A&A, 684, A104 (2024) https://doi.org/10.1051/0004-6361/202349029 c© The Authors 2024 Astronomy &Astrophysics Supernova environments in J-PLUS Normalized cumulative-rank distributions and stellar-population synthesis combining narrow- and broad-band filters? Raúl González-Díaz1,2 , Lluís Galbany1,3 , Tuomas Kangas4,5 , Rubén García-Benito6 , Joseph P. Anderson7,8 , Joseph Lyman9, Jesús Varela10, Lamberto Oltra1 , Rafael Logroño García10, Gonzalo Vilella Rojo10, Carlos López-Sanjuan10 , Miguel Ángel Pérez-Torres6,15,16 , Fabián Rosales-Ortega2 , Seppo Mattila4,5, Hanindyo Kuncarayakti4,5, Phil James11, Stacey Habergham11 , José Manuel Vílchez6 , Jailson Alcaniz12 , Raul E. Angulo13 , Javier Cenarro10 , David Cristóbal-Hornillos10, Renato Dupke12, Alessandro Ederoclite10, Carlos Hernández-Monteagudo10 , Antonio Marín-Franch10 , Mariano Moles10, Laerte Sodré Jr.14 , and Héctor Vázquez Ramió10 1 Institute of Space Sciences (ICE-CSIC), Campus UAB, Carrer de Can Magrans, s/n, 08193 Barcelona, Spain e-mail: raul.gonzalezD@autonoma.cat 2 Instituto Nacional de Astrofísica, Óptica y Electrónica (INAOE-CONAHCyT), Luis E. Erro 1, 72840 Tonantzintla, Mexico 3 Institut d’Estudis Espacials de Catalunya (IEEC), 08034 Barcelona, Spain 4 Finnish Centre for Astronomy with ESO (FINCA), 20014 University of Turku, Finland 5 Tuorla Observatory, Department of Physics and Astronomy, 20014 University of Turku, Finland 6 Instituto de Astrofísica de Andalucía (CSIC), Glorieta de la Astronomía s/n, Aptdo. 3004, 18080 Granada, Spain 7 European Southern Observatory, Alonso de Córdova 3107, Casilla 19, Santiago, Chile 8 Millennium Institute of Astrophysics MAS, Nuncio Monsenor Sotero Sanz 100, Off. 104, Providencia, Santiago, Chile 9 Department of Physics, University of Warwick, Coventry CV4 7AL, UK 10 Centro de Estudios de Física del Cosmos de Aragón (CEFCA), Unidad Asociada al CSIC, P. San Juan, 1, 44001 Teruel, Spain 11 Astrophysics Research Institute, Liverpool John Moores University, IC2, 146 Brownlow Hill, Liverpool L3 5RF, UK 12 Observatorio Nacional, Rua Gal. José Cristino 77, Rio de Janeiro 20921-400, RJ, Brazil 13 Donostia International Physics Center, Paseo Manuel de Lardizabal 4, 20018 Donostia-San Sebastian, Spain 14 Instituto de Astronomia, Geofísica e Ciencias Atmosfêricas, U. São Paulo, R. do Matão 1226, São Paulo 05508-090, SP, Brazil 15 Center for Astroparticles and High Energy Physics (CAPA), Universidad de Zaragoza, 50009 Zaragoza, Spain 16 School of Sciences, European University Cyprus, Diogenes street, Engomi 1516 Nicosia, Cyprus Received 19 December 2023 / Accepted 22 January 2024 ABSTRACT We investigated the local environmental properties of 418 supernovae (SNe) of all types using data from the Javalambre Photomet- ric Local Universe Survey (J-PLUS), which includes five broad-band and seven narrow-band imaging filters. Our study involves two independent analyses: (1) the normalized cumulative-rank (NCR) method, which utilizes all 12 single bands along with five continuum-subtracted narrow-band emission and absorption bands, and (2) simple stellar population (SSP) synthesis, where we build spectral energy distributions (SED) of the surrounding 1 kpc2 SN environment using the 12 broad- and narrow-band filters. Improve- ments on previous works include: (i) the extension of the NCR technique to other filters (broad and narrow) and the use a set of homogeneous data (same telescope and instruments); (ii) a correction for extinction to all bands based on the relation between the g − i color and the color excess E(B − V); and (iii) a correction for the contamination of the [N ii] λ6583 line that falls within the Hα filter. All NCR distributions in the broad-band filters, tracing the overall light distribution in each galaxy, are similar to each other. The main difference is that type Ia, II, and IIb SNe are preferably located in redder environments than the other SN types. The radial distribution of the SNe shows that type IIb SNe seem to have a preference for occurring in the inner regions of galaxies, whereas other types of SNe occur throughout the galaxies without a distinct preference for a specific location. For the Hα filter we recover the sequence from SNe Ic, which has the highest NCR, to SNe Ia, which has the lowest; this is interpreted as a sequence in progenitor mass and age. All core-collapse SN types are strongly correlated to the [O ii] emission, which also traces star formation rate (SFR), following the same sequence as in Hα. The NCR distributions of the Ca II triplet show a clear division between II-IIb-Ia and Ib-Ic-IIn subtypes, which is interpreted as a difference in the environmental metallicity. Regarding the SSP synthesis, we found that including the seven J-PLUS narrow filters in the fitting process has a more significant effect on the core-collapse SN environmental parameters than for SNe Ia, shifting their values toward more extincted, younger, and more star-forming environments, due to the presence of strong emission lines and stellar absorptions in those narrow bands. Key words. methods: observational – methods: statistical – techniques: photometric – supernovae: general – galaxies: general – galaxies: photometry ? Full Tables A.1–A.5 are available at the CDS via anonymous ftp to cdsarc.cds.unistra.fr (130.79.128.5) or via https://cdsarc.cds.unistra.fr/viz-bin/cat/J/A+A/684/A104 Open Access article, published by EDP Sciences, under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. This article is published in open access under the Subscribe to Open model. Subscribe to A&A to support open access publication. Article number, page 1 of 17 González-Díaz, R., et al.: A&A, 684, A104 (2024) 1. Introduction Supernovae (SNe) are one of the final stages in stellar evolu- tion and are key in driving the chemical evolution of galax- ies. Classical SNe (excluding superluminous ones) are basically divided in two main types: thermonuclear (those of type Ia) and core-collapse (CC). Type Ia SNe are those triggered by the ther- monuclear explosion of a carbon and oxygen white dwarf (WD; Hoyle & Fowler 1960). Different progenitor scenarios, such as WDs reaching the Chandrasekhar limit (MCh ∼ 1.44 M ) by mass accretion in a binary system or mergers of WDs, and explo- sion mechanisms, such as internal detonations of MCh, WD, or surface explosions on WDs with masses lower than MCh, have been proposed to explain how the WDs explode (see reviews by, e.g., Maoz et al. 2014). No direct progenitor detection has been reported to date (however, see McCully et al. 2014), but in all cases, the main spectral features are the lack of H lines and the presence of strong Si II lines up to a couple of weeks after maximum brightness (Filippenko 1997). SNe Ia occur in galaxies of all types, including elliptical galaxies which only contain old stellar populations, providing strong evidence that SNe Ia have long-lived, low-mass progenitors (Han et al. 2010). SNe Ia are more luminous than CC SNe, comprising about 30% of the observed SNe (Graur et al. 2017). Their distinctive fea- ture lies in the fact that generally normal SNe Ia exhibit similar spectra and light curves, making them valuable for tracing cos- mological distances (e.g., Phillips et al. 1999; Riess et al. 1996; Perlmutter et al. 1997, 1999; Astier et al. 2006). Core collapse SNe are those resulting from the gravitational collapse of the iron core of a massive star (>8 M ; Arnett et al. 1989). CC SNe are divided into three main subtypes depending on their spectral features, which reflect the state of the outer lay- ers of the progenitor star at the moment of explosion. Type II SNe show H lines because their progenitors have kept the H-rich outer