Download the entire proceedings as an Adobe PDF - Eastern Snow ...

26
63 rd EASTERN SNOW CONFERENCE
Newark, Delaware USA 2006
Main model settings
For simulating the hydrological year of 2004/2005 the calibrated model has been initialized at
the beginning of October 2004 by in field observed SWE data. The main calibrated parameters of
the model are shown in Table 1 in comparison to similar studies (Pellicciotti et al., 2005; Hock,
1999; Zappa et al., 2003). The melt factors presented in this paper are of the same size as the
compared values. The degree-day-factor for ice is slightly higher, which could be explained by the
dark ice surface and therefore quite low albedo of Goldbergkees glacier. The snowmelt model is
parameterized by the radiation melt factor and the maximum and the minimum temperature
dependent melt factors, which define the maximum at 21 st of June and the minimum at 21 st of
December of a sinus shaped function. The temperature dependent melt factor for ice is constant in
time.
Parameter
Table 1. Comparison of the model parameters calibrated at Goldbergkees watershed and the
parameters of the studies of Pellicciotti et al. (2005), Hock (1999) and Zappa et al. (2003)
RESULTS AND DISCUSSION
Cal.
Value
Zappa Hock Pellicciotti
Threshold temperature for snowmelt 0 0 0 1 [°C]
Max. degree day factor 3.2 – – –
[mm d –1 K –
1
]
Min. degree day factor 1 – – –
[mm d –1 K –
1
]
Degree day factor – 0.8 1.8 1.97
[mm d –1 K –
1
]
Radiation melt factor for snow 0.00015 0.00027 0.0008 0.00052
[mm W –1
m² K –1 h –1 ]
Temperature melt factor for ice 2.15 – 1.8 1.97
[mm d –1 K –
1
]
Radiation melt factor for ice 0.0003 – 0.0006 0.00106
[mm W –1
m² K –1 h –1 ]
Storage time for snowmelt 25 – – – [h]
Storage time for icemelt 2 – – – [h]
Translation time for snowmelt 3 – – – [h]
Translation time for icemelt 2 – – – [h]
Distributed snow-water-equivalent (SWE)
The PREVAH model makes a daily output of the distributed SWE storage possible. Thus the
validation using observed SWE data is an additional alternative. Figure 2 shows the validation of
the SWE for four different points in time starting with the initial SWE model settings of 6 October
2004. The simulated plot for this date shows the mean value of the SWE storage at the end of the
first simulated day described for the internal distribution of the meteorological units. Through this
semi-distributed approach some information on spatial distribution is lost. The simulated and
observed maximum snow accumulation at 6 May 2005 are in a very good agreement. At 2 June
2005 the simulation shows a little smaller SWE storage than the month before but the observed
SWE seems to show an overestimation through the field measurement. There is no possibility to
explain the high accumulation by observed precipitation data. At 7 July 2005 still a slightly
overestimation is seen. The last observation at 28 July 2005 shows a good accordance of the
simulated SWE, averaged over the catchment area. The simulated distribution of snow left in the
Unit