earth2studio.models.dx.DerivedSurfacePressure#

class earth2studio.models.dx.DerivedSurfacePressure(p_levels, surface_geopotential, surface_geopotential_coords, temperature_correction=True, corr_adjustment=(3.4257e-05, 1.5224))[source]#

Interpolates the surface pressure in hPa from pressure level geopotential, surface geopotential and optionally temperature. The calculation is based on linear interpolation of the logarithm of pressure, as well as an optional second-order correction based on temperature and an empirical adjustment.

Note

A more detailed description of the method is available at https://nvidia.github.io/earth2studio/userguide/notes/surface_pressure.md.

Parameters:

  • p_levels (list[int]) – Pressure levels (hPa) that are used as input for the surface pressure interpolation. At least 2 levels are needed. They must be ordered from highest to lowest pressure.

  • temperature_correction (bool, optional, default True) – Whether to apply the second-order correction to surface pressure accounting for temperature.

  • corr_adjustment (tuple[float, float], optional, default (3.4257e-5, 1.5224)) – An empirical adjustment to the temperature correction. The correction is modified as correction = corr_adjustment[0] + correction * corr_adjustment[1]. The default value was derived empirically optimizing the prediction of surface pressure in ERA5 data. The adjustment can be effectively disabled by passing corr_adjustment=(0.0, 1.0).

  • surface_geopotential (Tensor)

  • surface_geopotential_coords (OrderedDict[str, ndarray])

__call__(x, coords)[source]#

Forward pass of diagnostic

Parameters:

  • x (Tensor)

  • coords (OrderedDict[str, ndarray])

Return type:

tuple[Tensor, OrderedDict[str, ndarray]]