envelope intact, type Ib SNe show He but no H since the progenitor has lost the H envelope, and type Ic SNe lack both H and He lines due to the loss of both H- and He-rich layers (Filippenko 1997). Additionally, SNe IIn show narrow H emis- sion lines most probably due to interaction with circumstellar material (Smith et al. 2011), and SNe IIb are those intermedi- ate between SNe II and Ib, that show H only for a few days after explosion indicating that there was still a thin H layer before explosion (Filippenko 1988; Nomoto et al. 1993). There is a debate on the process responsible for the mass-loss lead- ing to stripped envelope (SE) SN types (Ib, Ic, IIb). Mass-loss through radiation-driven wind from single hot massive stars or the removal of the outer envelope via tidal stripping by a less massive companion in a binary are the two most viable possibil- ities (Filippenko 1997; Gal-Yam et al. 2017; Prentice & Mazzali 2017; Shivvers et al. 2017; Taddia et al. 2018). One outstanding question in the field of massive stellar evo- lution concerns the link between the nature of the progenitor and the resultant SN. Direct constraints on the nature of SN progenitors are limited. The bulk of these come from the small number of nearby events where deep, high-resolution (usually from the Hubble Space Telescope) pre-explosion imaging exists to directly image the progenitor star (e.g., Maund et al. 2013; Smartt 2015; Van Dyk 2017). For almost all discovered SNe, this is unfeasible. As such, methods investigating statistical prop- erties of SNe have been developed to exploit the large num- bers of observed SN environments (pre-SN or once the SN has faded), including historical ones. One of those statistical meth- ods is the normalised cumulative-rank (NCR; Fruchter et al. 2006; James & Anderson 2006) method. The NCR method, when applied to Hα narrow imaging, measures the correla- tion between SN locations and star-formation intensity (i.e., Hα emission) in their host galaxies. The strength of this correla- tion is an indicator of the progenitor lifetime, a proxy for the initial mass of the progenitor. The method has been employed on samples of SNe to deduce a sequence of ascending mass for the common SN types Ia, II, Ib, and Ic (Anderson & James 2008; Anderson et al. 2012; Kangas et al. 2013). Subsequently, Kangas et al. (2017) established a connection between the NCRs of these SNe and those of massive stars in the LMC and M33. This link indicated initial masses of approximately >20 M for Ic SNe progenitors and >9 M for II and Ib SNe. For the stripped-envelope SN types, Ic and Ib, these findings would imply that single Wolf-Rayet stars and/or massive binaries are the most probable progenitors for the former, while for the lat- ter interacting relatively low-mass binary systems, such as the directly detected 10−12 M progenitor of the SN Ib iPTF13bvn (Eldridge & Maund 2016), would be dominant. Another method that has proved useful in putting constraints on SN progenitors is the study of the stellar populations at SN locations. Galbany et al. (2014, 2016) presented the first statis- tical study of nearby SN host galaxies using integral field spec- troscopy. These were provided by the CALIFA survey, which consists of 132 SNe of all types in 115 galaxies, and it was later extended to 272 SNe in 232 galaxies in Galbany et al. (2018). Stellar parameters were inferred by fitting a set of single stellar population (SSP) models to the spectra of the locations where SNe occurred. They found that CC SNe tend to explode at posi- tions with younger stellar populations than the galaxy average, while at Ia SN locations local properties were, on average, the same as global ones. They also found a sequence from higher to lower metallicity, from SN Ia to SN Ic-BL, and a significant increasing relative number of SNe Ic at higher metallicities com- pared to other CC SNe types, which supports a large fraction of this SN subtype occurring under the single-star scenario (see also Kuncarayakti et al. 2013, 2018). In this paper, we present a comprehensive analysis of 418 SNe environments using the 12 filter system of the Javalambre Photometric Local Universe Survey (J-PLUS1; Cenarro et al. 2019), whose spatial resolution (. 382 pc, corresponding to an angular resolution of 1.14 arcsec at z = 0.016) and overall data quality of this dataset allow us to study the SN environments. Our investigation involves two main methodologies: the normal- ized cumulative-rank (NCR) analysis and the simple stellar pop- ulation (SSP) synthesis. Compared to previous works in the liter- ature, our study offers several advancements. Firstly, we extend the NCR technique to a number of broad- and narrow-band fil- ters other than Hα, which enables us to explore correlations with new spectral lines. This extension opens up possibilities to probe additional environmental factors, such as stellar popula- tion metallicity or age. Previous studies, such as Kangas et al. (2013, with UV imaging) and Anderson et al. (2015, with NIR imaging), have explored these factors to some extent. However, our research operates on a much larger scale and utilizes a set of homogeneous data acquired from the same telescope and instru- ment. Another improvement over past works is the correction for the contamination from the [N ii] λ6583 line that falls within the Hα filter. Vilella-Rojo et al. (2015) presented a method to cor- rect for such extra emission by combining narrow- and broad- band data and providing pure Hα emission from J-PLUS obser- vations. This method can also be extended to other narrow-band filters. 1 www.j-plus.es Article number, page 2 of 17 González-Díaz, R., et al.: A&A, 684, A104 (2024) 2. Data 2.1. J-PLUS The Javalambre Photometric Local Universe Survey is designed to observe 8500 deg2 of the northern sky from the Observato- rio Astrofísico de Javalambre (OAJ, Teruel, Spain; Cenarro et al. 2014) with the 83 cm Javalambre Auxiliary Survey Telescope (JAST80) and T80Cam, a panoramic camera of 9.2k × 9.2k pix- els that provides a 2 deg2 field of view (FoV) with a pixel scale of 0.55 arsec pix−1 (Marín-Franch et al. 2015). The J-PLUS filter system comprises the 12 broad, intermediate and narrow-band optical filters. J-PLUS is particularly designed to carry out the photometric calibration of the Javalambre Physics of the Accel- erating Universe Astrophysical Survey (J-PAS2; Benitez et al. 2014), which has observed the northern sky from Javalambre using 59 optical narrow-band filters. For this reason, some J- PLUS filters are located at key stellar spectral features that allow us to retrieve very accurate spectral energy distributions (SED) for more than five millions of stars in our galaxy. The 12 filters of J-PLUS are listed in Table 1 and their transmission is shown in Fig. 1, together with two template spectra of a star-forming and a passive galaxies to highlight the regions of interest covered by the intermediate- and narrow-band filters. The third J-PLUS Data Release (DR33) was made public in December 2022, com- prising 1642 pointings observed in the 12 optical bands amount- ing to ∼3200 sq. deg with around 30 million sources detected at magnitudes r < 21. We cross-matched all SN coordinates from the Open Supernova Catalogue (OSC; Guillochon et al. 2017) to the central coordinates of all J-PLUS DR3 tiles, and found 2168 SNe positions within the 2 sq. deg of the J-PLUS tile. The width of the narrow Hα filter (J0660 filter; ∼150 Å) puts an upper limit on the redshift of galaxies that are useful for an NCR study at z ∼ 0.0163 (or about 60 Mpc), since the J0660 transmission falls down dramatically at 6672 Å. Thus, in order to keep the Hα emission