27 FEATURE STORY GROUND TRUTHING Measuring diurnal changes in the flow of meltwater provided critical insight into the accuracy of mathematical models of glacial melting, as well as vital data that will help improve those models. Scientists engaged in a wide range of climate science disciplines rely on surface mass balance models that predict the effects of temperature, solar radiation and the reflectance of the ice surface on the rate of glacial melting. Until the 1980s or '90s, most of the mass deposited into the ocean from the Greenland Ice Sheet was the result of dynamical flow, the slow movement of the glacier as the weight of ice and snow causes it to deform and creep to the sea. At the end of their journey, the leading edges of the glaciers crumble into the sea, contributing water to the ocean as melting icebergs. However, in the past 20 to 30 years, glacial melting has eclipsed dynamical flow as the main source of water conveyance from glaciers to the ocean, says Cooper. He adds that meltwater flowing into moulins also affects the rate of dynamical flow. If meltwater makes its way to the base of the glacier through narrow channels, it can become so pressurized that it lifts the glacier like a hydraulic jack, Cooper says, speeding the glacier's slide to the sea and increasing the rate of iceberg calving. But if the meltwater runoff increases to the point that it creates a bigger, more efficient conveyance network, a higher rate of melting could actually reduce the pressure at the base of the glacier and slow the movement of the ice. So whether the question is how quickly meltwater is running into the ocean or how it may be speeding or slowing the movement of the glacier itself, surface mass balance glacial melting models are crucial tools in estimating the answers and predicting their results. To find out, the Smith/Rennermalm team chose a supraglacial river to study, which they named Rio Behar after a colleague. For 72 hours straight, they measured stream flow, as well as detailed weather readings, building a huge, real-life database of water movement and conditions at the site. They then teamed up with modeling teams who ran models with meteorological data from the site to compare their field data with the model outputs. The team found that the surface mass balance models overestimated the runoff by 21 to 58 percent. The team also noted a lag time of 0.4 to 9.5 hours between periods of high melting and increases in river discharge that was not reflected in the model predictions. Brandon Overstreet "does the compass calibration dance" on the ice with a SonTek RiverSurveyor M-9 mounted on a HydroBoard II. HYDROBOARD II The RiverSurveyor ADCP is mounted on a HydroBoard II that keeps it stable as the instrument is pulled back and forth across a river or stream. Rigged to enhance stability and minimize nose dive, and outfitted with adjustable fins, the HydroBoard II excelled in the rushing waters of Rio Behar at peak flow. Photo Courtesy: Lincoln Pitcher