sea_level_response.selrem.selrem_helpers

SELREM helper module.

Functions

bin_loads_to_points(lat_deg, lon_deg, ...[, ...])

Bin loads into coarse lat/lon cells and return one point load per occupied cell (at the cell centre).

build_lonlat(res_deg)

Build a regular global grid: lon = -180 to 180 (periodic), lat = 90 to -90.

compute_mean_mass_balance_over_time_window(ds)

Compute the mean mass balance and, if present, its error over a time range, collapsing the time dimension.

interp_ocean_mask_to(target_lons, ...)

Reproject an ocean mask onto a new longitude/latitude grid using interpolation.

interp_to_grid(field, lons_src, lats_src, ...)

Linearly interpolate a (lon, lat) field to the target grid.

load_ocean_mask()

Load the ocean mask from a NetCDF file.

load_selrem_parameters()

Load SELREM parameters from a JSON file.

preprocess_location_timeseries(...)

Preprocess annual SLR data for a specific location, returning plot-ready time series.

slice_and_prepare_dataset(vmb_dataset_path, ...)

Load, slice, and prepare the VMB dataset for SELREM computation.

wrap180(lon_deg)

Wrap any longitude (e.g. 0-360) to [-180, 180).
bin_loads_to_points(lat_deg: ndarray, lon_deg: ndarray, mass_gt: ndarray, masserr_gt: ndarray, bin_res_deg: float = 1.0) → tuple[ndarray, ndarray, ndarray, ndarray]

Bin loads into coarse lat/lon cells and return one point load per occupied cell (at the cell centre).

Masses are summed. Errors are combined as sqrt(sum(err^2)), i.e. assuming independent errors.

Parameters:

  • lat_deg (np.ndarray) – Latitude of the loads in degrees (any shape; they are flattened).

  • lon_deg (np.ndarray) – Longitude of the loads in degrees (any shape; they are flattened).

  • mass_gt (np.ndarray) – Mass in Gt, same shape as lat/lon.

  • masserr_gt (np.ndarray) – Mass error in Gt, same shape as lat/lon.

  • bin_res_deg (float) – Resolution at which mass-balance loads are binned to point loads.

Returns:

lat_centres, lon_centres (degrees), mass (Gt), mass error (Gt) of the occupied bins.

Return type:

tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray]

build_lonlat(res_deg: float) → tuple[ndarray, ndarray, ndarray, ndarray]

Build a regular global grid: lon = -180 to 180 (periodic), lat = 90 to -90.

Parameters:

res_deg (float) – Grid resolution in decimal degrees

Returns:

(Longitude, Latitude, 2D Longitude, 2D Latitude)

Return type:

tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray]

compute_mean_mass_balance_over_time_window(ds: Dataset, start_time: datetime | None = None, end_time: datetime | None = None) → Dataset

Compute the mean mass balance and, if present, its error over a time range, collapsing the time dimension.

Parameters:

  • ds (xr.Dataset) – Input mass balance dataset with a time dimension. The mass-balance error variable is optional.

  • start_time (datetime | None) – Start time for slicing, by default None. If both start_time and end_time are provided, the dataset is sliced accordingly before computing the mean. Otherwise, the entire time range is used.

  • end_time (datetime | None) – End time for slicing, by default None. If both start_time and end_time are provided, the dataset is sliced accordingly before computing the mean. Otherwise, the entire time range is used.

Returns:

Dataset with the same spatial dimensions (‘point’, ‘x’, ‘y’), where the ‘time’ dimension has been removed and each variable represents the mean value across the specified years. The error variable is only included if it was present in the input.

Return type:

xr.Dataset

interp_ocean_mask_to(target_lons: ndarray, target_lats: ndarray, ocean_mask: ndarray) → ndarray

Reproject an ocean mask onto a new longitude/latitude grid using interpolation.

Parameters:

  • target_lons (np.ndarray) – Target longitudes for the new grid

  • target_lats (np.ndarray) – Target latitudes for the new grid

  • ocean_mask (np.ndarray) – Spatial grid masking ocean cells

Returns:

The reprojected ocean mask on the target grid.

