Ctrl+K

pysteps.blending.steps.blend_means_sigmas

pysteps.blending.steps.blend_means_sigmas#

pysteps.blending.steps.blend_means_sigmas(means, sigmas, weights)#

Calculate the blended means and sigmas, the normalization parameters needed to recompose the cascade. This procedure uses the weights of the blending of the normalized cascades and follows eq. 32 and 33 in BPS2004.

Parameters:

  • means (array-like) – Array of shape [number_components, scale_level, …] with the mean for each component (NWP, nowcasts, noise).

  • sigmas (array-like) – Array of shape [number_components, scale_level, …] with the standard deviation for each component.

  • weights (array-like) – An array of shape [number_components + 1, scale_level, …] containing the weights to be used in this routine for each component plus noise, scale level, and optionally [y, x] dimensions, obtained by calling either pysteps.blending.steps.calculate_weights_bps() or pysteps.blending.steps.calculate_weights_spn().

Returns:

  • combined_means (array-like) – An array of shape [scale_level, …] containing per scale level (cascade) the weighted combination of means from multiple components (NWP, nowcasts and noise).

  • combined_sigmas (array-like) – An array of shape [scale_level, …] similar to combined_means, but containing the standard deviations.

Contents