Melt-pond extent is the primary driver of albedo evolution on sea ice after the onset of melt, and a key driver of summer surface heat budget. Due to the low topographic relief found on undeformed seasonal ice, pond extent can change dramatically in a matter of just a few days, particularly when meltwater is above hydrostatic equilibrium. Repeat surface-based LiDAR surveys represent a new method for monitoring important aspects of melt-pond evolution, building three-dimensional models over areas of hundreds of meters square, with centimeter-scale accuracy. Operating a surface-based LiDAR on a floating melting sea-ice platform, however, is challenging. We have developed techniques for stably collecting and overlaying multiple scans on floating ice that is subject to both vertical and horizontal displacement, as well as techniques for correcting and analyzing the data. Surveys of this kind were taken of two sites roughly 100 m × 200 m in size off Barrow, Alaska, during the 2009 melt season from 17 May to 25 June. After post-processing, the results allow us to track both pond volume changes and topographic evolution of the surface over the course of a melt season. Coupled with other observations at the site, these data shed light on several mechanisms of pond evolution and may serve as a validation dataset for high-accuracy pond models.