PGIR22akgylf: Slow-Rise Galactic Nova

• PGIR22akgylf is a classical nova in Cygnus distinguished by a 133-day gradual rise to optical maximum and severe reddening due to high extinction.
• Multi-telescope observations, including TESS, ZTF, and infrared data, reveal a stable 0.1802-day periodic brightness modulation linked to the binary system.
• The analysis supports a common-envelope interpretation with a dwarf donor, challenging conventional models of rapid nova eruptions.

PGIR22akgylf is a highly reddened Galactic classical nova in Cygnus, also designated MASTER OT J200029.27+345309.1, ZTF22abazrjk, and AT 2022sfe. Its distinguishing observational properties are an exceptionally slow ascent to optical maximum, lasting about $133$ days, and a stable periodic brightness modulation detected by TESS while the nova was still rising and the optical light was expected to be dominated by the nova photosphere rather than the underlying binary. This combination places PGIR22akgylf among the rare slowly rising novae, but with a phenomenology that indicates a close binary with a dwarf donor rather than a symbiotic system with a giant companion (Sokolovsky et al., 10 Jun 2026).

1. Identification, discovery, and observational setting

PGIR22akgylf was discovered on 2022-08-16.1900 UTC, adopted as

$t_0 = {\rm JD(UTC)}\ 2459807.6900.$

The adopted position, from the median of 227 ZTF detections, is $20{:}00{:}29.254\ +34{:}53{:}09.17$ (J2000), with a positional scatter of $0.06$ arcsec and a systematic tie to Gaia expected to be better than $0.014$ arcsec. Its Galactic coordinates are $l = 71.29366$ and $b = 2.53075$. The line of sight is strongly extinguished, with total line-of-sight extinction quoted as $A(V)=9.22$, and the raw optical spectrum has very little signal below $4500$ Å, consistent with severe reddening (Sokolovsky et al., 10 Jun 2026).

The nova was first found by Palomar Gattini-IR on 2022-08-16.1900 UTC at $J = 14.28$. MASTER independently detected it the same night at an unfiltered optical magnitude of about $t_0 = {\rm JD(UTC)}\ 2459807.6900.$0. By 2022-08-22.26 ($t_0 = {\rm JD(UTC)}\ 2459807.6900.$1 d), ZTF measured $t_0 = {\rm JD(UTC)}\ 2459807.6900.$2 and $t_0 = {\rm JD(UTC)}\ 2459807.6900.$3. A high-resolution spectrum obtained on 2022-08-29 with the Magellan Clay telescope established the classification as a classical nova.

The reconstruction of the eruption used a heterogeneous time-domain data set: PGIR $t_0 = {\rm JD(UTC)}\ 2459807.6900.$4-band photometry at roughly 1-day cadence, public ZTF $t_0 = {\rm JD(UTC)}\ 2459807.6900.$5 photometry, ATLAS cyan and orange forced photometry, and AAVSO $t_0 = {\rm JD(UTC)}\ 2459807.6900.$6 photometry. The AAVSO contribution included 1602 $t_0 = {\rm JD(UTC)}\ 2459807.6900.$7-band measurements over 22 nights from observer RS and 75 $t_0 = {\rm JD(UTC)}\ 2459807.6900.$8-band points over 27 nights plus 42 $t_0 = {\rm JD(UTC)}\ 2459807.6900.$9-band points over 14 nights from observer FDR. TESS provided the decisive high-cadence space-based coverage in Sector 55, spanning 2022-08-05 to 2022-09-01, corresponding to $20{:}00{:}29.254\ +34{:}53{:}09.17$0 to $20{:}00{:}29.254\ +34{:}53{:}09.17$1 d, with nominal 600 s full-frame cadence and effective on-source exposure of 475 s. There was one data-downlink gap near $20{:}00{:}29.254\ +34{:}53{:}09.17$2 to $20{:}00{:}29.254\ +34{:}53{:}09.17$3 d.

PGIR itself is a $20{:}00{:}29.254\ +34{:}53{:}09.17$4-band infrared time-domain survey instrument described as a 30 cm telescope with a $20{:}00{:}29.254\ +34{:}53{:}09.17$5 camera surveying $20{:}00{:}29.254\ +34{:}53{:}09.17$6 of the northern sky at a cadence of about 2 days, a configuration that motivated its use in transient discovery and monitoring in dusty sightlines (Karambelkar et al., 2020).

