A Statistically Based Method to Estimate Long‐Term Daily Air Temperature at High Elevations

Air temperature is a key variable influencing numerous chemical, physical, hydrological and biological processes; however, long‐term observations are lacking, particularly at high elevations. This study presents a statistical methodology for reconstructing daily air temperature time series at the highest permanently manned meteorological station in Switzerland, the Jungfraujoch (3571 a.s.l.), based on observations from 30 lower‐altitude stations (485–2691 m a.s.l.) dating back to 1900. The reconstructed time series were then compared with observations from 1933 to 2023, as well as with two high‐resolution gridded datasets (HISTALP and Imfeld et al. 2023) to validate the period from 1900 to 1933. We found that (i) the selection of stations with temporally consistent long‐term observations is a critical issue, (ii) model performance, efficiency and errors are primarily influenced by altitude, (iii) the Kling–Gupta Efficiency (KGE) is an appropriate metric for defining the ensemble simulation, considering correlation and bias on the mean and standard deviation values, (iv) the ensemble simulation increases the temporal consistency of the estimated time series and mediates the latitudinal and longitudinal gradients, (v) comparable performance to existing datasets is achieved, despite the low‐data requirements, but with greater computational efficiency, and (vi) the estimated time series can provide a benchmark for evaluating time series anomalies and for a deeper analysis of the elevation‐dependent warming issue.