Modelling variable glacier lapse rates using ERA-Interim reanalysis climatology: an evaluation at Vestari- Hagafellsjokull, Langjokull, Iceland

The near-surface air temperature lapse rate is an important tool for spatially distributing temperatures in snow- and ice-melt models, but is difficult to parameterize, as it is not simply correlated with boundary-layer meteorological variables, such as temperature itself. This contribution quantifies spring-autumn lapse rate variability over 5 years at Vestari-Hagafellsjökull, a southerly outlet of Langjökull in Iceland. It is observed that summer lapse rates (0.57 °C 100 m−1) are significantly lower than non-summer rates, and are also lower than the Saturated Adiabatic Lapse Rate (SALR), which is often adopted in melt models. This is consistent with reduced near-surface temperature sensitivity to free-atmosphere temperature change during the occurrence of melting. A Variable Lapse Rate (VLR) regression model is calibrated with standardized, 750 hPa temperature anomalies derived from ERA-Interim climatology, which is shown to be highly significantly correlated with near-surface temperatures. The modelled VLR overestimates cumulative June–September Positive Degree Days (PDDs) by 3% when used to extrapolate temperatures from 1100 to 500 m a.s.l. on the glacier, whereas the SALR overestimates cumulative PDDs by 14%. ERA-Interim data therefore appear to offer a good representation of free-atmosphere temperature variability over Vestari-Hagafellsjökull, and the modelling approach offers a simple means of improving lapse rate parameterizations in melt models.