2. Photometric evolution and the slow-rise phenomenon

The defining property of PGIR22akgylf is the absence of a normal rapid rise to maximum. The combined light curve shows that it reached optical peak on 2022-12-26.9296, corresponding to $20{:}00{:}29.254\ +34{:}53{:}09.17$7 d, at $20{:}00{:}29.254\ +34{:}53{:}09.17$8. The decline time by two magnitudes in $20{:}00{:}29.254\ +34{:}53{:}09.17$9 was estimated as $0.06$0 d. Both the ascent and decline therefore place it among slow novae (Sokolovsky et al., 10 Jun 2026).

This rise morphology is atypical for classical novae. Most novae brighten the bulk of the way to optical maximum within about a day, sometimes followed by only a later slow 1–2 mag climb. PGIR22akgylf instead appears to have climbed gradually over the entire $0.06$1-day interval. During the TESS observations the system was already at least $0.06$2 mag above quiescence but still about 4 mag below peak. The eruption was thus well underway while remaining far from its maximum photospheric extent and optical output.

The color evolution during the TESS-covered rise was minimal. Representative measurements include $0.06$3 and $0.06$4 at $0.06$5 d; $0.06$6, $0.06$7, $0.06$8, and $0.06$9 at $0.014$0 d; $0.014$1 and $0.014$2 at $0.014$3 d; $0.014$4 and $0.014$5 at $0.014$6 d; and $0.014$7 at $0.014$8 d. No significant color change was identified during the initial rise covered by TESS. This supports the interpretation of a gradually and roughly isothermally expanding photosphere rather than a simple rapidly expanding cooling fireball.

A possible pre-discovery brightening around $0.014$9 d appears in the TESS light curve, but that feature was treated cautiously because a similar pattern is present in a nearby check star and may be instrumental. The robust short-timescale variability attributed to the nova is therefore restricted to the second half of Sector 55.

3. TESS periodic modulation during the rise

The most distinctive time-series result is a coherent periodic brightness modulation detected in the TESS light curve during the second half of Sector 55, when the nova had brightened enough to dominate the source aperture. The measured period is

$l = 71.29366$0

equivalent to about 4.33 h, with light ephemeris

$l = 71.29366$1

The period uncertainty was estimated conservatively from

$l = 71.29366$2

where $l = 71.29366$3 is the period and $l = 71.29366$4 is the duration of the analyzed light-curve segment (Sokolovsky et al., 10 Jun 2026).

The abstract gives a peak-to-peak amplitude of $l = 71.29366$5 mag. In TESS flux units, the three chunks with significant detection yielded peak-to-middle sine-fit amplitudes of $l = 71.29366$6, $l = 71.29366$7, and $l = 71.29366$8, with mean fluxes of 228.7, 401.9, and 578.4, respectively. The corresponding fractional peak-to-peak variations were 0.025, 0.020, and 0.016. Thus the absolute flux amplitude increased as the nova brightened, while the fractional modulation decreased modestly from about 2.5% to 1.6%.

The periodicity was stable across the later three time chunks: $l = 71.29366$9 d, $b = 2.53075$0 d, and $b = 2.53075$1 d. No significant period change was found over the roughly two weeks in which the signal was present, and the light curve traces more than 70 cycles. The phased signal is approximately sinusoidal. This coherence is central to the interpretation, because some nova-envelope pulsation models predict only 5–10 cycles, whereas the observed modulation persists for far longer.

The TESS extraction used Lightkurve with custom source and background apertures because the field is crowded and TESS imaging is confusion-limited at $b = 2.53075$2 per pixel. Hard Lightkurve quality cuts rejected frames affected by stray light. To isolate short-timescale variability from the secular rise, the light curve was detrended in flux units with a Savitzky–Golay low-pass filter using a window width of 101 points and polynomial degree 5, with iterative $b = 2.53075$3 clipping before and after detrending. Lomb–Scargle periodograms provided the principal period search; piecewise linear detrending and the Deeming power spectrum gave consistent results.

4. Source association and contamination control

A nearby contaminating source posed a nontrivial identification problem. The star ATO J300.1356+34.8776, also ZTF J200032.54+345239.5 and Gaia DR3 2059317182140623872, lies only $b = 2.53075$4 arcsec from the nova and is a known W UMa eclipsing binary. Its orbital period is $b = 2.53075$5 d, with corresponding half-period $b = 2.53075$6 d, and its $b = 2.53075$7-band variability spans 14.921 to 15.203. Because TESS pixels are large, this eclipsing system leaks into the nova aperture (Sokolovsky et al., 10 Jun 2026).

