Observations and control run

An analysis of the observed snow cover at D88-89 and D98-99 shows that the snow duration is intermittent (Figure 19.3). In 1988-1989, the maximum snow is observed in December (Days of Year (DOY): 354) and the snowpack has melted at the beginning of March (DOY: 68), whereas in 1998-1999, the maximum snow depth was reached at the end of February (DOY: 50). At S88-89, snow starts building up in December (DOY: 317) or in November at S98-99 (DOY: 291), reaches its maximum depth in April (88-89, DOY: 119; 98-99, DOY: 111) and begins to melt by the end of April. The melting is steady at a constant rate (in 88 -89, 2 to 3 cm and in 98 -99, 7 to 8 cm per day).

The model is first calibrated at the two stations over the two investigated periods. For the control simulations, inputs of hourly data from the Swiss Meteorological Service are used to drive the SEBM. Simulated snowpacks show generally good agreement with the observed snow depth, although it can be noted that the lack of a snow compaction parameterization in the model prevents a closer agreement (Figure 19.3). This can be seen at the Disentis station at the end of the year for both investigated periods and at S88-89 for the beginning of 1989 and at S98-99 for the end of 1998. Snowpack duration in the control run compared to the observations differed at D88-89 by two days, at D98-99 by one day, at S88-89 by two days, and at S98-99 by four days (Table 19.3).

In Figure 19.4, two days are chosen to illustrate a case with continuous melting and one day is chosen with daytime melting. The latter is described for January 31, 1989 at Disentis. The radiation input reaches nearly 150 W m-2 at noon. This energy is first used to warm the snowpack to f and to increase the heat convergence (AQs); thereafter, snow starts melting, that is, AQm increases at the expense of AQs. Melt ceases when

Table 19.3 Maximum snow depth in centimeters (in parentheses the date of occurrence of the maximum snow depth), the number of snow-covered days above the thresholds of 2.5 cm for Disentis and 5.0 cm for Santis, and the day of the year with the peak runoff. Diff stands for the difference in days between the respective scenario and the control run

Maximum snow depth:

Obs Ctrl First Second Third

D88-89 D98-99 S88-89 S98-99

Number of snow-covered days:

Obs

Ctrl

Diff

First

Diff

Second

Diff

Third

Diff

D88-89

94

96

-2

62

-34

95

-1

89

-7

D98-99

146

145

-1

118

-27

140

-5

139

-6

S88-89

252

250

-2

219

-31

248

-2

239

-11

S98-99

294

290

-4

239

-51

277

-13

287

-4

Day of year with the peak runoff:

Day of year with the peak runoff:

Ctrl First Diff Second Diff Third Diff

"O

"D

-100

Disentis 31st January 1989

Figure 19.4 Daily evolution of radiation and energy fluxes for the January 31, 1989 at Disentis and May 25, 1989 at Santis. Note the different vertical axes

284 282 280 278

270 268 266 264

Qh Qe

AQmelt

Snow depth ot-co^rmcoracno

Figure 19.4 Daily evolution of radiation and energy fluxes for the January 31, 1989 at Disentis and May 25, 1989 at Santis. Note the different vertical axes the temperature of the pack drops below TfK owing to the declining radiation input. The case of daytime melting occurs early in the year because the negative fluxes of energy and radiation during night dominate. The case of continuous melting is shown for Santis on May 25, 1989. It happens when Q*,Qh and Qe act as energy sources. Melting occurs already to a small amount during nighttime owing to the energy input of Qe and

Qh, which increase AQs. During the day, the energy storage, AQs, increases the values of AQm beyond the radiation input.

The control runs of the two stations simulate radiation, energy, and hydrological fluxes for the two snow seasons. At D88-89, Q* has negative daily averages because the melting begins in January. It is a daytime melting where the snowpack temperature is close to the melting

Disentis 88-89

Disentis 98-99

Santis 88-89

Santis 98-99

Disentis 88-89

Disentis 98-99

Santis 88-89

Santis 98-99

Days of year Days of year Days of year Days of year

Figure 19.5 The evolution of the runoff for the two stations, the two winters, and the three scenarios. The runoff is shown during the melting period, that is, from the maximum snow depth to the end of melting. For the exact days of year, see Table 19.3

Days of year Days of year Days of year Days of year

Figure 19.5 The evolution of the runoff for the two stations, the two winters, and the three scenarios. The runoff is shown during the melting period, that is, from the maximum snow depth to the end of melting. For the exact days of year, see Table 19.3

point, which is reached during the afternoons by radiative warming. Qh,Qe, and Qpcp are about zero. The peak of runoff is concentrated at the end of the melting period at the beginning of March (Table 19.3, Figure 19.5). The situation at D98-99 differs slightly because the snow melts only at the beginning of April. The simulated radiative flux Q* becomes positive. The energy fluxes Qh and Qe are weak. The timing of the peak runoff is at the beginning of April (Figure 19.5).

At Santis, the situation is quite different. The snowpack continuously melts where Q* is always positive and it reaches cumulated values of 150 (S88-89) to 200 MJm-2 (S98-99). Qh is negative in both years and has cumulated values of-60 (S88-89) and -50 MJm-2 (S98-99). In S88-89, Qe is positive and it becomes negative only at the end of June, whereas it has values of -25 MJm-2 in S98-99. In both the periods, large runoff peaks are already simulated before the end of melting at the end of July (S88-89) and at the beginning of August (S98-99; Table 19.3, Figure 19.5).

Was this article helpful?

0 0

Post a comment