DETACH Coronal Hole Detection and Tracking Database (2010-2025)

Overview This repository contains a comprehensive, long-term statistical database of solar Coronal Holes (CHs) spanning 1.5 solar cycles (May 2010 to December 2025). The dataset was constructed with a 12-hour temporal resolution using observations from the Solar Dynamics Observatory (SDO). The CH boundaries and their associated physical parameters were automatically detected, extracted, and continuously tracked using the Detection and Tracking Algorithm for Coronal Holes (DETACH). Multi-wavelength data were utilized, specifically SDO/AIA 193 Å for geometric and radiative properties, and SDO/HMI for magnetic properties. This dataset is the official data release for the paper: Liu, J., Shen, C., et al. (2026). Statistical Analysis of Coronal Hole Spatiotemporal Evolution, Kinematics, and Lifetimes over 1.5 Solar Cycles. The Astrophysical Journal. Data Structure To accurately capture both instantaneous states and long-term evolution, the database is structured into two hierarchical levels: Snapshot Level: Stores the complete set of physical parameters for each individual CH detection at a specific 12-hour time frame. Trace Level: Aggregates all snapshots belonging to a single, continuously tracked CH. It contains lifecycle-defining properties (e.g., start time, end time, total lifetime) as well as statistical measures of the CH's evolution (temporal mean, standard deviation, and relative errors of the physical parameters). Data Dictionary Below is the detailed list of physical parameters extracted for each Coronal Hole. Note that to mitigate solar projection effects, all geometric parameters were rigorously calculated on an idealized solar sphere (radius = r_\odot). 1. Geometric Parameters Parameter Description Unit area_Mm2 Total area of the CH on the solar sphere \text{Mm}^2 area_ratio Area ratio on the solar sphere --- area_positive Area of positive magnetic polarity pixels \text{Mm}^2 area_negative Area of negative magnetic polarity pixels \text{Mm}^2 center_hgs_lat / center_hgs_lon Center latitude/longitude in Stonyhurst (HGS) coordinates deg center_hgc_lat / center_hgc_lon Center latitude/longitude in Carrington (HGC) coordinates deg boundary_hgs_lat_min / boundary_hgs_lat_max Minimum and maximum latitude in HGS deg boundary_hgs_lon_min / boundary_hgs_lon_max Minimum and maximum longitude in HGS deg 2. Magnetic Parameters Parameter Description Unit flux Signed magnetic flux Wb flux_positive / flux_negative Positive / Negative magnetic flux Wb flux_percent_unbalanced Unbalanced flux percentage % B_average Signed average magnetic field strength G B_abs_average Unsigned (absolute) average magnetic field strength G B_pos_avg / B_neg_avg Average magnetic field strength of positive/negative polarity pixels G unipolarity_degree Degree of unipolarity (U), where U=0 is perfectly unipolar --- 3. Radiative Parameters Parameter Description Unit average_aia_193_intensity Average EUV intensity inside the CH (SDO/AIA 193 Å) DN/s 4. Kinematic Parameters (Available predominantly at the Trace level) Parameter Description Unit hgs_longitude_speed Longitudinal rotation velocity deg/day hgs_latitude_speed Latitudinal migration velocity deg/day Usage The dataset includes a SQLite database file (detach.db) and a Jupyter Notebook (query_example.ipynb) demonstrating how to query, extract, and visualize the data. Prerequisites To run the example notebook, you will need Python installed along with the following standard scientific and solar physics packages: pip install sqlalchemy numpy matplotlib astropy sunpy pycocotools Example Notebook Highlights The provided query_example.ipynb offers practical code snippets for: Database Connection & ORM Definition: Using sqlalchemy to map the Trace and Instance tables and connect to the SQLite database. Data Filtering: Querying specific Coronal Hole instances within a defined time frame and spatial bounding box. Mask Decoding & Plotting: Decoding the RLE-compressed binary masks (mask, mask_hgc) using pycocotools, and overlaying the Coronal Hole contours onto a solar map using sunpy and matplotlib. Synoptic Map Generation: Aggregating masks over a specific Carrington Rotation to produce a CH synoptic map. To get started, simply place detach.db and query_example.ipynb in the same directory, launch Jupyter Notebook, and run the cells. Citation If you use this dataset in your research, please cite the following publication: Liu, J., Shen, C., et al. (2026). Statistical Analysis of Coronal Hole Spatiotemporal Evolution, Kinematics, and Lifetimes over 1.5 Solar Cycles. The Astrophysical Journal. Related Resources DETACH Algorithm: The DETACH algorithm used to generate this database is detailed in the paper: Liu, J., Shen, C., Pan, Y. et al. DETACH: Detection and Tracking Algorithm for Coronal Holes. Sol Phys 301, 17 (2026). DOI: 10.1007/s11207-026-02609-1 Source Code: The source code for the DETACH algorithm is open-source and available on GitHub at Junyan-L/DETACH. Contact For any questions regarding the dataset or the DETACH algorithm, please contact Junyan Liu ([e-mail]) or Chenglong Shen ([e-mail]).