The contamination pattern is diagnostically useful. During the first half of Sector 55, when the nova was still faint and the aperture was dominated by background and confusion, the periodogram showed a clear peak at $b = 2.53075$8 d, matching half the W UMa orbital period. During the second half of Sector 55, the contaminating binary signal remained present, but a new stronger peak at $b = 2.53075$9 d appeared. That period does not match the contaminant.

The most direct localization came from pixel-by-pixel TESS periodograms. These showed one cluster of pixels centered on the nova with best period near $A(V)=9.22$0 d and another cluster offset toward the contaminating binary with best period near $A(V)=9.22$1 d. The power maps therefore displayed two distinct islands, spatially separating the two signals. Temporal behavior supported the same conclusion: the $A(V)=9.22$2 d modulation appeared only when the nova had become bright enough to dominate the aperture and persisted through the remainder of Sector 55.

Additional controls strengthened the attribution. A synthetic $A(V)=9.22$3-day periodic signal injected into the smooth nova trend was recovered cleanly after the same detrending and period-search pipeline, showing that the nova’s secular brightening does not shift a contaminant frequency to $A(V)=9.22$4 d. A nearby check star, TYC 2678-1207-1, showed the contaminating $A(V)=9.22$5 d signal due to leakage from the eclipsing binary, but not the nova-associated $A(V)=9.22$6 d signal. Collectively, the temporal appearance, spatial localization, contamination modeling, and simulation tests identify the $A(V)=9.22$7 d modulation with PGIR22akgylf rather than with the nearby eclipsing binary.

5. Spectroscopic state and ejecta structure during the rise

Optical spectroscopy at $A(V)=9.22$8 d was obtained with DBSP on the Palomar 200-inch telescope, with 600 s total exposure at $A(V)=9.22$9. The spectrum exhibited a very red continuum, strong P Cygni Balmer profiles, He I, N II, and the Paschen series in the near-infrared. The H$4500$0 P Cygni absorption trough indicated a blueshifted velocity of about $4500$1. This spectrum is typical of a nova before peak in the He/N spectral phase (Sokolovsky et al., 10 Jun 2026).

Near-infrared spectroscopy at later times was obtained with TripleSpec on 2022-09-08 ($4500$2 d) and with SpeX on 2022-10-09 ($4500$3 d). Both spectra continued to show P Cygni lines of H I and He I, indicating that the nova remained in the early He/N phase even many weeks after discovery. Fe II lines were expected to emerge closer to peak, as often occurs in slow novae, but no post-peak spectroscopy was available.

The spectroscopic velocity scale is important for physical interpretation. The $4500$4 absorption is much faster than the orbital-scale outflow speeds expected if common-envelope interaction is shaping the envelope. This suggests that PGIR22akgylf hosted at least two kinematically distinct components: a slower, denser envelope responsible for the continuum and the prolonged rise to maximum, and a separate faster outflow responsible for the line absorption. This two-component picture recurs throughout the interpretation of the system.

6. Binary architecture, common-envelope interaction, and theoretical implications

The periodic signal is interpreted as likely tied to the binary orbit, although the data do not uniquely determine whether the observed period is the full orbital period or half of it. If the measured $4500$5 d is $4500$6, then the orbital period is 4.33 h; if it is half the orbital period, then $4500$7 d, or 8.65 h. In either case the system is a tight close binary and therefore requires a dwarf donor rather than a giant companion (Sokolovsky et al., 10 Jun 2026).

For a total system mass of $4500$8, the estimated binary separation is $4500$9 for a 5 h orbit and $J = 14.28$0 for an 8 h orbit. The corresponding orbital-scale outflow speeds expected for common-envelope interaction are $J = 14.28$1 for $J = 14.28$2 d and $J = 14.28$3 for $J = 14.28$4 d. These are much smaller than the observed $J = 14.28$5 H$J = 14.28$6 absorption velocity, reinforcing the inference of a slow orbit-shaped envelope plus a faster wind.

The preferred interpretation is that the detected modulation is not direct light from the binary stars but photometric variability of the nova envelope or photosphere induced by orbital distortion during a common-envelope-like phase. At the epoch of the TESS data the nova was already many magnitudes above quiescence, so the optical light should have been dominated by the expanding nova photosphere. The persistence of a stable orbital-timescale signal under those conditions implies that the continuum-forming region remained dynamically coupled to the binary geometry. The proposed mechanism is that the expanding envelope still engulfed the binary, with orbital motion imprinting a non-axisymmetric density distribution; the photosphere formed within that asymmetric structure, and its rotation with the binary modulated projected area and/or optical depth.

