Abstract. Seasonal meltwater lakes on the Greenland Ice Sheet form when surface runoff is temporarily trapped in surface topographic depressions. The development of such lakes affects both the surface energy balance and dynamics of the ice sheet. Although areal extents, depths, and lifespans of lakes can be inferred from satellite imagery, such observational studies have a limited temporal resolution. Here, we adopt a modelling-based strategy to estimate the seasonal evolution of surface water storage for the ~ 3600 km2 Paakitsoq region of W. Greenland. We use a high-resolution time dependent surface mass balance model to calculate surface melt, a supraglacial water routing model to calculate lake filling and a prescribed water-volume based threshold to predict lake drainage events. The model shows good agreement between modelled lake locations and volumes and those observed in 9 Landsat 7 ETM+ images from 2001, 2002 and 2005. We use the model to investigate the lake water volume required to trigger drainage, and the impact that this threshold volume has on the proportion of meltwater that runs off the ice supraglacially, is stored in surface lakes, or enters the subglacial drainage system. Model performance is maximised with prescribed lake volume thresholds between 4000 and 7500 times the local ice thickness. For these thresholds, lakes transiently store < 40% of meltwater at the beginning of the melt season, decreasing to ~ 5 to 10% by the middle of the melt season. 40 to 50% of meltwater runs off the ice surface directly, and the remainder enters the subglacial drainage system through moulins at the bottom of drained lakes.
Dynamics of surface water storage in the Amazon inferred from measurements of inter-satellite distance change