of the SN host galaxy within the coverage of the J-PLUS J0660 filter, we excluded all objects with a redshift larger than 0.0163. Moreover, NCR becomes less useful with distance, and samples at significantly different median distances may not be directly comparable (Kangas et al. 2017). This red- shift cut is passed by 282 SNe of the following types: 88 SNe Ia, 126 SNe II, 7 SNe IIn, 22 SNe Ib, 17 SNe Ic, 17 SNe IIb, and another five type Ibc SNe. 2.2. A dedicated SN host galaxy program at JAST80 To reach a significantly large number of SNe that cover all types compared to previous studies, and within the redshift limit imposed by the J-PLUS Hα filter width, we had to include objects outside the J-PLUS footprint, taking special care to increase the number of less represented subtypes (Ic, Ib, IIb, IIn). With this objective in mind, we initiated a dedicated program to acquire 73 additional fields using the same instrumental configu- ration and exposure times as in J-PLUS, which involves T80 and 12 filters. These fields encompassed the positions and host galax- ies of 136 SNe. For each of the 12 filters, we obtained three indi- vidual frames, all of which were subsequently processed using the same J-PLUS pipeline (Cenarro et al. 2019). Summing up the observations from the main survey and our dedicated program, 418 SNe are included in our sample, and they are listed in Table A.1. Taking into account the main SN types, we have 156 SNe II, 20 SN IIn, 108 SN Ia, 51 SN Ib, 49 SN Ic, 2 http://www.j-pas.org/survey 3 http://www.j-plus.es/datareleases/data_release_dr3 Table 1. J-PLUS photometric system properties. Filter name Central λ (nm) FWHM (nm) Comments uJAVA 348.5 50.8 ultraviolet continuum J0378 378.5 16.8 [O ii] emission J0395 395.0 10.0 Ca H+K absorption J0410 410.0 20.0 Hδ emission J0430 430.0 20.0 G-band absorption gSDSS 480.3 140.9 SDSS green continuum J0515 515.0 20.0 Mgb Triplet absorption rSDSS 625.4 138.8 SDSS red continuum J0660 660.0 14.5 Hα emission iSDSS 766.8 153.5 SDSS near infrared continuum J0861 861.0 40.0 Ca II triplet absorption zSDSS 911.4 140.9 SDSS near infrared continuum Notes. Transmission curves can be found at the J-PLUS website: http: //www.j-plus.es/survey/instrumentation. Fig. 1. Transmission of 12 J-PLUS filters, on top of a typical star- forming (in gray) and passive galaxy (in black) spectra. Labels, which represent the filter names, highlight the main spectral features covered by the intermediate and narrow band filters. 27 SN IIb, and seven other type Ibc SNe. The analysis is car- ried out with the sample divided into these seven main SN types, where Ibc SNe include Ib, Ic, IIb, and other Ibcs. 3. Generation of individual 3D data cubes Starting from the initial 12 images per field, one for each filter, we performed a number of steps to produce one 3D data cube per galaxy, namely galaxy cutout, flux calibration, correction for dust extinction, and continuum subtraction. All these steps are described in the following subsections. 3.1. Galaxy cutouts J-PLUS images cover ∼2 sq. deg of the sky with a pixel scale of 0.55 arcsec, with a median seeing of 1.14 arcsec, correspond- ing to .382 pc at z = 0.016. After registering all 12 images to the same reference frame, we took a squared cutout of a size that includes the full extent of the SN host galaxy centered at its core, defining the size by visual inspection (see Fig. 2). The 12 cutouts are then stored in a 3D cube, mimicking integral field spectroscopy observations, where two spatial dimensions corre- spond to RA and Dec, and the third dimension provides a spec- tral energy distribution (SED) at any position within the cutout. Article number, page 3 of 17 González-Díaz, R., et al.: A&A, 684, A104 (2024) 10h18m 16m 14m 12m 42°00' 41°40' 20' 00' Right Ascension De cli na tio n 0 200 400 600 800 1000 1200 0 200 400 600 800 1000 1200 SN 1921B Fig. 2. Example of J-PLUS field frame. The zoomed-in image corre- sponds to the cutout of the galaxy NGC 3184, which contains the SN 1921B; marked in white. The images were constructed combining the uJAVA (blue), gSDSS (green), and rSDSS (red) images as false colors. 3.2. Flux calibration To convert the electronic counts stored in each pixel to radiative flux, we use the expression from Logroño-García et al. (2019): Fλ = C · 10−0.4(ZP+48.6) c λ2pivot , (1) where C is the number count in the pixel, c is the speed of light, ZP is the zero point of the band used for the calibration to the standard AB magnitude system, and λpivot is the pivot wave- length of the filter, which is a source-independent measurement of the characteristic wavelength of a given pass band by λ2pivot = ∫ T (λ)dλ∫ T (λ)λ−2dλ , (2) where T (λ) represents the transmission curve of the filter. The values of ZP and pivot wavelength are given in Table 2. 3.3. Dust extinction correction The Milky Way interstellar dust reddening was corrected using the python code DUSTMAPS (Green 2018) and the Gaia Total Galactic Extinction map (Delchambre et al. 2023) as the source to obtain the E(B – V) color excess for every galaxy. In addi- tion, we applied a correction to the reddening due to dust in the SN host galaxy. Following Calzetti et al. (2000), the difference between the intrinsic and observed flux due to extinction is given by Fi(λ) = F0(λ)100.4E(B−V)k ′(λ), (3) where Fi(λ) and F0(λ) are the intrinsic and observed fluxes at a wavelength λ, E(B − V) is the color excess, and k′(λ) is the extinction law: an empirical relationship between the amount of extinction and wavelength. We employed the Calzetti et al. (2000) extinction law, defined as k′(λ) =  2.659(−1.857 + 1.040/λ) + R′V for 0.63 µm ≤ λ ≤ 2.20 µm 2.659(−2.156 + 1.509/λ − 0.198/λ2 + 0.011/λ3) + R′V for 0.12 µm ≤ λ ≤ 0.63 µm, , (4) where R′V is the obscuration in the V band. Here, we used the same value found by Calzetti et al. (2000), R′V = 4.05± 0.80, for stellar extinction4. To estimate the color excess E(B−V), we fol- lowed Vilella-Rojo et al. (2015), who found a relation between the observed g′ − i′ color and the spectroscopically measured E(B − V) obtained after convolving the SDSS spectra with the J-PLUS photometric system. These authors obtained the follow- ing expression by fitting a power-law function: E(B − V) = 0.206(g′ − i′)1.68 − 0.0457, (5) and assuming E(B − V) = 0 for g′ − i′ < 0.4. This is an empirical expression for the gas extinction. Vilella-Rojo et al. (2015) mixed in the same process of correc- tion to the stellar and gas components for all the J-PLUS data; so, we used the same method to be coherent. The gas extinc- tion is, at minimum, the same as the stellar extinction. However, on average, the extinction in HII regions can be twice as high (Cid et al. 2005). Therefore, it is essential to note that the extinc- tion correction applied by this method represents a conservative lower bound. With all these ingredients, we corrected the fluxes of the 12 images for all SNe in our sample, pixel by pixel. This improve- ment results in a change in our NCR values compared to studies that do not apply this correction to the fluxes. Hence, we caution against making direct comparisons of NCR values to previous works that employed different methodologies. 