Return type:

np.ndarray

interp_to_grid(field: ndarray, lons_src: ndarray, lats_src: ndarray, lon2d_t: ndarray, lat2d_t: ndarray) → ndarray

Linearly interpolate a (lon, lat) field to the target grid.

Points outside the source grid become NaN.

Parameters:

  • field (np.ndarray) – 2D field on the source grid, shape (nlon, nlat)

  • lons_src (np.ndarray) – Longitudes of the source grid in decimal degrees

  • lats_src (np.ndarray) – Latitudes of the source grid in decimal degrees

  • lon2d_t (np.ndarray) – 2D longitudes of the target grid in decimal degrees

  • lat2d_t (np.ndarray) – 2D latitudes of the target grid in decimal degrees

Returns:

The field interpolated onto the target grid, same shape as lon2d_t.

Return type:

np.ndarray

load_ocean_mask() → tuple[ndarray, ndarray, ndarray]

Load the ocean mask from a NetCDF file.

Returns:

A tuple containing the ocean mask array, longitude array, and latitude array.

Return type:

tuple[np.ndarray, np.ndarray, np.ndarray]

load_selrem_parameters() → tuple[ndarray, ndarray, ndarray, ndarray]

Load SELREM parameters from a JSON file.

Returns:

gdot_fit, udot_fit, gdot_unc, udot_unc arrays.

Return type:

tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray]

preprocess_location_timeseries(annual_slr_ds: Dataset, lat0: float, lon0: float) → dict

Preprocess annual SLR data for a specific location, returning plot-ready time series.

This function extracts time series data at the nearest grid point to the specified coordinates, computes cumulative sums, calculates linear trends, and returns data ready for plotting.

Parameters:

  • annual_slr_ds (xr.Dataset) – Annual SLR dataset containing ‘ndot’, ‘udot’, ‘sdot’ and their uncertainties.

  • lat0 (float) – Target latitude in decimal degrees (-90 to 90).

  • lon0 (float) – Target longitude in decimal degrees (-180 to 180).

Returns:

Dictionary containing: - ‘years’: List of year labels (integers) - ‘sdot_cumulative’: Cumulative relative SLR (mm) - ‘ndot_cumulative’: Cumulative absolute SLR (mm) - ‘udot_cumulative’: Cumulative vertical deformation (mm) - ‘sig_sdot_cumulative’: Cumulative uncertainty for sdot (mm) - ‘sig_ndot_cumulative’: Cumulative uncertainty for ndot (mm) - ‘sig_udot_cumulative’: Cumulative uncertainty for udot (mm) - ‘slope_sdot’: Linear trend for relative SLR (mm/yr) - ‘slope_ndot’: Linear trend for absolute SLR (mm/yr) - ‘slope_udot’: Linear trend for vertical deformation (mm/yr) - ‘actual_lat’: Actual latitude of nearest grid point - ‘actual_lon’: Actual longitude of nearest grid point

Return type:

dict

slice_and_prepare_dataset(vmb_dataset_path: Path | str, start_time: datetime, end_time: datetime) → Dataset

Load, slice, and prepare the VMB dataset for SELREM computation.

Parameters:

  • vmb_dataset_path (Path | str) – Path to the VMB GeoZarr dataset.

  • start_time (datetime) – Start time for slicing.

  • end_time (datetime) – End time for slicing.

Returns:

Sliced and prepared VMB dataset.

Return type:

xr.Dataset

Raises:

ValueError – If the dataset does not contain a ‘time’ dimension or if the time range is invalid

wrap180(lon_deg: ndarray | float) → ndarray | float

Wrap any longitude (e.g. 0-360) to [-180, 180).

Parameters:

lon_deg (np.ndarray | float) – Longitude in decimal degrees

Returns:

Longitude in decimal degrees wrapped to [-180, 180)

Return type:

np.ndarray | float