MDI active region evolution videos for BARD catalog

Citation and Acknowledgements Please cite both this database the paper describing it, as well as adding the following acknowledgement: "The BARD catalog of BMRs was downloaded from the solar dynamo dataverse (https://dataverse.harvard.edu/dataverse/solardynamo), maintained by Andrés Muñoz-Jaramillo." Database citation format is shown at the top of this page, underneath the database title. It can be downloaded in a variety of formats directly from this page The papers describing this database are S. Chatterjee, A. Munoz-Jaramillo, & D. Lamb. Human-efficient labeling of a solar flux emergence video dataset by a deep learning model, 27 May 2021, PREPRINT (Version 1) available at Research Square https://doi.org/10.21203/rs.3.rs-366908/v1 A. Munoz-Jaramillo, Z. A. Werginz, J. P. Vargas-Acosta, M. D. DeLuca, J. C. Windmueller, J. Zhang, D. W. Longcope, D. A. Lamb, C. E. DeForest, S. Vargas-Dominguez, J. W. Harvey, & P. C. H. Martens , IEEE International Conference on Big Data, 3194 (2016). https://ieeexplore.ieee.org/document/7840975 A. Munoz-Jaramillo, B. Navarrete, & L. A. Campusano , The Astrophysical Journal, Volume 920, Issue 1, id.31, 11 pp. https://ui.adsabs.harvard.edu/abs/2021ApJ...920...31M Main Limitations The main limitations of these data are: Videos are assembled in a constant Carrington rotation frame so, BMRs may drift within the video due to differential rotation. Many videos have gaps and thus, do not have a constant cadence. We addressed this by interpolating the missing frames. In these videos we do not include interpolated frames. Description The Active Region (AR) magnetic evolution videos are created by - Projecting full-disc line-of-sight magnetograms into Carrington grid Dividing the projected maps into 15 degrees x 15 degrees non-overlapping patches. Each patch is sampled into 90 pixels x 90 pixels resulting into pixel scale of (1/6) degree Stacking a set of consecutive frames covering about half a solar rotation tied to each AR from Bipolar Active Region Detection (BARD) catalogue A total of 2453 videos are available over the period 1996 - 2011 in 'fits' format. Within the 'fits' header- 'NAXIS1' represents number of x-pixels 'NAXIS2' represents number of y-pixels 'NAXIS3' represents number of frames 'TIMEXXXX' represents the time stamp for frame XXXX in YYYY-MM-DDTHH:MM:SS format 'CRVAL1' represents the starting Carrington Longitude (degrees) for each video 'CRVAL2' represents the starting Latitude (degrees) for each video Each video is manually verified and a label of '0' or '1' is added to the name of video with '0' representing 'no new AR emergence within the video' and '1' representing 'AR emergence'.