3.4. J0660 continuum subtraction J-PLUS narrow-band filters were designed to cover wavelength regions of a number of emission lines of interest. However, they also include the flux of the underlying stellar continuum that needs to measure the gas-phase component exclusively. In par- ticular, both the rSDSS and J0660 filters are collecting the flux of the Hα emission line, so to study the Hα emission, which is a tracer of the star formation, we have to subtract the stellar con- tinuum from the J0660 flux. Vilella-Rojo et al. (2015) presented two methods to remove the underlying continuum from a narrow-band filter. The sim- plest is by combining the narrow band that contains the emission with an adjacent or overlapping broad filter that traces the con- tinuum. For Hα, we can use the J0660 and the rSDSS filters, but taking into account that J0660 also includes the contribution of the forbidden transition of [N ii] doublet (at 654.8 nm and 658.53 nm). Assuming a flat continuum, the flux of the three emission lines can be recovered using the following expression: FHα+[N ii] = ∆J0600 F J0600 − FrSDSS 1 − ∆J0600 ∆rSDSS , (6) where F J0600 and FrSDSS correspond to the average flux inside the J0660 and rSDSS filters, respectively, and ∆x is defined for 4 The stellar extinction mainly affects those filters without emission lines, and gas extinction mainly affects those with emission lines (Thomas et al. 2013). Article number, page 4 of 17 González-Díaz, R., et al.: A&A, 684, A104 (2024) Table 2. Pivot wavelengths and zero points of the J-PLUS bands. Filter uJAVA J0378 J0395 J0410 J0430 gSDSS λpivot (nm) 352.29 378.64 395.06 410.07 430.04 474.46 ZP (AB mag) 21.078 (049) 20.416 (036) 20.332 (037) 21.283 (029) 21.380 (028) 23.553 (035) Filter J0515 rSDSS J0660 iSDSS J0861 zSDSS λpivot (nm) 514.98 622.98 659.98 767.65 860.25 892.20 ZP (AB mag) 21.500 (040) 23.494 (023) 20.953 (023) 23.136 (023) 21.439 (007) 22.498 (007) any pass band x at any wavelength of interest λs as ∆x ≡ ∫ Px(λ)λdλ Px(λ = λs)λs , (7) where Px is the transmission of the pass band x as a function of wavelength. In our case, we took λs = λHα. This method assumes a linear continuum, so the Hα absorp- tion is not taken into account, resulting in an overall bias of approximately 9% in all the results (Vilella-Rojo et al. 2015). When including more filters, such as those available in J-PAS, this method of continuum subtraction improves significantly (Martínez-Solaeche et al. 2021). 3.5. [N ii] removal The continuum-subtracted J0660 flux includes both the Hα and the [N ii] doublet emission. To remove the [N ii] contribution, we used the bimodal empirical relation between the spectro- scopic dust-corrected Hα flux and the total Hα+[N ii] flux found by Vilella-Rojo et al. (2015). This bimodality can be disentan- gled by using the same g′ − i′ color used for the dust correction. Vilella-Rojo et al. (2015) obtained the following expressions by fitting a line to each branch: log(FHα) = { 0.989 log(FHα+[N ii]) − 0.193, if g′ − i′ ≤ 0.5 0.954 log(FHα+[N ii]) − 0.753, if g′ − i′ > 0.5 . (8) In addition, we can easily obtain the [N ii] flux by simply taking the difference between the Hα flux and FHα+[N ii], F[N ii] = FHα+[N ii] − FHα. (9) We did not apply this procedure for SNe at z > 0.014, since the [N ii] λ6583 line is outside the J0660 filter range at a higher redshift and a subtraction of this line cannot be correctly per- formed. Therefore, we assume all emission in the J0660 filter comes from Hα flux for z > 0.014. 3.6. Other continuum subtractions We repeated the subtraction procedure for other narrow filters by applying a different method that was presented in Pascual et al. (2007) and Vilella-Rojo et al. (2015), in which three instead of two filters were used, one narrow filter that contain the feature of interest and two filters (one at each side if the main feature) to trace the continuum, Fcs = (FB1 − FB2) − ( αB1−αB2 αN−αB2 ) (FN − FB2) βB1 − βN ( αB1−αB2 αN−αB2 ) , (10) where Fcs is the flux of the narrow filter with the continuum subtracted, F x is the average flux, and β and α5 are defined as βx ≡ Px(λ = λs)λs∫ Px(λ)λdλ ; αx ≡ ∫ Px(λ)λ2dλ∫ Px(λ)λdλ . (11) The subindices B1, B2, and N refer to the first broad filter, the second one, and the narrow filter, respectively. We note that in this case the continuum is not flat, but it has a slope defined by the two reference filters. Similarly to the method of subtracting the continuum described above, this method also improves significantly when additional filters are incorporated (Martínez-Solaeche et al. 2021). We used this method to subtract the continuum of the J0378 and J0861 filter using the uJAVA and gSDSS and the iSDSS and zSDSS filters, respectively. In the latter case, this is an esti- mate of the calcium triplet absorption. More negative NCR val- ues for absorption filters indicate deeper absorption. Therefore, we inverted the sign of the resulting fluxes when constructing the NCR with this continuum-subtraction method to maintain coher- ence during analysis. 3.7. Error calculation Throughout the preceding calculations, it is essential to con- sider the respective error estimation through robust propaga- tion. The total uncertainty in our measurements of the flux (Eq. (1)) for a source and a given filter (Molino et al. 2014, Logroño-García et al. 2017) is given by σFλ = √( ∂Fλ ∂ZP σZP )2 + ( ∂Fλ ∂C σbn )2 + ( ∂Fλ ∂C σec )2 , (12) σZP being the error of the zero point reflected in Table 2, σec the uncertainly in the electron counting of the CCD, and σbn the large-scale background noise variation; the last two are given by σec = √ C G , σbn = S fit √ Npix ( afit + bfit √ Npix ) , (13) where G is the gain of the detector, Npix the number of pixels; and S fit, afit, and bfit are the resulting coefficients from the fit- ting. This information is found within headers of each image. We calculated the error of the other magnitudes by a quadratic propagation of errors. The error of the flux with the dust correc- tion (Eq. (3)) is given by σFi(λ) = √ (DF0 · σF0(λ))2 + (Dk′ · σk′(λ))2 + (DE · σE(B−V))2, (14) 5 We need to make a calculation that involves an integration of the transmission curve of the pass band; the data of the transmission curve are the same and can be found on the J-PLUS webpage. Article number, page 5 of 17 González-Díaz, R., et al.: A&A, 684, A104 (2024) where DF0 = ∂Fi(λ) ∂F0(λ) , Dk′ = ∂Fi(λ) ∂k′(λ) , and DE = ∂Fi(λ) ∂E(B−V) . σE(B−V) and σk′(λ) are the errors of the color excess (Eq. (5)) and the error given for the extinction law (Eq. (4)), respectively, that are given by σE(B−V) = √( ∂E(B − V) ∂i′ σi′ )2 + ( ∂E(B − V) ∂g′ σg′ )2 , (15) where σg′ and σi′ are the error of the flux in the gSDSS and iSDSS bands given by Eq. (12), and σk′(λ) = √( ∂k′(λ) ∂λ σλ )2 + ( ∂k′(λ) ∂R′V σR′V )2 . (16) Knowing the error of the fluxes with the dust correction, we can now calculate the error of the continuum subtracted fluxes. In the case of the subtraction of the red continuum of the J0660 filter (Eq. (6)), we have σFHα+[N ii] = √( ∂FHα+[N ii] ∂FF600 σFF600 )2 + ( ∂FHα+[N ii] ∂Fr′ σFr′ )2 , (17) with σFr′ and σFF600 being the error of the extinction-corrected flux of the J0660 and rSDSS filters obtained by Eq. (14). We can then calculate the error of the Hα flux with the [N ii] line subtracted (Eq. (8)) and the error of the [N ii] line flux (Eq. (9)) as σFHα = ∂FHα ∂FHα+[N ii] σFHα+[N ii] (18) and σF[N ii] = √( ∂F[N ii] ∂FHα σFHα )2 + ( ∂F[N ii] ∂FHα+[N ii] σFHα+[N ii] )2 . (19) Finally, we can calculate the error of the J0378 and J0861 flux with the continuum subtracted (Eq. (10)) as σFcs = √( ∂Fcs ∂FN σFN )2 + ( ∂Fcs ∂FB1 σFB1 )2 + ( ∂Fcs ∂FB2 σFB2 )2 , (20) with the subindices B1, B2, and N being referred to the first broad filter, the second one, and the narrow filter, respectively. Errors on αx, βx, and ∆x are negligible and are therefore not included in the calculation in order to simplify the procedure. 