Seasonal evolution of supraglacial lake volume from ASTER imagery

Water stored in and released from supraglacial lakes is an important factor when considering the seasonal and long-term evolution of the Greenland ice sheet. Here we use a radiative transfer model to estimate changes in the depth and volume of a supraglacial lake on the surface of Jakobshavn Isbrae, West Greenland, between 2002 and 2005. When compared to estimates of the lake depth determined from airborne lidar observations, we estimate that the root-mean-square departure of the modelled lake depths was 0.3 m during cloud-free conditions. The maximum lake area, depth and volume were 3.4 km(2), 9.6 +/- 1.0 m and (18.6 +/- 3.7) x 10(6)m(3), respectively. When sequenced according to the number of positive degree-days (PDDs) accumulated prior to each image, we observe that the lake volume evolves in three distinct phases. At the start of the melting season, the rate of filling is slow; after approximately 80 PDDs the rate of filling increases by a factor similar to 3, and after approximately 125 PDDs the lake drains rapidly. We estimate that the lake drains at a minimum rate of (2.66 +/- 0.53) x 10(6) m(3)d(-1) over a 6 day period.