Module `climate`

Brief summary

The climate module is a unified dispatcher for atmospheric forcing. It delegates to one of four implementations selected by the method parameter: simple, oggm, glacialindex, or station, under a single entry point with a consistent interface.

State variables

Reads: t, usurf

Writes: air_temp, air_temp_sd, precipitation

Choosing a method

Set processes.climate.method in your configuration:

processes:
  climate:
    method: simple   # or: oggm | glacialindex | station

Method: `simple`

A lightweight station-type forcing that computes spatially distributed monthly temperature and precipitation fields from a reference value and lapse rates. Suitable for idealized experiments or when neither OGGM data nor full climate files are available.

At each update the module:

Distributes temperature from a reference elevation using a constant lapse rate.
Optionally adds a sinusoidal seasonal cycle.
Distributes precipitation from a reference station value using an altitude-dependent lapse rate.
Generates 12-month fields for compatibility with SMB modules that expect monthly forcing.

Method: `oggm`

Processes monthly historical climate data from the GSWP3-W5E5 dataset obtained via the oggm_shop module. It generates monthly 2D raster fields for corrected precipitation, mean temperature, and temperature variability. The module applies a multiplicative correction factor to precipitation (prcp_fac) and a bias correction to temperature (temp_bias). Temperature is extrapolated across the glacier surface using a reference height and a constant lapse rate.

The module also supports generating climate data beyond the observational time frame by defining a reference period (ref_period) to randomly select years within it, then applying a user-defined temperature bias and precipitation scaling over time:

time   delta_temp  prec_scal
1900          0.0        1.0
2020          0.0        1.0
2100          4.0        1.0

If clim_trend_array is an empty list, the module reads from the file specified by file.

Method: `glacialindex`

Loads two climate snapshots corresponding to two end-member states and interpolates them using a climate signal and a glacial index (Jouvet et al., 2023)¹. Suitable for palaeo-glacier modelling.

A function GI(\(t\)) maps time to a scalar between 0 and 1, with GI=0 corresponding to nearly ice-free conditions (\(\mathrm{CL}_0\)) and GI=1 to maximum glaciation (\(\mathrm{CL}_1\)). Each climate state bundles mean temperature, temperature variability, mean precipitation, and lapse rate, \(\mathrm{CL}=(T^{\rm mean},T^{\rm std},P^{\rm mean},\mathrm{LR})\), and the climate at time \(t\) is the linear combination:

\[\mathrm{CL}(t) = \mathrm{GI}(t)\times \mathrm{CL}_1 + (1 - \mathrm{GI}(t))\times \mathrm{CL}_0.\]

The GI function is built by rescaling a climate proxy signal — e.g. the EPICA Antarctic temperature anomaly, available for the last 800,000 years — so that it is close to 1 at the ice maximum and 0 at the ice minimum. Because the two states are defined on different reference topographies, temperature is corrected for the elevation difference between the modelled ice surface and each reference surface using a vertical lapse rate.

Method: `station`

Generates spatially distributed temperature and precipitation fields from simplified climate parameters representing a weather-station-type forcing. Designed for mountain glacier simulations where climate is characterised by a local temperature lapse rate and a precipitation-altitude relationship.

At each update the module:

Computes a reference temperature field from the zero-degree isotherm elevation and the adiabatic lapse rate.
Adds an optional sinusoidal seasonal cycle (cosine approximation).
Computes a precipitation field using an altitude-dependent lapse rate relative to a reference station value.
Applies user-specified time-dependent temperature and precipitation offsets (climate_change_array).
Adjusts both fields for changes in ice surface elevation relative to the initial surface.

