Data for "Precovery Observations of 3I/ATLAS from TESS Suggests Possible Distant Activity"

Here, we present the data used in the AAS Journals manuscript, "Precovery Observations of 3I/ATLAS from TESS Suggests Possible Distant Activity" (accepted on August 19, 2025). 

In the data directory, we present three data sets within this repository:

1. The calibrated FFIs used in one of the manuscript figures (tess2025*.fits).
2. The location of 3I/ATLAS for the duration of the TESS observations in both (RA, Dec) and (x, y) pixel locations on the FFI (3I_pixel_locations*.csv).
3. The secular light curve data for 3I/ATLAS
4. The average value of the calibrated FFIs for the duration of the TESS observations (avg*.npy).
5. The shift-stacked cutouts around 896 Sphinx (stacked_A918PE_2-3_v3.npy).
6. The shift-stacked cutouts around 3I/ATLAS in Camera 2 CCD 3 (stacked_3I_2-3_v4.npy).
7. The shift-stacked cutouts around 3I/ATLAS in Camera 1 CCD 3 (stacked_3I_1-2_v4.npy).

We note that files which contain `2-3` are for TESS Camera 2 CCD 3 and `1-2` are for TESS Camera 1 CCD 2. If you use the secular light curve data, please also cite Ye et al. (2020) and Seligman et al. (2025).

Each stacked*.npy file contains a dictionary with the following information:

1. `raw` - the calibrated FFI cutouts shift-stacked on the object without background correction.
2. `subtracted` - the calibrated FFI cutouts shift-stacked on the object with background correction.
3. `error` - the calibrated FFI errors affiliated with the images.
4. `good_frames` - a quality flag indicating which timesteps are crowded versus not crowded. It is recommended to only anlyze frames that are not crowded (`good_frames == 0`).
5. `time` - the time array affiliated with the observations in units of MJD.
6. `loc` - the location of the object in the image. This will be the central pixel in all cases.

The files can be read using the following:

filename = ''

data = np.load(filename, allow_pickle=True).item()

Additionally, in the scripts directory, we present the Python scripts used to generate the figures within the manuscript. The files are named by the figure they generate. One should be able to generate the figures using all of the data presented in this repository.

The GitHub repository affiliated with this manuscript can be found at https://github.com/afeinstein20/atlas-tess. The manuscript of this work can be found on the arXiv at https://arxiv.org/abs/2507.21967. 

The calibrated TESS FFis are available for bulk download on the Mikulski Archive for Space Telescopes (doi:10.17909/5qpe-ek46).