Climatic characteristics of snow water equivalent in the Perm Krai area

In this study, we compared ERA5-Land reanalysis data on snow water equivalent (SWE) with values of SWE obtained from snow-measuring surveys on 18 field (non-forest) and 17 forest routes in the Perm Territory for 1967–2023 and analyzed the long-term trends of SWE. In general, the ERA5-Land reanalysis reproduces SWE in the Perm region satisfactorily. Mean relative error for SWE in March does not exceed 15%. The average correlation coefficient between the reanalysis data and the same from the observations is 0.72 for non-forest locations and 0.83 for locations in forest. In the southern part of the region, the reanalysis does mainly overestimate SWE by 10–40 mm, while in the north and east of the territory, there is an underestimation of the same order. The greatest divergence between snow surveys and reanalysis are found during snowmelt season, especially for non-forest snow-measuring routes. As it follows the ERA5-Land data, average date of formation of the SWE maximum in the southern part of the region is close to March 25, and in the eastern mountainous part it falls on the second decade of April, which is 4–7 days later than according to snow surveys in the forest. According to the ERA5-Land data and observations, a statistically significant negative trend of SWE was revealed all over the territory in the first half of the cold season, especially pronounced in November. It is related to the autumn warming and a shift of snow cover onset to later dates. In March, the negative trend according to the ERA5 data is statistically significant only in the southern part of the region, where it reaches –12 mm/10 years, but no statistically significant decrease in SWE is found according to the snow survey data. In May, a significant reduction of SWE in the northeast of the region (up to 15 mm/10 years) is found, which is due to the warming in April and May, and an earlier start of snowmelt. A comparison with the snow survey data shows that the reanalysis reproduces well the inter-annual variability of SWE accumulated by March, especially in forest locations. A statistically significant increase in SWE was revealed on five snow measuring routes in field, while a statistically significant decrease – on two forest routes, which is not confirmed by the reanalysis data. These discrepancies may be related to changes in local snow accumulation conditions on snow-measuring routes.