Parameters

Default configuration file (climate.yaml):

climate:
  # Choose the climate implementation: simple | oggm | glacialindex | station
  method: simple

  # ------------------------------------------------------------------
  # simple — constant temperature and precipitation at a reference
  # elevation, adjusted to state.usurf via vertical lapse rates.
  # Produces 12-month fields for compatibility with SMB modules.
  # ------------------------------------------------------------------
  simple:
    update_freq: 1.0
    temp: 0.0                          # mean annual temperature at ref_elevation [°C]
    prec: 1000.0                       # mean annual precipitation at ref_elevation [kg m-2 yr-1]
    temp_std: 5.0                      # temperature std deviation [°C]
    ref_elevation: 0.0                 # reference elevation [m a.s.l.]
    temp_lapse_rate: -0.0065           # temperature lapse rate [°C per m]
    prec_lapse_rate: 0.0               # precipitation lapse rate [fraction per m]
    cosine_yearly_cycle_temp: False    # add a cosine seasonal cycle to temperature
    cosine_yearly_cycle_amplitude: 5.0 # amplitude of the cosine cycle [°C]
    southern_hemisphere_climate: False # shift the cosine cycle by 6 months

  # ------------------------------------------------------------------
  # oggm — historical monthly temperature & precipitation from an OGGM
  # climate_historical.nc (per-RGI). Mirrors the legacy `clim_oggm`.
  # ------------------------------------------------------------------
  oggm:
    update_freq: 1.0
    file: file.txt
    clim_trend_array:
      - ["time", "delta_temp", "prec_scal"]
      - [1900, 0.0, 1.0]
      - [2020, 0.0, 1.0]
    ref_period: [1960, 1990]
    seed_par: 123

  # ------------------------------------------------------------------
  # glacialindex — two-end-member climate interpolated by a transient
  # signal. Mirrors the legacy `clim_glacialindex`.
  # ------------------------------------------------------------------
  glacialindex:
    update_freq: 100.0
    climate_0_file: data/climate.nc
    climate_1_file: data/climate1.nc
    signal_file: data/GI.dat
    vertical_lapse_rate_0: 6.0
    vertical_lapse_rate_1: 5.74
    temporal_resampling: 12

  # ------------------------------------------------------------------
  # station — synthetic climate built from a reference precipitation
  # station + adiabatic lapse rate + optional time-series of offsets.
  # Mirrors the legacy `clim_station`.
  # ------------------------------------------------------------------
  station:
    update_freq: 1.0
    zero_degree_isotherm: 2500.0
    adiabatic_lapse_rate: 0.0058
    air_temperature_stdev: 2.0
    precipitation_lapse_rate: 3.0
    reference_precipitation_elevation: 1000.0
    reference_precipitation: 800.0
    min_precipitation_allowed: 1.0
    cosine_yearly_cycle_temp: True
    cosine_yearly_cycle_amplitude: 5.0
    southern_hemisphere_climate: False
    export_climate_ref: True
    time_dependent_climate: True
    climate_change_array:
      - ["time", "temperature_offset", "precip_offset"]
      - [1750, 0.0, 100]
      - [1800, 0.0, 100]
      - [1850, 0.0, 100]
      - [1900, 0.0, 100]
      - [1950, 0.0, 100]
      - [2000, 0.5, 100]
      - [2050, 1.5, 100]
      - [2100, 2.0, 100]

Structure of the parameters:

climate
├── method
├── simple
│   └── ...
├── oggm
│   └── ...
├── glacialindex
│   └── ...
└── station
    └── ...

Description of the parameters:

Name	Description	Default value	Units
`method`	Climate implementation to use: `simple`, `oggm`, `glacialindex`, or `station`.	simple	—

simple

Name	Description	Default value	Units
`simple.update_freq`	Update climate fields every X years.	1.0	yr
`simple.temp`	Mean annual temperature at ref_elevation.	0.0	°C
`simple.prec`	Mean annual precipitation at ref_elevation.	1000.0	kg m\( ^{-2} \) yr\( ^{-1} \)
`simple.temp_std`	Temperature standard deviation used for PDD computation.	5.0	°C
`simple.ref_elevation`	Reference elevation for temperature and precipitation.	0.0	m
`simple.temp_lapse_rate`	Vertical temperature lapse rate.	-0.0065	°C m\( ^{-1} \)
`simple.prec_lapse_rate`	Fractional change in precipitation per metre of elevation.	0.0	m\( ^{-1} \)
`simple.cosine_yearly_cycle_temp`	Add a cosine seasonal cycle to temperature.	False	—
`simple.cosine_yearly_cycle_amplitude`	Amplitude of the cosine seasonal temperature cycle.	5.0	°C
`simple.southern_hemisphere_climate`	Shift the seasonal cycle by 6 months for southern-hemisphere glaciers.	False	—