3.8. Final 3D data cubes Once all these corrections are applied, we store these newly gen- erated images as new slices in the 3D data cube, where fluxes are in the primary extension and flux errors are in the first extension. The final cube has 17 slices: 12 corresponding to the extinction- corrected images in the 12 J-PLUS filters, the Hα+[N ii] contin- uum subtracted flux, the J0378 and J0861 continuum-subtracted flux, the flux corresponding only to the Hα line, and the [N ii] emission line flux. These resulting data cubes are available at Zenodo6 for all 418 SNe host galaxies presented in this work. Figure 3 shows the Hα image of the same galaxy from Fig. 2 at four different steps of the analysis: (a) the initial cutout from the observed image; (b) after dust reddening correction; (c) after r-band continuum subtraction; and (d) after [N ii] emission removal. 6 https://zenodo.org/records/10514632 4. Analysis 4.1. Normalized cumulative-rank distributions The NCR is obtained by sorting the flux values in increasing order (see Fig. 4), constructing the cumulative distribution, and normalizing it to the total emission of the galaxy. This procedure associates each pixel with an NCR value between 0 and 1, where the brightest pixel in the galaxy has an NCR value of one, and all pixels corresponding to sky have an NCR value of zero. Then, the ranked value in this distribution corresponding to the pixel where the SN occurred7 is the NCR associated with the SN. By compiling significant numbers of NCR values for different SN types, we can build NCR distributions and study the differences among types and use this information to infer properties of their progenitors. The NCR method has previously been applied to Hα imag- ing of SN hosts. Assuming that the Hα emission scales by the number of stars that are formed (Kennicutt 1998), a diagonal cumulative NCR distribution with a mean value of 0.5 indicates that the population traces the observed light, and the SN type fol- lows the number of stars formed and mapped by that particular SF tracer. Therefore, if SNe explode predominantly in locations with high SFR, they will favor higher NCRs. On the other hand, SNe that explode in random locations over the galaxy favor low NCRs. Following James & Anderson (2006), we included all nega- tive flux values in the construction of the NCR function; such values are the result of a background subtraction during the pro- cessing with the J-PLUS pipeline (Cenarro et al. 2019). This NCR calculation is applied independently to the 17 images of the 418 SN host galaxies. 4.2. SN environment SED fitting To extract the main properties of SN environments, we con- struct SEDs by measuring aperture photometry in all 12 J-PLUS bands in circular apertures of 1 kpc2 centered at SN locations. To ensure the analysis yields meaningful results, the fluxes need to be only corrected for Milky Way extinction, so here we use the constructed 3D cubes with only this correction applied (at the stage outlined in Sect. 3.3). The size of the aperture is selected so that the flux contained in the aperture has a high enough signal- to-noise ratio to reliably obtain SSP parameters. We made cal- culations using the host galaxy redshift obtained from the OSC and assuming a flat ΛCDM cosmology with H0 = 70 km s−1 Mpc−1. Within the redshift range of our sample this translates into apertures of a 26 to 2 arcsec radius. Figure 5 shows a couple of examples of the rSDSS filter galaxy cutout and the SED from the 12 J-PLUS bands. We note how the SED shape of an SN environment with ongoing star formation (SN 1997dq on top) shows a bluer continuum with a clear Hα emission, while the SED of a SN environment in a passive galaxy (SN 1957B on the bottom) looks redder and smoother. Once the 418 SEDs were built, we used FAST++ (Schreiber et al. 2018)8, a C++ implementation of FAST (Kriek et al. 2009), to fit them and obtain a number of galaxy physical properties. Given galaxy photometry, FAST++ deter- mines the best-fit SED from a library of simple stellar population (SSP) models. These models are initially synthesized in compos- ite stellar populations (CSPs) on a 5D grid, with each grid point 7 The precision of SN locations is lower than the image resolution (see Sect. 1), so what we are measuring is the immediate SNe environment. 8 Available at https://github.com/cschreib/fastpp Article number, page 6 of 17 González-Díaz, R., et al.: A&A, 684, A104 (2024) 2h47m04s 00s 46m56s 52s 36°40' 39' 38' 37' Right Ascension De cli na tio n 2h47m04s 00s 46m56s 52s 36°40' 39' 38' 37' Right Ascension De cli na tio n 2h47m04s 00s 46m56s 52s 36°40' 39' 38' 37' Right Ascension De cli na tio n 2h47m04s 00s 46m56s 52s 36°40' 39' 38' 37' Right Ascension De cli na tio n Fig. 3. J-PLUS NGC1058 J0660 image with the type II SN1961V marked in red. Top left: before dust extinction correction. Top right: after dust extinction correction. Bottom left: after continuum subtraction. Bottom right: after [N ii] removal. 0 2500 5000 7500 10000 12500 15000 17500 Pixel index 0.0 0.5 1.0 1.5 2.0 Pi xe l F lu x (W ·m 2 ·H z 1 ) 1e 17 0.0 0.2 0.4 0.6 0.8 1.0 NC R0.49 4.29e-18 15824 Fig. 4. Example of calculation of NCR value for SN pixel. The red curve is the sorted pixel value (left flux scale), while the blue curve is the cumulative distribution of the flux values, i.e., the normalized cumulative-rank pixel-value function (NCRPVF or NCR for simplic- ity). The blue dot represents the position and corresponding NCR value of the SN 2008az pixel, and the red dot represents the corresponding flux value. corresponding to a CSP SED with age t∗, stellar metallicity Z∗, V band extinction AV , timescale τ, and star formation history (SFH) at a given redshift z. At each point on the grid, a χ2 value is calculated as χ2F = N∑ i [ Fλ,i − Fλ,mod(t∗, τ, AV ,Z∗, z)]2 σ2i , (21) where N is the number of photometric points and σi is the error for point i. One-sigma confidence intervals for parameters and derived quantities (such as stellar mass and SFR) are deter- mined using Monte Carlo sampling of the grid around the lowest χ2. In this work, we used the BC03 (Bruzual & Charlot 2003) SSP library calculated using a delayed exponential (delayed-τ) SFH parametrization and a Chabrier (2003) initial mass func- tion. Dust extinction is modeled with a Calzetti et al. (2000) dust extinction law with a uniform and constant foreground dust screen. In all fits, the redshift was fixed to the reported spectro- scopic redshift of the SN host galaxy. The output of FAST++ includes the distribution of stellar populations of different ages and metallicities present at that location. Although several degeneracies are at play, with large enough samples one can disentangle the underlying differences Article number, page 7 of 17 González-Díaz, R., et al.: A&A, 684, A104 (2024) Fig. 5. Example of two SEDs of SN local environments of one squared kiloparsec extracted from J-PLUS observations. On top, the star-forming environment of SN 1997dq in its host galaxy NGC 3810 clearly shows a blue continuum based on the five broad-band filters and increased flux for the narrow-band filters where the main emission lines fall. On the bottom, for SN 1957B in the Messier 84 galaxy, a smooth red continuum is seen with no bumps at the narrow-band filter wavelengths. between the average properties of these parent populations for different SN types. In particular, we mainly focused on stellar population age, dust extinction, stellar mass, and star formation rate. 5. Results 5.1. NCR distributions The resulting NCRs for all individual SNe are provided in Tables A.2–A.4. With the NCR and error for all SNe in our sam- ple, we constructed cumulative distributions for all SN types and in all 17 frames of the data cubes as follows. We built a prob- ability distribution for every NCR using 30 000 NCR trials of the form NCR + (σNCR · α), α being a random variable between −1 and 1. We sorted all the