Snow is a form of precipitation; however, in hydrology it is treated differently because of the lag between when it falls and when it produces runoff, groundwater recharge, and is involved in other hydrologic processes. Remote sensing is a valuable tool for obtaining snow data for predicting snowmelt runoff as well as climate studies. Rango [1993] presents a good review of the status of remote sensing in snow hydrology. Nearly all regions of the electromagnetic spectrum provide useful information about the snowpack. Depending on the need, one may like to know the areal extent of the snow, its water equivalent, or the ``condition'' or grain size, density and presence of liquid water. Although no single region of the spectrum provides all these properties, techniques have been developed to provide all of the properties to some degree or other.
The water equivalent of snow can be measured from low elevation aircraft carrying sensitive gamma radiation detectors [ Carroll and Vadnais, 1980]. This approach is limited to low aircraft altitudes (approximately 150 m) because the atmosphere attenuates a significant portion of the gamma radiation. Currently, this operational program covers over 1400 flight lines annually in the United States and Canada [ Carroll and Carroll, 1989]. This method is effective for measuring snow cover in open plains, but is less effective in more hilly terrain or when there is extensive forest cover or regions with deep snowpacks.
Snow can readily be identified and mapped with the visible bands of satellite imagery because of its high reflectance in comparison to non-snow areas. Although snow can be detected at longer wavelengths, i.e., in the near infrared region, the contrast between snow and non-snow areas is considerably reduced compared to the visible region of the spectrum. However, the contrast between clouds and snow is greater in the infrared region and serves as a useful discriminator between clouds and snow [ Dozier, 1984]. Thermal data are perhaps the least useful of the common remote sensing products for measuring snow and its properties; but they can be useful for helping identify snow/no-snow boundaries and discriminating between clouds and snow.
Use of satellite data for snow mapping has become operational in several regions of the world. Currently, NOAA develops snow cover maps for about 3000 river basins in North America of which approximately 300 are mapped according to elevation for use in streamflow forecasting [ Carroll, 1990]. NOAA also produces regional and global maps of mean monthly snow cover [ Dewey and Heim, 1981].
Microwave remote sensing offers great promise for future applications to snow hydrology. This is because the microwave data can provide information on the snowpack properties of most interest to hydrologists; i.e., snow cover area, snow water equivalent (or depth), and the presence of liquid water in the snowpack which signals the onset of melt [ Kunzi et at., 1982]. With the availability of satellite microwave data (Scanning Multichannel Microwave Radiometer (SMMR and SSM/I), Chang et al [1982] have developed an algorithm for estimating snow water equivalent for dry snow and mapped the depth and global extent of snow cover [ Chang et al, 1987]. The passive microwave systems are limited by their interaction with other media such as forest areas although a method to correct for the absorption of the snow signal by forest cover has been developed [ Chang et al, 1990]. The spatial resolution attainable by the passive satellite systems is also a limitation but Rango et al, [1989] have shown that reasonable snow water equivalent estimates can be made on basins smaller than 10000 sq km.
Active microwave remote sensing also has the potential to provide important information about the snow pack and at very high resolution with Synthetic Aperture Radar (SAR), [ Stiles et al, 1981, and Rott, 1986]. Unfortunately, analysis of radar data is more complex than passive microwave data and until very recently, there have been no orbiting SAR systems for collecting snow data. In spite of that, aircraft SAR measurements have shown that SAR can discriminate between snow and glaciers from other targets and discriminate between wet and dry snow [ Shi and Dozier, 1992, and Shi et al, 1994]
Snowmelt runoff procedures that use remote sensing can be grouped into empirical approaches and modeling. Early use of remote sensing focused on empirical relationships between snow cover area or percent snow cover and monthly or accumulated runoff [ Rango et al, 1977, Ramamoorthi, 1987]. These simple relationships work very well for some applications and particularly in data sparse regions of the world. A number of models have been modified to use remote sensing data for predicting snowmelt runoff [ Hannaford and Hall, 1980, Dilard and Orwig, 1979]. On the other hand the Snowmelt Runoff Model (SRM) [ Martinec et al, 1983] was specifically developed for using remote sensing of snow cover by elevation zone as the primary input variable. SRM has been extensively tested on basins of different sizes and regions of the world [ Rango 1992, WMO 1992]. Although SRM is a degree day model that uses only snow cover as remote sensing derived input, energy balance models [ Leavesley and Stannard, 1990 and Marks and Dozier, 1992] should be able to use additional remote sensing data such as albedo and other energy balance parameters (see Evaporation).