Integral Field Unit Spectroscopy with One Fiber

2026-06-08

AI Summary

Traditional integral field unit (IFU) spectroscopic observations are prohibitively expensive, hindering large-scale studies of galaxy evolution. This work proposes a multimodal probabilistic foundation model capable of reconstructing high-resolution spectra—and their associated uncertainties—at arbitrary spatial locations within galaxies using only single-fiber spectra and broadband imaging, without any IFU data during training. The method innovatively integrates masked autoencoding, multimodal alignment, and spatial position– and redshift-aware wavelength encoding, achieving IFU-like capabilities through large-scale pretraining. The generated emission-line flux maps show excellent agreement with independent MaNGA IFU observations and match the performance of supervised baseline models trained directly on IFU data.

Abstract

Integral field unit (IFU) spectroscopy provides spatially resolved spectra across galaxies, offering crucial insights into their evolution. However, its high observational cost limits current IFU datasets to $\sim 10^4$ objects. We present a multi-modal, probabilistic foundation model that predicts high-resolution spectra with calibrated uncertainties at arbitrary spatial locations within a galaxy directly from broadband images. Built on a masked autoencoder framework, our architecture injects fiber positional encodings and redshift aware wavelength encodings, enabling spatially conditioned predictions. Trained on 4.7 million images and single fiber spectroscopic observations from the Dark Energy Spectroscopic Instrument (DESI) survey, our model exploits the natural variance of fiber placements and the morphological self-similarity of galaxies to achieve IFU-like capabilities without any IFU training data. Predicted emission line flux maps match independent IFU observations from the Mapping Nearby Galaxies at APO (MaNGA) survey, with performance comparable to a supervised baseline trained directly on IFU data.

Problem

• Integral Field Unit Spectroscopy
• galaxy evolution
• spatially resolved spectra
• broadband imaging
• spectroscopic prediction

Innovation

• Integral Field Unit Spectroscopy
• Foundation Model
• Masked Autoencoder
• Spatially Resolved Spectra
• Probabilistic Prediction