221
Modeling Snowmelt Runoff under Climate Change Scenarios
distribution, and it is also the basic input to SRM. Accurate determination of the
time series of SCA is, therefore, essential in order to simulate correctly melt processes and predict daily flows (Walter et al. 2005), forecast runoff (Salomonson and
Appel 2004), and understand the impacts of climate change (Robinson et al. 1993;
Wang et al. 2010).
10.3.1 IN SITu MeaSuReMent hiStoRy
Starting in the 1930s, conventional methods including the use of snow courses/
surveys and snow pits began to be used to obtain point measurements of snow water
equivalent (SWE; Dressler et al. 2006). These manual methods worked well in providing historical records of SWE; however, the low spatial and temporal resolution
of the data, coupled with the labor intensiveness of data collection from these snow
courses, was a significant drawback to the usefulness of the conventional methods.
Installation of an automated network of snow telemetry (SNOTEL) sites began in
1963, with the aim of supplementing and, to some extent, replacing the manually
operated snow courses (Serreze et al. 1999). Real-time data of SWE can be collected
from these SNOTEL stations; likewise, SCAs can be derived from the snow melt-out
dates as described by Garen and Marks (2005). The high temporal resolution of the
SNOTEL sites and the automated nature of data collection serve as improvements
over the snow courses. Notwithstanding the high temporal resolution of the information available from these automatic stations (Egli 2008), the spatial resolution of
this information is generally coarse (Farinotii et al. 2010). For example, a study conducted by Bales and Rice (2006) in Sierra Nevada showed the presence of snow at
higher elevations, even when all snow at lower elevations where the surface measurement stations were located had melted. Furthermore, SNOTEL stations and other
point stations are not present in many areas of the world (Ault et al. 2006); hence,
they cannot be employed to monitor snow activities or the distribution of snow on a
global level (Molotch 2009).
10.3.2 Satellite hiStoRy
A more spatially complete and comprehensive view of snow cover extent requires
information from satellite-borne sensors (Robinson et al. 1993) due to the large spatial coverage of satellite remotely sensed data (König 2001) in contrast to the low
spatial coverage of the SNOTEL sites. This realization coupled with the importance
of knowing the distribution of snow over a large area led the National Oceanic and
Atmospheric Administration (NOAA) to commence the first operational snow mapping over Northern Hemisphere land surfaces in 1966 (Robinson et al. 1993). Since
then, progress has been recorded in the utilization of space-borne sensors toward
monitoring the variability of snow extent in space and time (Salomonson and Appel
2004). In 1972, the very high resolution radiometer (VHRR) with a spatial resolution
of 1.0 km was launched, followed by the advanced VHRR launched in 1978 having
a spatial resolution of 1.1 km.
In order to generate snow cover information on a larger scale as well as to generate
information on snow volume, which was not possible with optical sensors launched
Modeling Snowmelt Runoff under Climate Change Scenarios
distribution, and it is also the basic input to SRM. Accurate determination of the
time series of SCA is, therefore, essential in order to simulate correctly melt processes and predict daily flows (Walter et al. 2005), forecast runoff (Salomonson and
Appel 2004), and understand the impacts of climate change (Robinson et al. 1993;
Wang et al. 2010).
10.3.1 IN SITu MeaSuReMent hiStoRy
Starting in the 1930s, conventional methods including the use of snow courses/
surveys and snow pits began to be used to obtain point measurements of snow water
equivalent (SWE; Dressler et al. 2006). These manual methods worked well in providing historical records of SWE; however, the low spatial and temporal resolution
of the data, coupled with the labor intensiveness of data collection from these snow
courses, was a significant drawback to the usefulness of the conventional methods.
Installation of an automated network of snow telemetry (SNOTEL) sites began in
1963, with the aim of supplementing and, to some extent, replacing the manually
operated snow courses (Serreze et al. 1999). Real-time data of SWE can be collected
from these SNOTEL stations; likewise, SCAs can be derived from the snow melt-out
dates as described by Garen and Marks (2005). The high temporal resolution of the
SNOTEL sites and the automated nature of data collection serve as improvements
over the snow courses. Notwithstanding the high temporal resolution of the information available from these automatic stations (Egli 2008), the spatial resolution of
this information is generally coarse (Farinotii et al. 2010). For example, a study conducted by Bales and Rice (2006) in Sierra Nevada showed the presence of snow at
higher elevations, even when all snow at lower elevations where the surface measurement stations were located had melted. Furthermore, SNOTEL stations and other
point stations are not present in many areas of the world (Ault et al. 2006); hence,
they cannot be employed to monitor snow activities or the distribution of snow on a
global level (Molotch 2009).
10.3.2 Satellite hiStoRy
A more spatially complete and comprehensive view of snow cover extent requires
information from satellite-borne sensors (Robinson et al. 1993) due to the large spatial coverage of satellite remotely sensed data (König 2001) in contrast to the low
spatial coverage of the SNOTEL sites. This realization coupled with the importance
of knowing the distribution of snow over a large area led the National Oceanic and
Atmospheric Administration (NOAA) to commence the first operational snow mapping over Northern Hemisphere land surfaces in 1966 (Robinson et al. 1993). Since
then, progress has been recorded in the utilization of space-borne sensors toward
monitoring the variability of snow extent in space and time (Salomonson and Appel
2004). In 1972, the very high resolution radiometer (VHRR) with a spatial resolution
of 1.0 km was launched, followed by the advanced VHRR launched in 1978 having
a spatial resolution of 1.1 km.
In order to generate snow cover information on a larger scale as well as to generate
information on snow volume, which was not possible with optical sensors launched