J-PLUS: Local Universe Photometric Survey

Updated 14 November 2025

J-PLUS is a wide-field photometric survey offering a 12-band system that captures key stellar and nebular features for detailed astrophysical analysis.
It employs innovative calibration integrating Gaia, Pan-STARRS, and machine learning techniques to achieve high-precision photometry and systematic control.
The survey enables comprehensive studies from stellar parameter mapping to cosmological tomography, providing a benchmark for future hyperspectral surveys.

The Javalambre Photometric Local Universe Survey (J-PLUS) is a Northern Hemisphere, wide-field, multi-band photometric survey designed to deliver hyperspectral imaging—quasi-integral-field-like, 12-band optical photometry—across thousands of square degrees. Utilizing a purpose-built 0.83 m telescope (JAST80) and panoramic camera (T80Cam) at the Observatorio Astrofísico de Javalambre, J-PLUS systematically samples key stellar and nebular features through a unique combination of broad, intermediate, and narrow-band filters. The survey enables stellar population studies, multi-dimensional galaxy analysis, high-precision photometric calibration, object classification, chemo-dynamical mapping within the Milky Way, and cosmological tomography, bridging the gap between traditional imaging surveys and narrow-field spectroscopic integral field surveys.

1. Survey Architecture and Instrumentation

J-PLUS employs the JAST80 telescope (0.83 m, f/4.5, Ritchey–Chrétien optics) equipped with a 9,216 × 9,232 pixel e2v CCD (10 μm px⁻¹; 0.55″ px⁻¹), yielding a ∼2 deg² field-of-view per pointing. A hexapod-controlled secondary, wide-band anti-reflection coatings, and dual 7-position filter wheels maintain near-diffraction-limited imaging and stable throughput across the focal plane. The photometric system is composed of 12 filters spanning 3,500–10,000 Å:

Filter	λ₀ (nm)	Δλ (nm)	Class	Key Feature
u	348.5	50.8	Medium	Balmer break
J0378	378.5	16.8	Narrow	[O II] λ3727
J0395	395.0	10.0	Medium	Ca II H+K
J0410	410.0	20.0	Medium	Hδ
J0430	430.0	20.0	Medium	G-band (CH)
g	480.3	140.9	Broad	Continuum (blue)
J0515	515.0	20.0	Medium	Mg b triplet
r	625.4	138.8	Broad	Continuum (green)
J0660	660.0	13.8	Narrow	Hα λ6563
i	766.8	153.5	Broad	Continuum (red)
J0861	861.0	40.0	Medium	Ca II triplet
z	911.4	140.9	Broad	Continuum (far-red)

Filter sequencing and response are optimized for temperature, gravity, metallicity diagnostics, and emission line extraction. With a median V-band seeing of 0.71″, typical 5σ AB depths reach 21–22 mag per band in 3″ apertures (Cenarro et al., 2018, López-Sanjuan et al., 2023).

2. Calibration, Data Processing, and Systematic Control

J-PLUS calibration strategy developed rapidly from initial SDSS-based zero-points to a sophisticated framework integrating Gaia BP/RP low-resolution spectra, Pan-STARRS photometry, stellar-locus regression, and white dwarf loci. The current DR3 pipeline achieves 1% accuracy and 0.7–1.3% internal precision:

Zero-point anchoring: Three filters (g, r, i) are anchored to Pan-STARRS DR1 using color-term transformations.
Homogenization: Synthetic Gaia BP/RP photometry (GaiaXPy) is matched to J-PLUS instrumental magnitudes, with per-pointing, per-CCD spatial corrections.
Color-dependent systematic removal: Empirical “transformation surfaces” correct magnitude/color biases in Gaia BP/RP, and a Bayesian fit to the white dwarf locus provides the absolute color scale.
Red sensitivity: Sky-projected polynomial fits further correct for spatial/metallicity-driven color trends—crucial in blue filters affected by the Galactic metallicity gradient (López-Sanjuan et al., 2021). The inclusion of metallicity-dependent stellar loci reduces calibration biases in u, J0378, and J0395 from >0.06 mag to ≪0.01 mag, essential for unbiased stellar parameter and SED inference.
Data products: Each pointing delivers calibrated images, PSF-homogenized coadds, detailed source catalogs (per-band and forced-photometry, wide morphological suite), star-galaxy-quasar probabilities, and photometric redshifts with odds PDFs (López-Sanjuan et al., 2023, López-Sanjuan et al., 2019).

3. Science Cases: Stellar and Galactic Applications

3.1. Stellar Parameter Mapping

The filter set is engineered to break degeneracies between effective temperature, surface gravity, and chemical composition, validated by artificial neural networks (ANNs) trained on SEGUE and LAMOST labels. Key results from DR1 include σ(T_eff) ≈ 55 K, σ(log g) ≈ 0.15 dex, σ([Fe/H]) ≈ 0.07 dex, and σ([C/Fe]), σ([N/Fe]), σ([Mg/Fe]), σ([Ca/Fe]), σ([α/Fe]) ≈ 0.04–0.08 dex, with two million stars mapped over ≈1,000 deg² (Yang et al., 2021). The combination with Gaia enables precise, spatially widespread chemo-dynamical studies.

