Snow reservoir

In Rameau, the snow reservoir produces snow melting \(q_{sm}\) (mm) and snow runoff \(q_{sr}\) (mm) according to the values of rainfall \(R\) (mm), PET \(E_P\) (mm), temperature \(T\) (°C) and, if available, snow \(R_S\) (mm). \(q_{sm}\) and \(q_{sr}\) are then added to the effective rainfall produced by the soil reservoir.

The SWE \(H_n\) (mm) determines how much water the snowpack in the reservoir contains. The liquid water retention \(r\) (%) corresponds to the fraction of \(H_n\) bounded by capillarity to the snowpack.

During a time step, snow melting occurs according to a succession of physical processes:

  1. Snow melting by soil calories \(M_s\) (mm),

  2. Snow melting using a degree day model \(M_d\) (mm),

  3. If \(R > E_P\), snow melting by rainfall calories \(M_r\) (mm).

If \(R < E_P\), PET may not be fully satisfied by rainfall and the rest may be taken from the snow by sublimation, leading to the calculation of an AET from the snow \(E_S\) (mm) and an UPET \(E_U\).

\(M_s\) participates in feeding the soil reservoir while \(M_d + M_r\) are partitioned between storage in the snow pack, \(q_{sm}\) and \(q_{sr}\).

Activate snow

To activate snow in the model, the user needs to provide at least a temperature input data file.

[files]
temperature = "temperature.csv"

In this case, snow is produced if \(T < 0\). Alternatively, a snow input data file can directly been provided in addition to the temperature data file.

[files]
temperature = "temperature.csv"
snow = "snow.csv"

Physical parameters

Temperature correction

Additive temperature correction \(T_c\) (°C). Before proceeding with snow melting processes, \(T\) is potentially corrected by \(T_c\) in order to correct the possible discrepancies affecting the model input temperature:

\[T = T + T_c.\]

Should be optimised. Default value is 0 °C.

[watershed.1]
snow.correction.temperature = { value = 0.0, lower = -3.0, upper = 3.0, opti = false, sameas = 0 }

Snow melting by soil calories

Amount of SWE likely to melt under the action of the calories released by the soil \(C_s\) (1/10 \(\mathrm{mm.day^{-1}}\)).

Should be optimised. Default value is 5 1/10 \(\mathrm{mm.day^{-1}}\).

[watershed.1]
snow.melting = { value = 5.0, lower = 0.001, upper = 20.0, opti = false, sameas = 0 }

Degree day model coefficient

Degree day coefficient of the degree day model \(C_d\) (\(\mathrm{mm.°C^{-1}.day^{-1}}\)).

Should be optimised. Default value is 4 \(\mathrm{mm.°C^{-1}.day^{-1}}\).

[watershed.1]
snow.degree_day.coefficient = { value = 4.0, lower = 0.001, upper = 7.0, opti = false, sameas = 0 }

Degree day model temperature threshold

Temperature threshold of the degree day model \(T_d\) (°C).

Should be optimised. Default value is 0 °C.

[watershed.1]
snow.degree_day.temperature = { value = 0.0, lower = -2.0, upper = 2.0, opti = false, sameas = 0 }

Correction factor of snow melting by rain

Correction factor \(F_m\) (%) applied to the amount of snow that will melt under the action of rainfall.

Should not be optimised in most cases. Default value is 0%.

[watershed.1]
snow.correction.rainfall = { value = 0.0, lower = -20, upper = 20.0, opti = false, sameas = 0 }

Correction factor of PET

Correction factor \(F_s\) (%) applied to \(E_P\) for calculating the amount of SWE that will evaporate by sublimation. \(F_s\) corrects \(E_p\) since the evaporation rate is not necessarily the same between the soil reservoir and the snowpack.

Should not be estimated in most cases. Default value is 0%.

[watershed.1]
snow.correction.pet = { value = 0.0, lower = -20.0, upper = 20.0, opti = false, sameas = 0 }

Maximum snow retention

The maximum snow retention of the snow pack \(r_m\) (%).

Should be optimised. Default value is 5%.

[watershed.1]
snow.maximum_retention = { value = 5.0, lower = 0.001, upper = 30.0, opti = false, sameas = 0 }

Calculating snow melting

In the following section, \(H_0\) and \(r_0\) designate the initial SWE and retention values, respectively, at the beginning of the time step of duration \(\Delta t\) (s).

Snow melting by soil calories

If \(H_n > 0\), snow can melt daily under the action of the calories released by the soil. The SWE likely to melt in this way is noted as \(C_s\) (1/10 \(\mathrm{mm.day^{-1}}\)):

\[\begin{split}H_n & = \max \left (H_0 - \frac{\Delta t}{86400}C_s, 0 \right ) \\ M_s & = H_0 - H_n.\end{split}\]

Degree day model

A degree day model assumes that, for each 1°C over 0°C, a certain depth of snow will be melted. The degree day parameters are the degree day factor \(C_d\) (\(\mathrm{mm.°C^{-1}.day^{-1}}\)) and the temperature threshold \(T_d\) (°C) above which melting occurs. If \(T > 0\):

\[\begin{split}H_n' & = \max \left (H_n - \frac{\Delta t}{86400}C_d(T - T_d), 0 \right ) \\ M_d & = H_n' - H_n \\ H_n & = H_n'.\end{split}\]

Snow melting by rainfall calories

If \(R > E_p\) and \(T > 0\), snow can melt under the action of the rainfall calories:

\[\begin{split}H_n' & = \max \left (H_n - (1 + 0.01F_r)\frac{(R - E_p)T}{79.7}, 0 \right ) \\ M_r & = H_n'- H_n \\ H_n & = H_n'.\end{split}\]

The latent heat of fusion of water is 79.7 calories per gram.

Sublimation of snow

If \(R < E_p\), \(E_a\) and \(E_u\) (mm) are calculated as follows:

\[\begin{split}H_n' & = \max (H_n - ((1 +0.01F_s)E_p - R), 0) \\ E_S & = H_n - H_n' \\ E_U & = \max(E_P - E_S, 0) \\ H_n & = H_n'.\end{split}\]

Snow melting and runoff calculation

Assuming that \(M_s\) feeds directly the soil reservoir, the free liquid water in the snowpack \(W_r\) (mm) is:

\[W_f = M_d + M_r + \max (R - E_p, 0)\]

The total available liquid water in the snowpack is equal to the addition of the bounded water \(r_0H_0\) and the free water \(W_f\). If \(r_0 < r_m\), a part of this total available liquid water, denoted \(W_d\), will replenish the snowpack in order for \(r\) to reach the maximum snow retention \(r_m\) (%):

\[\begin{split}r & = \min \left (\frac{r_0H_0 + W_f}{H_n}, 0.01r_m \right ) \\ W_d & = \min((0.01r_m - r)H_n, W_f) \\ H_n & = H_n + W_d.\end{split}\]

\(1 - \frac{W_d}{W_f}\) is the fraction of the total liquid water available that contribute either to \(q_{sm}\) or to \(q_{sr}\):

\[\begin{split}q_{sm} & = M_s \left (1 - \frac{W_d}{W_f} \right ) \\ q_{sr} & = \max \left ( \left ( R - E_p \right) \left( 1 - \frac{W_d}{W_f} \right), 0 \right)\end{split}\]