Monthly temperature index model calibrated on geodetic MB data#

As of OGGM v1.6, the simplest and standard mass balance model available in OGGM is a monthly temperature index model that can be calibrated on any mass balance product (the default is Hugonnet et al., 2021, see Geodetic MB data).

The monthly mass balance $$B_i$$ at elevation $$z$$ is computed as:

$B_i(z) = P_i^{Solid}(z) - d_f \, max \left( T_i(z) - T_{Melt}, 0 \right)$

where $$P_i^{Solid}$$ is the monthly solid precipitation, $$T_i$$ the monthly temperature and $$T_{Melt}$$ is the monthly mean air temperature above which ice melt is assumed to occur (-1°C per default). Solid precipitation is computed out of the total precipitation. The fraction of solid precipitation is based on the monthly mean temperature: all solid below temp_all_solid (default: 0°C) and all liquid above temp_all_liq (default: 2°C), linear change in between. Total precipitation is obtained from the climate dataset, multiplied by a precipitation correction factor $$P_f$$. The parameter $$d_f$$ indicates the temperature sensitivity of the glacier, and it needs to be calibrated. The model needs to compute the temperature and precipitation at the altitude $$z$$ of the glacier grid points. The default is to use a fixed lapse rate of -6.5K km $$^{-1}$$ and no gradient for precipitation.

Calibration#

New in version 1.6!

A major change from previous OGGM versions, the calibration is now much more flexible, explicit, and adaptable. We recommend all users to spend some time in getting familiar with the calibration procedure in order to adapt it for their own purposes.

Visit the new mass balance calibration tutorial for an overview.

Notes#

Although this mass balance model is a clear improvement to previous OGGM versions (mostly, for better using observations and for getting rid of the residual parameter $$\epsilon$$), more physical approaches are possible. Importantly, we need to take the uncertainty estimates into account, and we need to tackle the issue of daily data for hydrological models.

More exigent users might have a look at PyGEM or OGGM sandbox, which are offering clever ways to deal with these issues.