# Survey of Orion Disks with ALMA (SODA) IV: Millimeter variability of protoplanetary disks

**Authors:** D. Krebs, A. Hacar, J. Forbrich, S. E. van Terwisga  
**arXiv:** [2610.23640](https://arxiv.org/abs/2610.23640)  
**Date:** 2026-10-16  
**Summary page:** https://www.summarizepaper.com/en/arxiv-id/2610.23640v1/

## Key points

- Purpose: Assess short-term millimeter variability in Class II young stellar objects, whose emission is commonly assumed to be stable thermal dust radiation when estimating disk masses.
- Data: ALMA observations of 873 objects in Orion at approximately 1.3 millimeters, with two or three visits per source spanning roughly one hour to three days.
- Method: Separate imaging of observing blocks produced 1,946 snapshots and 377 detections; conservative variability criteria required a maximum-to-minimum flux ratio above 1.2 with high statistical significance.
- Main finding: Twenty sources met the conservative criteria—5.3% of detections and 2.3% of all targets—with flux ratios of approximately 1.23–2.05 and scatter exceeding noise expectations.
- Timescales: Twelve significant changes occurred over approximately one hour, five over one day, and three over three days.
- Extreme case: MGM2012 1041 was detected in one epoch but undetected an hour later, implying a variability factor greater than seven.
- Threshold sensitivity: Relaxed criteria identified 60 candidates, approximately 16% of detections; candidate counts also depended on flux-extraction methods.
- Implications: Short-term millimeter variability could bias disk masses inferred from unresolved, single-epoch observations.
- Possible mechanisms: Rapid variability may involve magnetic reconnection and nonthermal emission, while slower changes may reflect accretion-driven heating; neither explanation is established.
- Limitations: Sparse sampling and unequal timescale coverage prevent a definitive variability census or precise intrinsic occurrence rate; dedicated monitoring is needed.

## Summary

This paper investigates the prevalence of short-term variability in the millimeter continuum emission of Class II young stellar objects. This emission is commonly interpreted as thermal radiation from circumstellar dust and used to estimate disk masses, often under the assumption that the flux remains stable during observations. Although previous studies have detected dramatic millimeter flares, their frequency across large disk populations remains uncertain. The authors revisit the Survey of Orion Disks with ALMA (SODA) to assess whether variability on timescales of approximately one hour to three days can be detected in a large survey not originally designed for time-domain studies. The analysis uses ALMA Band 6 observations of 873 young stellar objects in the LDN 1641 and LDN 1647 clouds of Orion. These observations were obtained in November and December 2019 at frequencies of 224–242 GHz, corresponding to a wavelength of approximately 1.3 millimeters. Each source was observed in two or three execution blocks, with approximately 16 seconds of integration per visit. All sources have observations separated by roughly one hour, while subsets also have measurements separated by about one or three days. Rather than combining all observations, as in the original survey, the authors image each execution block separately, producing 1,946 snapshots. They smooth the images to a common beam for each target and extract fluxes using masks defined from the higher-sensitivity combined images. This conservative extraction procedure yields 377 detections, compared with 502 in the original survey. The principal variability metric, termed the variability factor, is the ratio of the maximum to minimum measured flux. A source is conservatively classified as a variability candidate if this factor exceeds 1.2 and the factor divided by its propagated uncertainty exceeds seven. The authors also examine the standard deviation of individual flux measurements relative to the combined-image flux, comparing it with the scatter expected from observational noise. Calibrator fluctuations are approximately 3% or less, substantially below the changes required by the main selection criteria. Twenty sources satisfy the conservative criteria, representing 5.3% of the 377 detections and 2.3% of all 873 targets. Their variability factors range from approximately 1.23 to 2.05, with normalized flux scatter of roughly 8–30%. All twenty also exhibit scatter exceeding the authors’ noise-based expectations. Twelve significant changes occur over approximately one hour, five over one day, and three over three days. Separately, MGM2012 1041, a source detected in one epoch but undetected an hour later, has a reported lower-limit variability factor greater than seven. Relaxing the thresholds to a variability factor above 1.15 and a factor-to-uncertainty ratio above four yields 60 candidates, or approximately 16% of detections. The authors conclude that short-term millimeter variability may be more common than generally assumed and could bias disk masses inferred from unresolved, single-epoch measurements. Rapid changes may be associated with magnetic reconnection and nonthermal synchrotron or gyrosynchrotron emission, whereas slower changes might involve accretion-driven heating. However, the observations do not establish the underlying physical mechanisms. Important limitations include sparse sampling, with only two or three epochs per source; unequal coverage of different timescales; and sensitivity to flux-extraction methods and selection thresholds. Alternative reductions change the number of candidates, although the twelve strongest conservative candidates remain identifiable across methods. The study therefore provides evidence for variability rather than a definitive census or a precise estimate of its intrinsic occurrence rate. Dedicated monitoring across additional star-forming regions is needed to confirm candidates, characterize their behavior, and assess their impact on disk-mass surveys.