NCRs in increasing order, and the median and one sigma of those distributions of 30 000 NCRs are shown in the plots. In Fig. 6, we show the 12 panels, one for each broad and narrow band, with the corresponding NCR dis- tributions for the seven SN types. In general, we observe that as we move to redder wave- lengths (from gSDSS to zSDSS), all SN type distribution tends to be equal, following the black diagonal distribution, with an average NCR & 0.4 for every distribution. Type II, IIb, and Ia SN distributions begin to separate from the diagonal, and type Ibc, Ib, Ic, and IIn distributions begin to separate at bluer wave- lengths. Also, type Ib SNe distribution gets closer to the diagonal as we move to bluer bands. The order Ia-II-IIb-IIn-Ibc remains mostly equal in all filters; however, IIn and IIb types present higher and lower NCRs in red bands, respectively. We conducted Kolmogorov–Smirnov (KS) tests to compare two samples and assess the probability of both being drawn from the same probability distribution. By setting a significance level of 2σ, if the p value computed by the KS test falls below the threshold of 0.05, it means the rejection of the null hypothesis. In other words, it indicates they originated from different popula- tions. Table 3 shows the p value for every pair of NCR distribu- tions and the diagonal hypothetical distribution that accurately traces the respective flux ([1:1]), and also for every filter. The table summarizes what is shown in Fig. 6. For the 12 J-PLUS filters, all distributions for filters redder than J0410 show p val- ues > 0.05 (with the exception of the combination of II vs. Ibc types for J0515 and Ia vs. Ibc for J0410). For uJAVA, J0378, and J0395, Ia and II-IIb-IIn shows p-value > 0.05. Only com- binations of CC-types with Ia, and combinations of II with Ib, Ic and Ibc presents p-values < 0.05, indicating that types Ia, II, IIb, and IIn come from different possible progenitors than type Ib, Ic, and Ibc SNe that are located in redder environments in comparison with the rest of the types. The distributions of the five continuum-subtracted narrow- band filters are presented in Fig. 7. We note again that SNe with z > 0.014 are not included in the N ii distributions as explained in Sect. 3.5. The most significant characteristic is the high number of zero NCR values in all cases. As we discuss later (Sect. 6), we associated part of the zero values found with the small aperture of the JAST80 telescope and the exposure times Article number, page 8 of 17 González-Díaz, R., et al.: A&A, 684, A104 (2024) 0.0 0.2 0.4 0.6 0.8 1.0 NCR 0.0 0.2 0.4 0.6 0.8 1.0 C um ul at iv e di st ri bu tio n uJAVA II 156 Ia 108 Ib 51 Ic 49 Ibc 134 IIn 20 IIb 27 0.0 0.2 0.4 0.6 0.8 1.0 NCR 0.0 0.2 0.4 0.6 0.8 1.0 C um ul at iv e di st ri bu tio n J0378 II 156 Ia 108 Ib 51 Ic 49 Ibc 134 IIn 20 IIb 27 0.0 0.2 0.4 0.6 0.8 1.0 NCR 0.0 0.2 0.4 0.6 0.8 1.0 C um ul at iv e di st ri bu tio n J0395 II 156 Ia 108 Ib 51 Ic 49 Ibc 134 IIn 20 IIb 27 0.0 0.2 0.4 0.6 0.8 1.0 NCR 0.0 0.2 0.4 0.6 0.8 1.0 C um ul at iv e di st ri bu tio n J0410 II 156 Ia 108 Ib 51 Ic 49 Ibc 134 IIn 20 IIb 27 0.0 0.2 0.4 0.6 0.8 1.0 NCR 0.0 0.2 0.4 0.6 0.8 1.0 C um ul at iv e di st ri bu tio n J0430 II 156 Ia 108 Ib 51 Ic 49 Ibc 134 IIn 20 IIb 27 0.0 0.2 0.4 0.6 0.8 1.0 NCR 0.0 0.2 0.4 0.6 0.8 1.0 C um ul at iv e di st ri bu tio n gSDSS II 156 Ia 108 Ib 51 Ic 49 Ibc 134 IIn 20 IIb 27 0.0 0.2 0.4 0.6 0.8 1.0 NCR 0.0 0.2 0.4 0.6 0.8 1.0 C um ul at iv e di st ri bu tio n J0515 II 156 Ia 108 Ib 51 Ic 49 Ibc 134 IIn 20 IIb 27 0.0 0.2 0.4 0.6 0.8 1.0 NCR 0.0 0.2 0.4 0.6 0.8 1.0 C um ul at iv e di st ri bu tio n rSDSS II 156 Ia 108 Ib 51 Ic 49 Ibc 134 IIn 20 IIb 27 0.0 0.2 0.4 0.6 0.8 1.0 NCR 0.0 0.2 0.4 0.6 0.8 1.0 C um ul at iv e di st ri bu tio n J0660 II 156 Ia 108 Ib 51 Ic 49 Ibc 134 IIn 20 IIb 27 0.0 0.2 0.4 0.6 0.8 1.0 NCR 0.0 0.2 0.4 0.6 0.8 1.0 C um ul at iv e di st ri bu tio n iSDSS II 156 Ia 108 Ib 51 Ic 49 Ibc 134 IIn 20 IIb 27 0.0 0.2 0.4 0.6 0.8 1.0 NCR 0.0 0.2 0.4 0.6 0.8 1.0 C um ul at iv e di st ri bu tio n J0861 II 156 Ia 108 Ib 51 Ic 49 Ibc 134 IIn 20 IIb 27 0.0 0.2 0.4 0.6 0.8 1.0 NCR 0.0 0.2 0.4 0.6 0.8 1.0 C um ul at iv e di st ri bu tio n zSDSS II 156 Ia 108 Ib 51 Ic 49 Ibc 134 IIn 20 IIb 27 Fig. 6. Cumulative NCR distributions of 12 broad- and narrow-band filters. The straight black diagonal line represents a hypothetical distribution, infinite in size, which accurately traces the respective observed flux. Each SN distribution corresponds to the median and one sigma of the 30 000 NCR trial distributions as detailed in Sect. 5.1 of Lyman et al. (2021). It can be observed that as we move to redder filters (from gSDSS to zSDSS), the distributions for each type tend to become more similar along the diagonal. In contrast, there is a clearer distinction between them in the blue filters with more zero values. Article number, page 9 of 17 González-Díaz, R., et al.: A&A, 684, A104 (2024) Table 3. p-values of the Kolmogorov–Smirnov test for every SN type combination and filter. Combination uJAVA J0378 J0395 J0410 J0430 gSDSS J0515 rSDSS J0660 iSDSS J0861 zSDSS Hα+[N ii] [O ii] Ca II triplet Hα [N ii] II vs. Ia 0.213 0.740 0.709 0.757 0.646 0.896 0.999 0.442 0.408 0.997 0.974 0.973 0.076 0.013 0.990 0.399 0.527 II vs. Ib 0.002 0.096 0.023 0.638 0.392 0.386 0.083 0.279 0.169 0.496 0.632 0.925 0.095 0.347 0.158 0.037 0.017 II vs. Ic 0.021 0.164 0.184 0.438 0.251 0.810 0.318 0.777 0.298 0.421 0.367 0.391 0.017 0.461 0.001 0.001 0.002 II vs. Ibc 0.001 0.037 0.019 0.194 0.073 0.289 0.023 0.229 0.117 0.399 0.380 0.495 0.000 0.134 0.003 0.000 0.000 II vs. IIn 0.410 0.321 0.773 0.695 0.819 0.343 0.487 0.211 0.750 0.814 0.446 0.434 0.193 0.364 0.109 0.069 0.026 II vs. IIb 0.998 0.918 0.817 0.899 0.995 0.985 0.751 0.695 0.889 0.805 0.971 0.987 0.421 0.848 0.903 0.859 0.543 Ia vs. Ib 0.000 0.017 0.021 0.334 0.117 0.270 0.135 0.224 0.173 0.614 0.777 0.977 0.000 0.009 0.256 0.001 0.002 Ia vs. Ic 0.000 0.027 0.046 0.104 0.165 0.770 0.525 0.669 0.439 0.704 0.588 0.740 0.000 0.003 0.014 0.000 0.000 Ia vs. Ibc 0.000 0.003 0.006 0.046 0.074 0.217 0.092 0.342 0.217 0.506 0.697 0.959 0.000 0.000 0.026 0.000 0.000 Ia vs. IIn 0.086 0.184 0.362 0.449 0.926 0.155 0.595 0.271 0.774 0.832 0.557 0.557 0.006 0.076 0.263 0.014 0.010 Ia vs. IIb 0.627 0.837 0.700 0.938 0.999 0.837 0.893 0.700 0.999 0.771 0.996 0.999 0.018 0.043 0.965 0.310 0.121 Ib vs. Ic 0.988 0.999 0.926 0.999 1.000 0.879 0.701 0.890 0.999 0.994 0.991 0.984 0.990 1.000 0.409 0.380 0.895 Ib vs. Ibc 0.967 1.000 0.975 1.000 0.999 0.985 0.924 0.998 0.999 1.000 1.000 0.999 0.989 0.997 0.998 0.978 0.995 Ib vs. IIn 0.806 0.999 0.889 0.999 0.405 0.851 0.987 0.494 0.999 0.947 0.967 0.687 0.903 0.707 0.947 0.754 0.754 Ib vs. IIb 0.131 0.352 0.862 0.987 0.553 0.528 0.221 0.295 0.271 0.624 0.760 0.972 0.964 0.964 0.770 0.630 0.630 Ic vs. Ibc 1.000 1.000 0.999 1.000 1.000 0.999 0.999 0.991 1.000 0.999 0.999 0.999 0.997 0.999 0.744 0.827 0.995 Ic vs. IIn 0.934 0.997 0.966 0.999 0.469 0.731 0.975 0.789 0.999 0.995 0.951 0.951 0.857 0.707 0.984 0.681 0.883 Ic vs. IIb 0.286 0.349 0.897 0.835 0.620 0.793 0.865 0.512 0.438 0.280 0.488 0.512 0.798 0.964 0.118 0.179 0.335 Ibc vs. IIn 0.965 0.995 0.927 1.000 0.466 0.853 0.999 0.579 0.999 0.973 0.903 0.772 0.569 0.652 0.999 0.840 0.813 Ibc vs. IIb 0.264 0.393 0.927 0.957 0.732 0.673 0.613 0.393 0.346 0.443 0.673 0.788 0.810 0.880 0.372 0.314 0.487 IIn vs. IIb 0.890 0.688 0.996 0.996 0.996 0.551 0.749 0.284 0.565 0.714 0.376 0.578 0.549 0.873 0.498 0.553 0.553 II vs. [1:1] 0 0 0 0 0 0.00038 0 0 0 0 0 0 0 0 0 0 0 Ia vs. [1:1] 0 0 0 0 0 0.00125 0 0.00027 0.00028 0.00055 0 0 0 0 0 0 0 Ib vs. [1:1] 0.014 0.001 0.001 0.003 0.128 0.357 0.667 