History#

0.6.0 (unreleased)#

  • Added aidas.model.detect_velocity_waves(), which detects radial velocity wave signatures by differencing two consecutive radar volumes following Miller et al. (2022, https://doi.org/10.5194/amt-15-1689-2022). The negative half of the velocity change is thresholded, regridded onto a 0.5 km Cartesian grid and area filtered into a binary mask.

  • Added aidas.vis.visualize_velocity_waves(), which draws the mask in a panel of its own beside the velocity it was detected from. That velocity is the dealiased, quality controlled sweep, kept on the RadarImage() as wave_sweep_radar, rather than the raw folded field in the original volume. aidas.vis.plot_velocity_wave_mask() draws just the mask, for comparing detections made with different settings side by side.

  • Added aidas.io.get_previous_scan(), which fetches the volume collected immediately before a given scan. The S3 listing behind it now skips the _MDM metadata files, which share a timestamp with their volume but hold no radar data, and recognises the older _V03.gz volume names.

  • aidas.util.azimuth_point(), aidas.util.azimuth_from_ellipse() and the lidar triggering functions take a mask argument, so an instrument can be cued at a detected wave train as well as at the lake breeze. The default is unchanged.

  • Fixed aidas.io.preprocess_radar_image() overwriting its own radar and bucket_name arguments, which made it fetch KLOT from the Unidata bucket whatever the caller asked for.

  • Renamed the project to the Argonne AI-Driven Adaptive Sampling System (AIDAS). The import package is now aidas (previously adam) and the distribution is now aidas-atmos on PyPI (previously adam-atmos).

  • adam-atmos is deprecated. A final transitional release forwards import adam to aidas with a DeprecationWarning; see compat/adam-atmos/.

  • Replaced the “digital twin” framing in the project description with adaptive sampling / instrument tasking, which describes what the system actually does.

  • Added aidas.util.aeqd_to_lonlat() and compute radar gate geolocation directly rather than delegating to PROJ, fixing a unit test regression across pyproj versions.

0.5.0 (2026-03-19)#

  • Added aidas.util.azimuth_from_ellipse() for deriving the instrument pointing direction from an ellipse fitted to the lake breeze mask.

  • Added unit tests covering the new azimuth utilities.

  • Expanded the notebook documentation.

0.4.0 (2026-02-09)#

  • Added the aidas.triggering module with adaptive scanning support for Halo Photonics lidars.

  • Added the aidas.testing module, including a reference test dataset and fake SSH/SFTP clients for exercising scan triggering without hardware.

  • Added docstrings and documented references for the testing and triggering modules.

  • Allowed the user to specify their own S3 bucket when fetching radar data, and changed the default bucket.

  • Fixed handling of radar_object when passed as either a list or an array.

0.3.0 (2025-09-15)#

  • Added the Sphinx-Gallery example gallery.

  • Fixed the instrument pointing feature.

0.2.0 (2025-09-08)#

  • Added aidas.util.azimuth_point() to determine the optimal instrument pointing direction for adaptive scanning.

  • Added unit tests for instrument pointing.

  • Published the documentation to GitHub Pages.

0.1.1 (2025-07-14)#

  • Renamed the distribution to adam-atmos on PyPI.

0.1.0 (2025-07-14)#

  • First release on PyPI.