A central quantitative comparison concerns photospheric size versus binary separation. Using an average nova peak absolute magnitude of $J = 14.28$7, a $J = 14.28$8 K blackbody at peak corresponds to a radius of about $J = 14.28$9. During the TESS observations, when the nova was about 5 to 3 magnitudes below peak, the implied blackbody radius would be roughly $t_0 = {\rm JD(UTC)}\ 2459807.6900.$00 to $t_0 = {\rm JD(UTC)}\ 2459807.6900.$01 if the temperature were similar. Compared with orbital separations of $t_0 = {\rm JD(UTC)}\ 2459807.6900.$02 or $t_0 = {\rm JD(UTC)}\ 2459807.6900.$03, the photospheric radius was about 4 to 16 times larger than the binary separation. That ratio is large but not so extreme that few-percent departures from spherical symmetry would be implausible. This suggests that the observed few-percent modulation can be reconciled with an orbit-shaped photosphere.

The slow rise itself argues against a freely expanding high-velocity shell as the continuum source. If the envelope had been ejected at $t_0 = {\rm JD(UTC)}\ 2459807.6900.$04 for 133 days, it would have reached $t_0 = {\rm JD(UTC)}\ 2459807.6900.$05 AU, or $t_0 = {\rm JD(UTC)}\ 2459807.6900.$06, whereas the expected maximum photospheric radius near optical peak is only about $t_0 = {\rm JD(UTC)}\ 2459807.6900.$07, or $t_0 = {\rm JD(UTC)}\ 2459807.6900.$08 AU. The continuum-forming material therefore could not have been a simple ballistic shell at the spectroscopic line velocity. This supports the existence of a slow, dense envelope remaining close to the binary while the nova rose to maximum.

A broader implication is that slow-rise behavior is not confined to symbiotic binaries. Symbiotic systems contain giant donors in wide orbits, whereas the period in PGIR22akgylf implies a close binary with a dwarf donor. PGIR22akgylf therefore demonstrates that a months-long slow rise can occur in a non-symbiotic classical nova. A plausible implication is that slow classical novae and thermonuclear-powered symbiotic eruptions may represent related regimes differentiated in part by orbital scale and the availability of companion interaction to remove the envelope.

The longer-term periodic phenomenology remains nontrivial. A prior AAVSO-based study reported $t_0 = {\rm JD(UTC)}\ 2459807.6900.$09 d with amplitude $t_0 = {\rm JD(UTC)}\ 2459807.6900.$10 mag from $t_0 = {\rm JD(UTC)}\ 2459807.6900.$11-band data spanning $t_0 = {\rm JD(UTC)}\ 2459807.6900.$12 to $t_0 = {\rm JD(UTC)}\ 2459807.6900.$13 d, differing from the TESS period by about 11 minutes. A true orbital-period change of that size is implausible; for isotropic outflow, the relative period change was written as

$t_0 = {\rm JD(UTC)}\ 2459807.6900.$14

and a 4% period change would require mass loss of order $t_0 = {\rm JD(UTC)}\ 2459807.6900.$15, well above a nova envelope mass of $t_0 = {\rm JD(UTC)}\ 2459807.6900.$16 to $t_0 = {\rm JD(UTC)}\ 2459807.6900.$17. The differing photometric periods are therefore interpreted as either evolution in the envelope-generated modulation pattern or methodological and systematic effects in the ground-based analysis, not as a real change in the binary orbit.

The most secure conclusions are observational: PGIR22akgylf is a heavily reddened classical nova, it took about 133 days to reach optical maximum, TESS detected a coherent $t_0 = {\rm JD(UTC)}\ 2459807.6900.$18 d modulation during the rise, the signal is spatially and temporally associated with the nova rather than solely with a nearby contaminant, and the binary must host a dwarf donor. The common-envelope interpretation, the low white-dwarf mass, and the possibility of a weakly explosive or non-impulsive ignition are more tentative. Even so, PGIR22akgylf constitutes a rare case in which slow-rise nova photometry, orbital-timescale modulation, and common-envelope-assisted mass-loss scenarios are linked within a single well-observed eruption.