The heat budget of a proglacial stream, Phelan Creek, from the Gulkana Glacier, Alaska, is estimated by using the hourly time series of stream water temperature, discharge and meteorology obtained in the summer of 2006. As an analytical result, the net shortwave radiation and bed friction occupied 32.1 and 38.2% of the total thermal input to the stream, respectively. The time series of the water temperature were simulated by a deterministic model with the coupling of the heat budget and the thermal advective diffusion equation. The simulated result is agreeable to the observed one with the Nash-Sutcliffe efficiency (NASH) of 0.747 and the root-mean-square error (RMSE) of 0.2368C over the observation period. However, under condition that the rainfall of more than 1 mm h À1 occurs continually for more than a day, the simulation was less reasonable with NASH ¼ 0.225 and RMSE ¼ 0.2268C. This is probably because the relatively warm subsurface flow input to the stream channel from the non-glacial area.

Calving glaciers are rapidly retreating in many regions under the influence of ice‐water interactions at the glacier front. In contrast to the numerous researches conducted on fjords in front of tidewater glaciers, very few studies have been reported on lakes in which freshwater calving glaciers terminate. To better understand ice‐water interactions at the front of freshwater calving glaciers, we measured lakewater temperature, turbidity, and bathymetry near Glaciar Perito Moreno, Upsala, and Viedma, large calving glaciers of the Southern Patagonia Icefield. The thermal structures of these lakes were significantly different from those reported in glacial fjords. There was no indication of upwelling subglacial meltwater; instead, turbid and cold glacial water discharge filled the region near the lake bottom. This was because water density was controlled by suspended sediment concentrations rather than by water temperature. Near‐surface wind‐driven circulation reaches a depth of ~180 m, forming a relatively warm isothermal layer (mean temperature of ~5–6°C at Perito Moreno, ~3–4°C at Upsala, and ~6–7°C at Viedma), which should convey heat energy to the ice‐water interface. However, the deeper part of the glacier front is in contact with stratified cold water, implying a limited amount of melting there. In the lake in front of Glaciar Viedma, the region deeper than 120 m was filled entirely with turbid and very cold water at pressure melting temperature. Our results revealed a previously unexplored thermal structure of proglacial lakes in Patagonia, suggesting its importance in the subaqueous melting of freshwater calving glaciers.

The hydrodynamics of turbidity currents from spring runoffs of an influent river is examined by measuring velocity, water temperature, and concentration of suspended sediment in a reservoir. Crosssectional measurement of turbidity currents, made from a bridge, shows the effect of centrifugal force on the downstream movement. This is due to considerable meandering of the old river channel. The turbidity currents, initiated at high sediment concentration of 1,700 mg/L, were bifurcated into an underflow and some interflows in the downstream reach, because of the relatively dense bottom water of the reservoir. The "lower layer" of turbidity currents exhibits a log linear relationship between dimensionless velocity and height under a hydraulically smooth condition. The velocity distribution in the upper and lower layers could be explained by overall Richardson number. Two drag coefficients, related with shear stresses at bottom bed and interface, generally decrease with increasing densimetric Froude number, but are independent of Reynolds number. The conditions of sediment deposition and erosion by turbidity currents can be explained by using the "extended Shields diagram." centration (25-2500 mg/L) and at low velocity (the order 10-• to 10 -2 m/s), while the former currents, called high-density turbidity currents, usually transport sand and gravel in high concentration (25-250 g/L) and at high velocity (the order 10 m/s) [Stow, 1986]. Actual events of turbidity currents have been noted by the sequential breakage of submarine cables (high density) [Heezen and Ewinq, 1952] and through field measurements (low density) in the sea [Shepard et al., 1977], reservoirs [Chikita and Okumura, 1987], and lakes (especially, proglacial lakes by Gustat•son [1975], Smith et al. [1980], and Weirich [ 1986]). Mathematical models of the dynamics of turbidity currents, with erosion and deposition, have recently been developed by Parker [1982], Akiyama and Stefan [1985], Fukushima et al. [1985], and Parker et al. [1986]. However, these models are based on a small amount of field data and few experimental results; they leave some doubts about reality in the field. Particular effects which are not taken into account are the topography on the downstream movement of turbidity currents, interaction among wind-driven currents and turbidity currents, and the grain size distribution of suspended and bottom sediments. Furthermore, the shear stress at an interface between turbidity currents and clear flows is negected, and the

No abstract

No abstract