Glacier melt is strongly impacted by climate and meteorology. Temperature lapse rates are used to model glacier melt, and the accuracy of the spatial distribution of modelled melt can be impacted by the lapse rate used in modelling. Additionally, the observed spatial distribution of melt is highly temporally variable. Whether this variability is caused by lapse rate is unknown. Storglaciären and Rabots glaciär in Northern Sweden were equipped with temperature measurement stations at both low and high glacier elevations over the 2014 ablation season. From these measurements, surface lapse rates 0.1 m above the glacier surface, and near-­surface lapse rates 2 m above the surface, were calculated for the two glaciers. The lapse rates were then compared to meteorological variables measured in the middle of the glaciers’ elevation range. In addition, a comparison was made with ablation data collected via ablation stakes throughout the melt season. On both Storglaciären and Rabots glaciär, the surface lapse rate is −0.28 °C (100m)−1 averaged over the ablation season. The season ­average near­-surface lapse rate is also the same for both glaciers, at −0.37 °C (100m)−1. The lapse rate values are shallow in comparison to non-­glaciated mountain areas. The meteorological variables of wind speed and precipitation affect surface lapse rate on short timescales. Long-­term patterns in surface lapse rate are influenced by incoming radiation, humidity and precipitation. In addition, topographic shading and albedo impact the incoming short­-wave radiation, causing diurnal and seasonal fluctuations in surface lapse rate. A cumulative approach to lapse rate using a positive degree day gradient reflects the pattern of ablation gradients measured through the ablation season. However, a lack of data means no robust conclusions can be drawn from this comparison. Many melt modelling studies use steeper lapse rates in ablation and mass balance modelling than observed over Storglaciären and Rabots glaciär. This can lead to underestimation of ablation at high glacier elevations. Measurements of local lapse rates recorded over glacier surfaces are necessary in order to produce more accurate ablation modelling results.

The reliance on freshwater released by mountain glaciers and ice caps demands that the effects of climate change on these thermally-sensitive systems are evaluated thoroughly. Coupling climate variability to processes of mass and energy exchange at the glacier scale is challenged, however, by a lack of climate data at an appropriately fine spatial resolution. The thesis addresses this challenge through attempting to reconcile this scale mismatch: glacier boundary-layer observations of meteorology and ablation at Vestari Hagafellsjökull, Iceland, and Storglaciären, Sweden, are related to synoptic-scale meteorological variability recorded in gridded, reanalysis data. Specific attention is directed toward synoptic controls on: i) near-surface air temperature lapse rates; ii) stationarity of temperature-index melt model parameters; and iii) glacier-surface ablation. A synoptic weather-typing procedure, which groups days of similar reanalysis meteorology into weather categories , forms the basis of the analytical approach adopted to achieve these aims. Lapse rates at Vestari Hagafellsjökull were found to be shallowest during weather categories characterised by warm, cloud-free weather that encouraged katabatic drainage; steep lapse rates were encountered in weather categories associated with strong synoptic winds. Quantitatively, 26% to 38% of the daily lapse-rate variability could be explained by weather-category and regression-based models utilizing the reanalysis data: a level of skill sufficient to effect appreciable improvements in the accuracy of air temperatures extrapolated vertically over Vestari Hagafellsjökull. Weather categories also highlighted the dynamic nature of the temperature-ablation relationship. Notably, the sensitivity of ablation to changes in air temperature was observed to be non-stationary between weather categories, highlighting vulnerabilities of temperature-index models. An innovative solution to this limitation is suggested: the relationship between temperature and ablation can be varied as a function of weather-category membership. This flexibility leads to an overall improvement in the simulation of daily ablation compared to traditional temperature-index formulations (up to a 14% improvement in the amount of variance explained), without the need for additional meteorological data recorded in-situ. It is concluded that weather categories are highly appropriate for evaluating synoptic controls on glacier meteorology and surface energetics; significant improvements in the parameterization of boundary-layer meteorology and ablation rates are realised through their application.