3.2. Milky Way Archaeology

J-PLUS × Gaia enables full reconstruction of the Milky Way disc’s star formation and chemical enrichment history through Bayesian multi-isochrone fitting across >1 million stars (Alzate-Trujillo et al., 10 Nov 2025). This approach utilizes the 12-band photometry to empirically constrain metallicity distributions (from faint main sequence stars) and combine these as strong priors for precise age dating (from the upper main sequence and subgiant branch). The derived SFH demonstrates clear bifurcation into α-enhanced (thick disc, 12.5–8 Gyr) and solar-scaled (thin disc, 8–3 Gyr) sequences, with resolved spatial trends in metallicity and age. J-PLUS photometry, by offering independent metallicity estimates via narrow-band indices, alleviates the age–metallicity degeneracy endemic to broad-band CMD analyses.

3.3. Stellar Populations and Emission Lines in Galaxies

Multi-band SED fitting (e.g., MUFFIT, AlStar) enables 2D stellar population mapping and nebular emission measurements across galaxies and groups. For instance, in nearby systems such as M101 and satellites, 12-band photometry, when processed through Voronoi tessellation and spectral-synthesis codes, yields spatially resolved maps of stellar mass surface density (Σ*), light-weighted age, mass-weighted metallicity, dust attenuation, SFR surface density, emission line fluxes, and equivalent widths (Thainá-Batista et al., 19 May 2025). Local relations—age-Σ, resolved main sequence, Σ_-metallicity, and Σ_*-O/H—can be robustly constructed, rivaling IFU survey results but at increased areal throughput.

3.4. Extragalactic Redshift and Classification Science

Using twelve-band photometry, J-PLUS achieves photometric redshift precision of Δz ≃ 0.01–0.03 for galaxies with r ≲ 21 mag, and Δz < 0.01 at r < 18 (Cenarro et al., 2018). The inclusion of intermediate and narrow bands improves photo-z scatter by up to a factor 2.5 over ugriz-only surveys. Angular density and redshift fluctuation tomography in DR3 constrains linear galaxy bias evolution ((b_g(z=0.05)=0.9±0.06 to b_g(z=0.15)=1.5±0.05)), provides the first empirical photo-z error measurement from angular clustering (σ_Err ≈ 0.014), and sets methodological groundwork for future 3D velocity and growth rate inference (Hernández-Monteagudo et al., 19 Dec 2024).

4. Machine Learning and Classification Pipelines

J-PLUS leverages large, spectroscopically labelled datasets and multi-modal input vectors to drive state-of-the-art object classification:

Star/galaxy/quasar classification: BANNJOS (Bayesian Neural Network Joint Optical Surveyor) produces probabilistic, PDF-aware classifications of all DR3 sources. Trained on >1.2 million spectroscopically classified sources, it achieves >95% accuracy at r < 21.5 mag—and ~90% at r < 22 mag (Pino et al., 25 Apr 2024). PDF-based selection allows dynamic purity/completeness tuning and identification of ambiguous objects (e.g. AGN/quasar/host blends).
Pipeline optimization: TPOT-driven AutoML confirms the superior AUC and average precision of XGBoost for 3-way (star/galaxy/quasar) classification, robustly outperforming morphological classifiers such as SExtractor’s CLASS_STAR. All features—12-band photometry, colors, morphology, extinction, PSF—are used (Marttens et al., 2022).
Specialized populations: For UCDs, a two-layer ML pipeline (PCA, SVM) reduces photometric false positives and efficiently pre-filters for SED fitting. Recall in test/blind regions reaches 91–92%, yielding a 7,827-object candidate sample over 2,176 deg² (Mas-Buitrago et al., 2022).

5. Survey Operations, Automation, and Data Releases

The J-PLUS Tracking Tool underpins survey execution and data quality assurance. It integrates:

Real-time scheduler: Assigns pointings nightly based on visibility, scientific priority, and lunar avoidance, using a weighted composite score S_i including airmass, moon distance, and priority (Civera, 2022).
Adaptive exposure control: Scales filter exposure times dynamically using current moon phase and separation to maintain uniform depth (RMS < 0.02 mag across the footprint).
Feedback system: Automated web-services assess raw and reduced image diagnostics, filter completion, and mark pointings for reobservation if <100% filter validation is met. The typical time from exposure to final VALIDATED label is <24 hours, achieving >99% completion in first attempt and robust audit logging.
Data releases: DR3 covers 3,192 deg² (2,881 after masking), containing 44+ million sources, with fully calibrated, PSF-homogenized images, object classification PDFs, and forced-photometry catalogs.

6. Scientific Legacy and Future Prospects

J-PLUS’s design as both an independent scientific program and a pathfinder for J-PAS (which will use 56 narrow bands over the same footprint) establishes its legacy as the earliest true “hyperspectral” local universe survey. Its 12-band system:

Enables cost-effective chemo-dynamical and resolved mapping that complements expensive IFU spectroscopy;
Provides a foundational calibration base for J-PAS (and S-PLUS in the south), with a shared calibration hierarchy (Gaia photometry → white dwarf locus → Pan-STARRS absolute flux).
Serves as a benchmark for machine learning-driven object and population studies, variably enabling “photometric spectroscopy” of individual stars, asteroids, planetary nebulae, emission-line galaxies, and more.
Demonstrates the key transformative potential of statistical multi-band photometry for massive, multi-disciplinary sky survey projects.

As J-PLUS approaches its 8,500 deg² goal and as J-PAS comes online, the homogeneous legacy dataset delivered by J-PLUS will continue to underpin wide-field stellar, galactic, extragalactic and cosmological studies, offering a photometrically precise reference for the astronomical community and future spectro-photometric surveys (Cenarro et al., 2018, Alzate-Trujillo et al., 10 Nov 2025, López-Sanjuan et al., 2023, Thainá-Batista et al., 19 May 2025, Yang et al., 2021, Pino et al., 25 Apr 2024, Hernández-Monteagudo et al., 19 Dec 2024).