Initialization

[anonymous34]

How should the initial conditions of most state variables be estimated?

[rchlumsk]

How should the initial conditions of most state variables be estimated?

For most parameters, the default values and/or zero values are likely fine. The main exception is soil moisture and groundwater storage, which in almost any simulation should be non-zero initially. The other exception is snow depth, if a model is first run in the winter for a region that typically has snow in the winter. The reason for these exceptions is that although the model will eventually account for even a zero initial soil moisture, the impact can last for a very long time (much longer than a typical spin-up period), and can provide incorrect results in model results.

The typical spin-up period depends on model complexity, although typically one year for most relatively simple models is sufficient.

[anonymous34]

How should the initial conditions for soil capacity be estimated?

[rchlumsk]

How should the initial conditions for soil capacity be estimated?

Initial conditions can be estimated in several ways, depending on the model setup. Generally, the soil storage can be estimated as porosity*soil thickness, with a correction for stone fraction. An estimate of 25% saturation for an October 1st start is likely ‘good enough’ to allow the model to estimate the correct soil moisture within the spin-up period.

The other method is to use Raven to do the work, i.e. inputting the default soil moisture, running the model, and using the model run results as initial conditions for the next run.

In place of other information, the field capacity of soil can be used as the initial moisture condition. Note that for some models, soils are not represented with a physically meaningful porosity, and the storage calculation may slightly different.

[anonymous34]

When is a good time of year to begin a model simulation?

[rchlumsk]

When is a good time of year to begin a model simulation?

The common times to start running a model are at the start of the water year (October 1st in Canada), when soil conditions can be estimated safely as being relatively dry, or in spring, when the soil conditions can be estimated as relatively wet.