0.107 0.311 0.035 0.060 0.019 0.047 0.056 0.003 0.100 0.004 Ic vs. [1:1] 0.002 0.000 0.001 0.022 0.071 0.257 0.069 0.213 0.339 0.216 0.144 0.071 0.295 0.018 0.278 0.908 0.193 Ibc vs. [1:1] 0.000 0.000 0.000 0.000 0.000 0.146 0.016 0.031 0.036 0.004 0.002 0.000 0.011 0.001 0.000 0.053 0.000 IIn vs. [1:1] 0.028 0.021 0.036 0.146 0.016 0.855 0.643 0.952 0.568 0.366 0.496 0.496 0.396 0.473 0.317 0.711 0.554 IIb vs. [1:1] 0.000 0.000 0.001 0.016 0.005 0.107 0.017 0.020 0.068 0.015 0.023 0.023 0.037 0.017 0.000 0.007 0.001 Notes. We also included the p-values for the SN types and the diagonal hypothetical distribution that accurately traces the respective flux ([1:1]). Black font boxed values represent those distributions without an underlying causative relationship; i.e., we reject the null hypothesis that the two samples were drawn from the same probability distribution (p-value< 0.05), and both NCR distributions come from different populations. The distribution used for the five continuum-subtracted filters have been rebuilt by proportionally removing the number of zero NCR values in each distribution that correspond to the lowest observed fraction of zeros in an individual distribution (see Sect. 6). of the images used, which is not enough to collect enough pho- tons to provide a high signal-to-noise ratio in these continuum- subtracted images for all galaxies. Moreover, Hα is associated with stars of M > 15−20 M , so many SN progenitors are not expected to be associated with this emission, contributing to more zero NCRs in the distributions. The removal of the [N ii] line has no appreciable effect on the NCR distributions, with the plots of Hα+[N ii], Hα and [N ii] being similar. In any case, we observe a similar scenario between the broad- and narrow-band filters, where the type Ia distribution is clearly the one with the lowest NCRs in all cases, followed by type II. The [O ii] emis- sion also traces the star formation rate (Kennicutt 1998), sim- ilarly to Hα. With the change of signs in Ca II triplet fluxes, a higher NCR after subtracting the continuum suggests deeper absorption, which is analogous to high NCR values in emission- line filters and indicates strong emission (see Sect. 3.6). The order of the distributions appears to remain consistent, with Ia, II, and IIb displaying lower NCR values, distinguishing them from the other SN types. However, the large amount of zero val- ues makes the analysis difficult in these cases. 5.2. SN environment parameter distributions The local stellar population parameters AV , t∗, SFR, and sSFR for all 418 SN environments are reported in Table A.5. Addi- tionally, their distributions are illustrated in Fig. 8 and segre- gated according to the same seven SN groups. These have been constructed similarly to in the previous section, that is, by per- forming 30 000 different distributions where in each realization the best output values from the best FAST++ fits are randomly varied within their 1σ uncertainty, and the median and the 1σ variation of all realizations is taken as the final distribution. We recover previous trends already found in the literature. The average local AV is larger for SNe Ibc (1.13 ± 0.08 mag) compared to SNe Ia (0.63 ± 0.09 mag), with the SNII distribu- tion being in the middle of the two (0.85 ± 0.07 mag), as pre- viously reported by Galbany et al. (2017). This trend is driven by the SNIb and SNIc distributions, which are clearly shifted to higher values compared to all others. Similarly, all CC SNe distributions show younger average ages (38 ± 10 Myr for SNe II and 43 ± 13 Myr for SNe Ibc) compared to those of SNe Ia environments (222 ± 73 Myr), and CC SNe environments have both higher SFR (0.08 ± 0.06 and 0.16 ± 0.04 M yr−1 for SNe Ibc and SNe II, respectively) and sSFR (1.88 ± 1.40 and 4.35±1.29×10−8 yr−1 for SNe Ibc and SNe II, respectively) than SNe Ia (0.01 ± 0.01 M yr−1 and 0.19 ± 0.16 × 10−8 yr−1) that occur in more passive local environments (Galbany et al. 2018). 5.3. Radial distributions of the SNe To investigate potential correlations between the emission of a specific band and the distribution of SNe in galaxies, we calculated two parameters: the normalized galactocentric dis- tance (NGCD) and the fraction of the r-band emission flux (Fr; Anderson & James 2009; Habergham et al. 2010). Our approach involved fitting an elliptical 2D Gaussian model to the rSDSS images of the galaxies to determine the shape and size of each galaxy. If the SN was located at the center of the ellipse, NGCD was assigned a value of 0; if the SN was positioned at the outer ellipse’s edge, the NGCD equaled 1. We then fit a second ellipse, Article number, page 10 of 17 González-Díaz, R., et al.: A&A, 684, A104 (2024) 0.0 0.2 0.4 0.6 0.8 1.0 NCR 0.0 0.2 0.4 0.6 0.8 1.0 C um ul at iv e di st ri bu tio n H +[NII] II 156 Ia 108 Ib 51 Ic 49 Ibc 134 IIn 20 IIb 27 0.0 0.2 0.4 0.6 0.8 1.0 NCR 0.0 0.2 0.4 0.6 0.8 1.0 C um ul at iv e di st ri bu tio n H II 156 Ia 108 Ib 51 Ic 49 Ibc 134 IIn 20 IIb 27 0.0 0.2 0.4 0.6 0.8 1.0 NCR 0.0 0.2 0.4 0.6 0.8 1.0 C um ul at iv e di st ri bu tio n [OII] II 156 Ia 108 Ib 51 Ic 49 Ibc 134 IIn 20 IIb 27 0.0 0.2 0.4 0.6 0.8 1.0 NCR 0.0 0.2 0.4 0.6 0.8 1.0 C um ul at iv e di st ri bu tio n Ca II triplet II 156 Ia 108 Ib 51 Ic 49 Ibc 134 IIn 20 IIb 27 0.0 0.2 0.4 0.6 0.8 1.0 NCR 0.0 0.2 0.4 0.6 0.8 1.0 C um ul at iv e di st ri bu tio n [NII] II 116 Ia 72 Ib 39 Ic 30 Ibc 94 IIn 12 IIb 20 Fig. 7. Cumulative NCR distributions of five constructed filters. Hα+[N ii], [O ii] and Ca II triplet distributions correspond to the J0660, J0378, and J0861 continuum-subtracted filters, respectively. The Hα filter is the result of removing the [N ii] contribution, which is only applied to SNe with z < 0.014. All distributions correspond to the median and one sigma of the 30 000 NCR trial distributions as outlined in Sect. 5.1 of Lyman et al. (2021). The straight black diagonal line represents a hypothetical distribution, infinite in size, that accurately traces the respective observed flux. The gray shading marks the lowest values of zeros in a distribution in each filter. with the same parameters as the first one, but scaled to place the SN at the edge of this second ellipse. Fr was subsequently com- puted as the ratio between the integrated fluxes in the rSDSS band of the second and first ellipses. We find in Fig. 9 that all SN types exhibit nearly identical distributions with respect to each other, except for type IIb SNe, which appear to be concentrated in the inner regions of galaxies (within the range of 0.2 to 0.6 in NGCD). For the other SN types, there does not seem to be a preference for any specific location within the galaxies. The Fr are also similar among all SN types, and close to the diagonal, tracing the rSDSS emission as the NCR distributions of Fig. 6 do. Since all J-PLUS bands present the NCR distri- butions close to the diagonal, and there is no preference in the location of the SN, we only traced the general shape and size of the galaxies. 