Improving current and future satellite observations of snow water equivalent

Snow water equivalent (SWE) is the liquid equivalent of water of a known area of snowpack. Seasonal and inter-annual changes in the global extent of SWE have a strong and complex influence on estimates of the global energy balance. As the global energy balance is an important part of global climate models, which are used to predict climate change, it is vital that current uncertainties in SWE estimates are identified and minimised to reduce their impact on predictions of future climate scenarios. The distribution of SWE can be highly variable over space and time; even over flat, uncomplicated land surfaces. Consequently, to get an adequate estimate of SWE on a global scale, observations are required at a horizontal resolution of 200-500 m and a temporal resolution of 15 days (or even less when snowpacks are melting). Observations of SWE at these resolutions are required to adequately test how well global climate models predict SWE; particularly as the accuracy of SWE predictions by global climate models has an important knock-on effect as to how well such models predict future climate scenarios. However, current observations of SWE do not meet these horizontal and temporal requirements. The global distribution of ground-based SWE observations are too sparse and, although satellite observations more closely match the greater spatial extents required to evaluate modelled estimates, none of the currently available satellite sensors are designed specifically to measure SWE; those that are used to get some estimate of SWE only have a horizontal resolution of 25,000 m. Consequently, we urgently need to find out: 'How can we reduce uncertainty in estimates of SWE from current satellite sensors and can we provide the scientific justification for new sensors specifically designed to observe SWE?' Recent technological advances in ground-based radar has meant that, for the first time, observations of SWE (to an accuracy of 10%) are possible at a rate of up to 50 observations a second using a cheap, lightweight, low-power radar system attached to a snowmobile. This proposal will capitalise on such technological advances to make high horizontal resolution measurements (~10 cm) within the footprint of current satellite sensors (25 x 25 km), which will allow the uncertainty in SWE to be accurately assessed. Observations of SWE and other snowpack properties will be made periodically from snowpits to provide a double check on the accuracy of radar observations. Also, hourly changes in SWE will be observed using this radar system as a snowpack first accumulates and then melts-out throughout an annual cycle. Hourly radar observations will be made throughout the winter at a range of frequencies and angles relative to the snowpack surface. This will mimic potential new sensors which have been proposed to specifically measure SWE. Currently, the abilities of proposed new sensors designed to observe SWE have only been justified by theoretical studies. This work will provide the first data set that is able to test these theoretical studies over a wide range of snowpack conditions. Estimates of SWE and other snowpack properties (e.g. vertical profiles of temperature, grain size and liquid water content) using a computer model will provide essential hourly information to interpret and compare with the radar observations. Periodic snowpits will again be used to double check the accuracy of modelled estimates and radar observations. The timing and focus of this proposal takes advantage of exceptional logistical and scientific opportunities currently scheduled for 2007-10 as part of ongoing work by NASA and the European Space Agency. It will add great value to current and future proposals for satellites dedicated to the observation of SWE and, more generally, it will advance the collaborative and international nature of snow science research as part of the International Polar Year.