oggm.diagnostics.compute_geodetic#
- oggm.diagnostics.compute_geodetic(prun, reference_period=None)[source]#
Compares the modelled mass change to Hugonnet et al. (2021).
The model value is computed exactly the way OGGM’s own calibration does it (see run_dynamic_melt_f_calibration): the mass change of the glacier over the reference period, divided by its RGI area and by the length of the period. The regional value is the area weighted mean of these, i.e. the total mass change of the region divided by its total RGI area - the same aggregation Hugonnet et al. use.
Three references are provided:
hug_pergla_*: the per glacier observations, aggregated over exactly the glaciers OGGM modelled. This is the apples-to-apples comparison: same glaciers, same area weights, no extrapolation.
hug_reg_*: the regional averages published by Hugonnet et al., which include an extrapolation to the unmeasured glaciers, and an error estimate. Note that hug_reg_dmdtda_published is relative to the measured area only: hug_reg_dmdtda (derived from dmdt and the full regional area) is the one to compare to. The published regional file is RGI6 based.
the per glacier calibration residuals of the run itself, in
compute_spinup().
- Parameters:
- prunPreproRun
the run to diagnose
- reference_periodstr, optional
the geodetic period to compare over, e.g. ‘2000-01-01_2020-01-01’. Defaults to cfg.PARAMS[‘geodetic_mb_period’].
- Returns:
- a DataFrame indexed by region, with a global row.