6. Discussion 6.1. NCR results Concerning the NCR distributions of the broad- and narrow- band filters and the average NCR values, it is worth noting that the broad filters exhibit slightly lower errors than the narrow- band filters, particularly for those where the continuum has been subtracted. The errors for the uJAVA filter are higher than the rest of the broad-band filters. This is what we expected since the broad filters have higher S/Ns, and the uJAVA filter has the lower FWHM of all J-PLUS broad-band filters. We also observe approximately the same sequence between the main types in all panels in Fig. 6; this is type Ibc between II and Ia for bluer wave- lengths, moving upward in the cumulative plot and downward for NCRs for redder wavelengths. This shift of the Ibc distribu- tion is due to the Ib contribution, which correlates more with bluer bands, similarly to what was reported by Kangas et al. (2013). Kelly & Kirshner (2012) also found that the environ- ments of SN Ib exists preferentially in environments with bluer surface brightness, especially in the ultraviolet band. In our case, we also found that the filter that best distinguishes the type Ib distribution is uJAVA. By looking at continuum-subtracted filters, the main charac- teristic is the large amount of zeros present in all distributions. For instance, in the Hα plot we have a noticeable amount of zero NCR values: ∼35% for Ibc, ∼53% for II, and ∼65% for Ia. This is partly because many SNe do not follow the Hα emis- sion. However, this fraction of zeros is larger than in previous studies (e.g., Anderson et al. 2012 found ∼25% for Ibc, ∼39% for II, and ∼58% for Ia). Since these filters represent the residual flux once a pseudo-continuum (from broad-band filters) has been subtracted from a narrow-band filter, and since the diagonal line assumes a population that specifically traces the observed light, the distribution will include a fraction of zeros that depends on the image depth. Therefore, we have two possible explanations for this large amount of zeros; either most of the flux collected by the narrow-band filters is due to the underlying continuum, or the combination of a small telescope aperture (0.8 m) and low Article number, page 11 of 17 González-Díaz, R., et al.: A&A, 684, A104 (2024) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 AV 0.0 0.2 0.4 0.6 0.8 1.0 C u m u la ti ve d is tr . SNe Ia (108) SNe II (156) SE SNe (134) SNe Ib (51) SNe Ic (49) SNe IIb (27) SNe IIn (20) 6 7 8 9 10 log10 (t [yr]) −3 −2 −1 0 1 2 log10 (SFR [M¯ yr−1]) 0.0 0.2 0.4 0.6 0.8 1.0 C u m u la ti ve d is tr . −11 −10 −9 −8 −7 −6 log10 (sSFR [yr −1]) Fig. 8. Cumulative distributions of each of three main SN types for main environmental parameters obtained from best FAST++ fits to SED: extinction AV , average stellar age 〈t〉, star-formation rate log10 SFR, and specific star-formation rate log10 SFR/Mass. exposure time was not enough to get significant contrast in these continuum-subtracted filters. To further analyze this, we rebuilt the NCR distributions, this time proportionally removing the number of zero NCR values in each distribution that correspond to the lowest observed frac- tion of zeros in an individual distribution, following a proce- dure similar to Ransome et al. (2022). This can be done as all our images are of a similar depth, since all the images were obtained using the same telescope with the same conditions, exposure time, and instrument configuration. This is equivalent to excluding the shaded region in all panels of Fig. 7, because we assume we have reached the background level at that point, and tracing the diagonal from the intersection of the dotted hor- izontal line with the zero NCR to the upper right corner. This assumption is grounded in the fact that we successfully retrieved a certain level of signal, and anything below that threshold corre- sponds to the background we were not able to recover due to the combination of the telescope and exposure time used, as previ- ously mentioned. In Fig. 10, we show the resulting distributions. Hα+[N ii], Hα, and [N ii] plots still show similar distributions; the main type that correlates the emission is Ic, with the high- est NCR values on average (Fig. 11). For the J0660 (Fig. 6) and Hα (Fig. 10) panels for type Ia, II, and all Ibcs combined, we recovered the tendency of the distributions seen in previous works (e.g., Anderson & James 2008), where type Ic/Ibc SNe is the closest distribution to the diagonal line, followed by type II and type Ia SNe. We interpreted this as a sequence in progen- itor age and then inferred a sequence of progenitor mass, with the distributions closer to the diagonal being the SN type with higher mass progenitors (Kuncarayakti et al. 2018). Type IIn SN distribution also correlates with the Hα emis- sion, with lower NCR values than Ic (∼0.05 on average), which is also consistent with the literature (Ransome et al. 2022), with the same methodology), however, our nonzero IIn sample is reduced to only seven SNe. All CC SNe types are strongly corre- lated to the [O ii] emission, that also traces SFR (Kennicutt 1998) following the sequence of Ia-II-IIb-Ic-Ibc-Ib-IIn. If we exclude the bias in the IIn type, which is limited to only four SNe and not numerous enough to perform reliable statistics, we can observe that type Ib exhibits the most pronounced correlation with this emission, primarily because this type of SN is more closely associated with bluer bands, as previously mentioned. The same Article number, page 12 of 17 González-Díaz, R., et al.: A&A, 684, A104 (2024) Fig. 9. Cumulative distributions of normalized galactocentric distances (top) and Fr (bottom) for every SN type. correlations between SN types are found for the Hα+[N ii], Hα, and [N ii] filters. Besides, the Ia, IIb, and IIn types are correlated when the [N ii] is subtracted, the p-value being > 0.05 after the removal. The Ca II triplet distribution makes a clear division between types II, IIb, and Ia and types Ib, Ic, and IIn (closer to the diago- nal because of the change in the sign). These differences between populations are also evident when computing the p value of the KS test (Table 3): II versus Ic, II versus Ibc, Ia versus Ic, and Ia versus Ibc present p values < 0.05 for the Ca II triplet NCRs; II versus Ia, II versus IIb, and Ia versus IIb present p values > 0.05, as well as the rest of CC SNe. The interpretation of the intensity of the Ca II triplet absorption is not trivial. Some works (e.g., Diaz et al. 1989; Garcia-Vargas et al. 1998) have shown that the Ca II triplet equivalent width (EW) presents a degen- eracy between stellar population age and metallicity. At high metallicity, the Ca II EW reaches its maximum value for stel- lar populations of ∼10 Myr, primarily attributed to the presence of red supergiant (RSG) stars. A secondary, lower peak emerges around 100 Myr, driven by stars in the asymptotic giant branch (AGB) phase. Furthermore, for ages over 1 Gyr, the Ca II EW becomes a reliable indicator of increasing metallicity. Based on our results, and compared to SNII that explode in low Ca II NCR environments, one possible interpretation would be to attribute the higher values for Ib/Ic to young progenitors exploding at locations with metal-rich older populations. For SNIIn, although with low numbers, our results are in line with these progenitors being a combination of very young (∼10 Myr peak) and older (∼100 Myr peak) stars (Galbany et al. 2018). Finally, in Fig. 11 we show the average NCR values in all filters for the 7 SN types. On the left panel, we see that on average, through the 12 J-PLUS filters, SNe Ia present the lower NCR values (between 0.10 ± 0.02 and 0.33 ± 0.02), fol- lowed by II AV (mag) log SFR log M SN2021yok II 6.600.08−0.13 2.10 0.03 −0.15 0.55 0.19 −0.23 6.85 0.07 −0.12 SN2022prv II 6.801.31−0.17 0.80 0.41 −0.80 -0.44 0.18 −1.29 6.06 0.05 −0.16 SN1998ar II 6.600.10−0.11 1.40 0.03 −0.17 0.12 0.17 −0.27 6.42 0.07 −0.19 SN1990L Ia 7.650.05−0.10 0.00 0.00 −0.00 -3.31 0.07 −0.03 3.99 0.07 −0.04 SN2019ltw IIb 6.550.01−0.30 1.40 0.05 −0.09 0.93 0.52 −0.01 7.18 0.24 −0.02 SN2018cow Ic 9.500.02−2.47 0.00 1.43 −0.00 -2.16 1.79 −0.02 6.87 0.02 −0.28 SN2020acac II 8.400.02−1.78 0.00 1.50 −0.00 -1.35 1.74 −0.03 6.72 0.05 −0.05 ... Note: The full table is available at the CDS. Article number, page 17 of 17