oggm

Name	Description	Default value	Units
`oggm.update_freq`	Update climate fields every X years.	1.0	yr
`oggm.file`	Path to the OGGM climate_historical.nc file.	file.txt	—
`oggm.clim_trend_array`	Time-dependent temperature and precipitation scaling table: rows of [time, delta_temp (°C), prec_scal (-)].	[['time', 'delta_temp', 'prec_scal'], [1900, 0.0, 1.0], [2020, 0.0, 1.0]]	—
`oggm.ref_period`	Reference period [start, end] for the OGGM climate baseline.	[1960, 1990]	yr
`oggm.seed_par`	Random seed for stochastic perturbations.	123	—

glacialindex

Name	Description	Default value	Units
`glacialindex.update_freq`	Update climate fields every X years.	100.0	yr
`glacialindex.climate_0_file`	NetCDF file for the modern (warm) end-member climate.	data/climate.nc	—
`glacialindex.climate_1_file`	NetCDF file for the glacial (cold) end-member climate.	data/climate1.nc	—
`glacialindex.signal_file`	File containing the glacial index signal (time, index) used to interpolate between end members.	data/GI.dat	—
`glacialindex.vertical_lapse_rate_0`	Vertical lapse rate applied to the modern end-member temperature field.	6.0	°C (100 m)\( ^{-1} \)
`glacialindex.vertical_lapse_rate_1`	Vertical lapse rate applied to the glacial end-member temperature field.	5.74	°C (100 m)\( ^{-1} \)
`glacialindex.temporal_resampling`	Number of sub-steps per model time step for temporal resampling of the glacial index signal.	12	—

station

Name	Description	Default value	Units
`station.update_freq`	Update climate fields every X years.	1.0	yr
`station.zero_degree_isotherm`	Elevation at which mean annual temperature is 0 °C.	2500.0	m
`station.adiabatic_lapse_rate`	Adiabatic lapse rate for air temperature.	0.0058	°C m\( ^{-1} \)
`station.air_temperature_stdev`	Space-independent standard deviation of daily temperature variability.	2.0	°C
`station.precipitation_lapse_rate`	Change in precipitation with altitude as a fraction of reference_precipitation.	3.0	% per 100 m
`station.reference_precipitation_elevation`	Elevation of the reference precipitation station.	1000.0	m
`station.reference_precipitation`	Mean cumulative annual precipitation at the reference station.	800.0	mm yr\( ^{-1} \)
`station.min_precipitation_allowed`	Minimum allowed precipitation to avoid unrealistically low values.	1.0	mm yr\( ^{-1} \)
`station.cosine_yearly_cycle_temp`	Add an idealised cosine seasonal temperature signal.	True	—
`station.cosine_yearly_cycle_amplitude`	Amplitude of the cosine seasonal temperature variation.	5.0	°C
`station.southern_hemisphere_climate`	Shift the seasonal cycle by 6 months for southern-hemisphere glaciers.	False	—
`station.export_climate_ref`	Export a climate_ref.nc file for the original topography.	True	—
`station.time_dependent_climate`	Allow climate to vary in time via climate_change_array.	True	—
`station.climate_change_array`	Time-dependent offsets: rows of [time, temperature_offset (°C), precip_offset (mm yr^{-1})].	[['time', 'temperature_offset', 'precip_offset'], [1750, 0.0, 100], [1800, 0.0, 100], [1850, 0.0, 100], [1900, 0.0, 100], [1950, 0.0, 100], [2000, 0.5, 100], [2050, 1.5, 100], [2100, 2.0, 100]]	—

Contributors: Guillaume Jouvet, Tancrède Leger.

Jouvet, G., Cohen, D., Russo, E., Buzan, J., Raible, C. C., Haeberli, W., Kamleitner, S., Ivy-Ochs, S., Imhof, M. A., Becker, J. K., Landgraf, A., & Fischer, U. H. (2023). Coupled climate-glacier modelling of the last glaciation in the alps. Journal of Glaciology, 69(278), 1956--1970. https://doi.org/10.1017/jog.2023.74 ↩

Module climate

Choosing a method

Method: simple

Method: oggm

Method: glacialindex

Method: station