Spatial and temporal variability of air temperature on a melting glacier: Atmospheric controls, extrapolation methods and their effect on melt modeling, Juncal Norte Glacier, Chile

Temperature data from three Automatic Weather Stations and twelve Temperature Loggers are used to investigate the spatiotemporal variability of temperature over a glacier, its main atmospheric controls, the suitability of extrapolation techniques and their effect on melt modeling. We use data collected on Juncal Norte Glacier, central Chile, during one ablation season. We examine temporal and spatial variability in lapse rates (LRs), together with alternative statistical interpolation methods. The main control over the glacier thermal regime is the development of a katabatic boundary layer (KBL). Katabatic wind occurs at night and in the morning and is eroded in the afternoon. LRs reveal strong diurnal variability, with steeper LRs during the day when the katabatic wind weakens and shallower LRs during the night and morning. We suggest that temporally variable LRs should be used to account for the observed change. They tend to be steeper than equivalent constant LRs, and therefore result in a reduction in simulated melt compared to use of constant LRs when extrapolating from lower to higher elevations. In addition to the temporal variability, the temperature-elevation relationship varies also in space. Differences are evident between local LRs and including such variability in melt modeling affects melt simulations. Extrapolation methods based on the spatial variability of the observations after removal of the elevation trend, such as Inverse Distance Weighting or Kriging, do not seem necessary for simulations of gridded temperature data over a glacier.