users can specify parameters of interest, including flux, spectral energy distribution, and morphology. We utilize JWST/NIRCam imagery of the z =2.481 lensed galaxy SGAS1110, confirming the functionality of SUMAC by automatically tagging points in the UMAP topological space, mapping them back to the imagery of the lensed galaxy to show alignment with visual star forming clusters. We additionally analyze spectroscopic data for the galaxy, ensuring pixels that SUMAC identifies as corresponding to star-forming clumps match characteristics such as age, metallicity, and emission line ratios that are indicative of star formation. SUMAC’s ability to handle large datasets efficiently, without requiring manual validation or extensive parameter tuning, ensures a more reproducible and scalable approach to high-redshift galactic analysis. SUMAC has the potential to be a valuable tool in the field of astronomical image processing, increasing the efficiency and accuracy of galactic dynamics studies.

Gravitational lensing is a naturally occurring phenomena in which foreground galaxies magnify the light of background galaxies, enabling observations that are otherwise too faint or distant to resolve. With the imaging capabilities of the James Webb Space Telescope (JWST), strongly lensed galaxies are now being spatially resolved to a degree previously unachievable. It is now not only possible but crucial to study lensed galaxies to completely unpack the properties and processes of galaxies in the early universe at these spatial scales. I use spectral energy distribution (SED) fitting and modeling tools on spatially resolved data from JWST. The data includes observations of COOLJ1241+2219, the brightest galaxy at Cosmic Dawn i.e., the first billion years of the Universe, and other high-redshift gravitationally lensed galaxies. These observations allow me to produce maps of key properties within the inner regions of these galaxies, revealing a diversity of star formation rates (SFR), star formation histories (SFH), and other stellar properties at the smallest spatial scales. This analysis is important for understanding how early galaxies evolved and quenched (stopped forming stars) not just as a single entity, but through distinct regions that otherwise cannot be resolved if not for magnification from gravitational lensing. This work is expected to significantly improve the methodologies employed to study galaxies as the sum of their individual parts, as we usher in a new era of larger telescopes in the next decade.