SnowMIP2

Snow Model Intercomparison Project for forest snow processes

Questions and Answers

Last updated on 24 May, 2006

Drafts of the instructions were sent to the data providers on 21 April 2006 and to the modellers on 3 May 2006 for review. We are replying individually to all comments that are received, but a summary of the responses is given here.

Driving data

  • Surface pressure
  • Precipitation
  • Vegetation parameters

  • Vegetation height and LAI for open sites
  • Seasonal LAI
  • Snow-free albedos and emissivities
  • Root depths
  • Specification of LAI
  • Soil parameters

  • Field capacity at Alptal
  • Soil profile depth
  • Output data

  • Time interval of output data
  • Time averaging of output data
  • Definition of Ecan

  • Driving data

    Q: Should a constant surface atmospheric pressure be used for each site?

    Yes. As we do not have pressure measurements for all of the sites, a standard atmospheric pressure is provided in the instructions for each site. Fluctuations in pressure at a site are much smaller than averages, so the use of a constant value should have little influence on simulations.

    Q: Can models combine the provided rainfall and snowfall rates and repartition the total precipitation using their own algorithms?

    Measured total precipitation has been partitioned into rainfall and snowfall rates using methods suggested by the data providers for each site. We strongly discourage modellers from repartitioning the total precipitation themselves, as this will make intercomparisons between models much harder to interpret. If your model will not accept rainfall and snowfall as separate inputs, please output the rainfall and snowfall rates produced by the model.


    Vegetation parameters

    Q: Can vegetation height and LAI for open sites be provided?

    These are sites with short vegetation submerged by snow, so we expect the vegetation characteristics of the open sites to have little influence on snow simulations. We will, however, include any further information from the data providers in the instructions.

    Q: Can seasonal changes in LAI (e.g. monthly values) be provided?

    We don't have seasonal measurements of LAI, but we expect any variations to be small and have little influence on simulations for coniferous canopies over a winter.

    Q: Can you provide snow-free canopy albedos and emissivities?

    Measured or estimated snow-free albedos for all sites will be added to the instructions. If your model requires emissivity as an input, set it to 1.

    Q: Can you provide root depths?

    We do not have root depth measurements; please estimate them if required.

    Q: LAI appears to be estimated differently at the different sites. Can you specify how it was measured and better yet, reference the technique so that investigators can convert this to whatever definition of LAI they use in their models?

    We are preparing notes on how LAI was measured at each site and guidance on how it should be interpreted.


    Soil parameters

    Q: Can you provide values for field capacity of individual soil layers at Alptal?

    You should set each layer to your model's saturation point for the specified soil texture (or similar). Lacking detailed soil moisture measurements, the data providers recommended this choice from the observation that the soil is wet, with ground water levels at 30 - 60 cm depth.

    Q: Are the modellers free to choose soil profile depth in the simulations?

    Yes.


    Output data

    Q: What is the time interval of output data?

    The output data should be on the same 30-minute time intervals as the input data.

    Q: Should output variables be time averages or instantaneous values?

    This was carefully considered in the PILPS2e intercomparison; for discussion, see
    PILPS2e output variable discussion. In SnowMIP2, prognostic or state variables should be given at the end of each timestep, fluxes should be average values over a timestep, and storage change variables (DelCanHeat, DelCanSWE etc) should be accumulated over a timestep.

    Q: Does Ecan in the water balance components include transpiration (i.e. transpiration + interception evaporation) or does it denote interception evaporation only?

    Ecan should be evaporation of intercepted water only, and Tcan is transpiration.