earth2studio.models.dx.TCTrackerVitart#
- class earth2studio.models.dx.TCTrackerVitart(vorticity_threshold=3.5e-05, mslp_threshold=99000.0, temp_dec_threshold=0.5, lat_threshold=60.0, exclude_border=True, path_search_distance=300, path_search_window_size=3)[source]#
Finds a list of tropical cyclone centers using the conditions in Vitart 1997
Note
For more information about this method see:
- Parameters:
vorticity_threshold (float, optional) – The threshold for vorticity at 850, below which a possible tropical cyclone center is rejected, by default 3.5e-5 1/s
mslp_threshold (float, optional) – The threshold for minimum sea level pressure for local minimums to be considered tropical cyclone, by default 99000 Pa
temp_dec_threshold (float, optional) – The value for which average temperature must decrease away from the warm core for a possible center to be considered a tropical cyclone, by default 0.5 degrees celsius
lat_threshold (float, optional) – The maximum absolute latitude that a point will be considered to be a tropical cyclone, by default 60 degrees (N and S).
exclude_border (bool | int, optional) – If positive integer, exclude_border excludes peaks from within exclude_border-pixels of the border of the image. If tuple of non-negative ints, the length of the tuple must match the input array dimensionality. Each element of the tuple will exclude peaks from within exclude_border-pixels of the border of the image along that dimension. If True, takes the min_distance parameter as value. If zero or False, peaks are identified regardless of their distance from the border.
path_search_distance (int, optional) – The max radial distance two cyclone centers will be considered part of the same path in km, by default 300
path_search_window_size (int, optional) – The historical window size used when creating TC paths, by default 2
Examples
The cyclone tracker will return a tensor of TC paths collected over a series of forward passes which are held inside of the models state. Namely given a time series of n snap shots, the tracker should be called for each time-step resulting in a tensor consisting of a set number of paths with n steps. Any non-valid / missing data will be torch.nan for filtering in post processing steps.
>>> model = TCTrackerVitart() >>> # Process each timestep >>> for time in [datetime(2017, 8, 25) + timedelta(hours=6 * i) for i in range(3)]: ... da = data_source(time, tracker.input_coords()["variable"]) ... input, input_coords = prep_data_array(da, device=device) ... output, output_coords = model(input, input_coords) >>> # Final path_buffer shape: [batch, path_id, steps, variable] >>> output.shape # torch.Size([1, 6, 3, 4]) >>> model.path_buffer.shape # torch.Size([1, 6, 3, 4]) >>> # Remove current paths from models state >>> model.reset_path_buffer() >>> model.path_buffer.shape # torch.Size([0])
