Adjustment of Monthly Air Temperature and Precipitation Data from Automatic System to Align with Manually Measured Long-Term Data at High-Altitude Observatory

Mountain weather stations with continuous measurements maintained over decades serve as indispensable data sources for analysing climate dynamics. The transition from manual to automatic measuring systems may introduce inhomogeneities in these climate data series, therefore, it may be necessary to adjust the data, particularly when considering a transition to solely automatic measurements.

In our study, we analysed climate data parallelly measured by the automatic (AWS) and manual weather station (MWS) at the Skalnaté Pleso Observatory (1778 m a.s.l.) in the High Tatra Mountains. Manual meteorological measurements have been performed at this location since 1943 using the same methods and devices. The AWS Physicus with the parallel recording of meteorological data was installed at the observatory in 2014. Excluding the first years of trial operation with several data gaps in the AWS data, we processed six years of parallel measurements from 2017 to 2022 to derive corrections applicable to AWS data, ensuring the continuity and homogeneity of the conventional long-term data series. We utilized monthly regressions (MR) and cumulative distribution functions (Generalised extreme value – GEV and Gaussian distribution - GAUSS) proposed in Lukasová et al. (2023, DOI:10.1127/metz/2023/1200) and applied them to AWS data measured in 2023. Our focus was on two essential climatological parameters: air temperature and atmospheric precipitation.

Comparison of data from parallel manual and automatic measurements revealed the underestimation of monthly precipitation totals in AWS data, with a mean bias error (MBE) of -6.8 mm and root mean squared error (RMSE) of 20 mm. Among the methods considered, correction by MR yielded the lowest errors, reducing them to 1.5 mm and 11.3 mm for MBE and RMSE, respectively. For air temperature, the monthly AWS data were overestimated by 0.07, 0.28 and 0.11 °C for Tmax, Tmin and Tmean, respectively. The lowest errors after corrections of Tmax were achieved with MR and GAUSS methods with MBE of 0.0 °C and RMSE of 0.1°C for both. For Tmin and Tmean, MR and GEV methods resulted in MBE of 0.0 °C and RMSE of 0.3 °C and MBE of 0.0 °C and RMSE of 0.1 °C for both methods, respectively. Based on these results, AWS data corrected using MR for atmospheric precipitation and MR/GAUSS for air temperature can be considered suitable for maintaining the continuity of historical climate data series at Skalnaté Pleso Observatory.

Despite our findings, we recommend continuing parallel manual and automatic measurements at high-altitude meteorological observatories exposed to extreme weather events. In case of equipment failure, is often difficult to repair it, especially in harsh weather conditions and limited access to high mountains. For the observatories at unique positions, this may cause data loss that is difficult to compensate with data from other stations.