Coastline-aware interpolation

Scattered observations, analyzed around land.

Ocean-DIVA turns irregular spatial samples into a gridded field and relative posterior error while respecting coastlines, islands, and disconnected basins.

Explore the model Use with xarray

12.0 °C 18.0 °C

What makes it useful

Spatial structure without crossing the coast.

Ocean-DIVA combines a compact Python API with path-aware covariance for marine and bounded-domain analyses.

Start from observations

Pass x/y coordinates, measured values, optional weights, and a correlation length.

Describe the domain

Use contour rings, a water mask, or an xarray land mask whose coordinates define the grid.

Keep uncertainty

Receive the analysis, relative posterior error, and water-domain mask together.

Install

NumPy-first, with optional integrations.

Core

python -m pip install ocean-diva

pandas + xarray

python -m pip install "ocean-diva[accessors]"

Plots

python -m pip install "ocean-diva[plot]"

Interactive playground

See the barrier change the field.

Drag observations, click ocean to add one, or Shift-click near a point to remove it. The browser preview is illustrative; the Python API runs DIVA’s full solver.

10 °C 20 °C · patterned area is land

Parameters reflected in Python

Coordinate-aware masks

Use 1 for land and 0 for ocean.

Pass an xarray DataArray or a single-variable Dataset as land_mask. Its named coordinates are sorted and used as the output grid.

1Create observations with longitude and latitude coordinates.
2Create a binary land mask on the desired output coordinates.
3Analyze, then use the returned water variable for display or export.

xarray land-mask example

import numpy as np
import xarray as xr

temperature = xr.DataArray(
    [12.0, 10.0, 20.0, 18.0],
    dims="observation",
    coords={
        "longitude": ("observation", [0.2, 0.3, 0.7, 0.8]),
        "latitude": ("observation", [0.25, 0.75, 0.75, 0.25]),
    },
    name="temperature",
    attrs={"units": "degree_Celsius"},
)

longitude = np.linspace(0, 1, 51)
latitude = np.linspace(0, 1, 51)
land = xr.DataArray(
    (
        (latitude[:, None] >= 0.20)
        & (latitude[:, None] <= 0.80)
        & (longitude[None, :] >= 0.42)
        & (longitude[None, :] <= 0.58)
    ).astype(np.uint8),
    dims=("latitude", "longitude"),
    coords={"longitude": longitude, "latitude": latitude},
)

result = temperature.diva.analyze(
    x="longitude",
    y="latitude",
    correlation_length=0.2,
    land_mask=land,
)

water_temperature = result["analysis"].where(result["water"])

Result dataset

Three aligned variables.

The xarray accessor preserves coordinate names and source attributes.

`analysis`

The interpolated field on the requested grid, with land set to the exclusion value.

`error`

Relative posterior error, from lower uncertainty near observations toward 1 farther away.

`water`

A boolean domain mask: true in analyzed ocean cells and false on land.