If using fit_normalizer=True during variogram estimation, one can't access the fitted normalizer afterwards (in v1.3.0rc2):
bin_center, gamma = gs.vario_estimate((x, y), field, normalizer=gs.normalizer.BoxCox, fit_normalizer=True)
It should return the fitted normalizer as well, when fitting it:
bin_center, gamma, normalizer = gs.vario_estimate((x, y), field, normalizer=gs.normalizer.BoxCox, fit_normalizer=True)
In order to use it afterwards in SRF or Krige.