Discharge measurements at Hannigalp

The two studied winters had very different weather conditions. Winter 2000/01 was characterized by early snow (60 cm were measured at Hannigalp on 1st November) and a thick snowpack. As a result, the soil remained unfrozen until the end of the snowmelt period. The next winter, a deep and persistent soil frost built up. In late autumn 2001, cold air temperatures (average of—13 °Cin

December) were recorded. As the snow cover was shallow (5 cm in November and December), the soil froze to a depth of 50 cm (Figure 7.6). At the end of the snowmelt, the frozen layer thawed only slowly so that the ground was still frozen to a depth of 45 cm on 15 May.

Discharge was influenced by the physical state of the soil. Under unfrozen conditions, the soil infiltration capacity was high, and the meltwater infiltrated entirely into the ground. As a result, no lateral flow1 was recorded in spring 2001 (Figure 7.6). Under frozen conditions, a significant amount of meltwater was collected as surface and subsurface runoff. The soil infiltration capacity was

1In this text, the term lateral flow refers to both surface and subsurface runoff

- Subsurface runoff measured

- Subsurface runoff simulated

- Accum. subsurf. runoff measured

---- Accum. subsurf. runoff simulated

1

<

1

I I

¡ ¡

, ,

- r ---,--

1 1

1

i :

1

1

1 1 1 1

1

i i i i i

Figure 7.7 Simulated and measured subsurface runoff and surface runoff (daily fluxes and total accumulation) for spring 2002 at Hannigalp

06.04 10.04 14.04 18.04 22.04 26.04 30.04 04.05 08.05 12.05 16.05

Figure 7.7 Simulated and measured subsurface runoff and surface runoff (daily fluxes and total accumulation) for spring 2002 at Hannigalp reduced by the pore ice and the presence of a basal ice sheet. This ice sheet built up in March 2002 when a first snowmelt event wetted the whole snowpack. During the main snowmelt, at the end of April 2002, approximately 30% of the total meltwater ran off as lateral flow. When no basal ice sheet was present, the snowmelt discharge was only marginally influenced by the soil frost. This was observed after the massive snowfall (54 cm) in May 2002, which recovered an already bare soil (i.e. no basal ice), the meltwater infiltrated entirely into the ground, despite the fact that the soil was still frozen below a depth of 10 cm.

At Gd St Bernard, the discharge patterns showed a similar behavior in both years. However, because of the steeper slope and to the more intense snowmelt (southerly exposure), a larger portion of meltwater ran off as surface flow. In total, 10% of the meltwater ran off as lateral flow during the first unfrozen winter, whereas this proportion increased to 40% vol. during the frozen winter 2001-02. Similarly to Hannigalp, the presence of a basal ice sheet was the main reason for the increase in the surface runoff.

The COUP-model accurately reproduced the timing and intensity of the surface and subsurface discharge during the main snowmelt event in spring 2002 (Figure 7.7). However, no discharge from the snowpack was simulated during the first snowmelt event in early April 2002, resulting in a slight shift (one day) in the lateral flow during the final snowmelt. At the onset of the final melt period, the simulated soil was still too dry compared to the measurements, and therefore all meltwater was able to infiltrate into the ground. Consequently, in a second run the soil infiltration capacity was reduced to account for the presence of basal ice, which steadily increased the surface runoff.

0 0

Post a comment