Snowmelt water in spring is an important source of soil water, which is critical to supporting crop growth. Quantifying the contribution of snowmelt water to soil water and its dynamic changes is essential for evaluating soil moisture and allocating agricultural water resources. In this paper, through controlled outdoor experiments, different snow depths and soil depth gradients were set; and snow, precipitation, and soil samples were collected regularly. To analyze the contribution of snowmelt water to soil water and its dynamic changes, the MAT-253 stable isotope ratio mass spectrometer was adopted for hydrogen and oxygen isotope analyses. The results showed that the snowmelt water for snow depths of 10 cm, 30 cm, and 50 cm all contributed to the 0–30 cm soil layer. The contribution increased with soil depth, contributing 8.13%, 8.55%, and 11.24%, respectively. The contribution of the snow cover at the same depth to the soil moisture at different depths also varied, i.e., the contribution increased with increasing soil depth. The snowmelt water retention time at depths of 10 cm, 30 cm, and 50 cm was inconsistent, i.e., it was the longest at 0–10 cm (average of 69 days), followed by 20–30 cm (average of 59 days), and the shortest at 10–20 cm (average of 54 days). The greater the snow depth, the shorter the retention time of the snowmelt water in the different soil layers. For surface soil, the contribution of the snowmelt water at greater depths was significantly different; while for deep soil, the contribution was more sensitive to the snow depth. Regardless of snow depth, soil contributions at different depths were significantly different. Precipitation also affected the contribution of the snowmelt water to the soil water, exhibiting different effects at different depths.

In the context of global warming, a key scientific question for the sustainable development of the Arctic tourism industry is whether the region’s climate is becoming more suitable for tourism. Based on the ERA5-HEAT (Human thErmAl comforT) dataset from the European Center for Medium-range Weather Forecasts (ECMWF), this study used statistical methods such as climatic tendency rate and RAPS to analyze the spatial-temporal changes in Arctic summer climate comfort zones from 1979 to 2019 and to explore the influence of changes in climate comfort on Arctic tourism. The results showed the following: (1) With the increase in the Arctic summer temperature, the universal thermal climate index (UTCI) rose significantly from 1979 to 2019 at a rate of 0.457 °C/10a. There was an abrupt change in 2001, when the climate comfort changed from “colder” to “cool”, and the climate comfort has remained cool over the past decade (2010–2019). (2) With the increase in Arctic summer temperatures, the area assessed as “comfortable” increased significantly from 1979 to 2019 at a rate of 2.114 × 10⁵ km²/10a. Compared with the comfortable area in the 1980s, the comfortable area increased by 6.353 × 10⁵ km² over the past 10 years and expanded to high-latitude and high-altitude areas, mainly in Kola Peninsula, Putorana Plateau, and Verkhoyansk Mountains in Russia, as well as the Brooks Mountains in Alaska. (3) With the increase in Arctic summer temperatures, the number of days rated comfortable on 30% of the grid increased significantly from 1979 to 2019 (maximum increase: 31 days). The spatial range of the area with a low level of comfortable days narrowed and the spatial range of the area with a high level of such days expanded. The area with 60–70 comfortable days increased the most (4.57 × 10⁵ km²). The results of this study suggest that global warming exerts a significant influence on the Arctic summer climate comfort level and provides favorable conditions for further development of regional tourism resources.

In this study, on the basis of the temperature data collected at 612 meteorological stations in China from 1961 to 2019, cold regions were defined using three indicators: an average temperature of <−3.0 °C during the coldest month; less than five months with an average temperature of >10 °C; and an annual average temperature of ≤5 °C. Spatial interpolation, spatial superposition, a trend analysis, and a spatial similarity analysis were used to obtain the spatial distribution of the cold regions in China from 1961 to 2019. Then, the areas of the cold regions and the spatial change characteristics were analyzed. The results reveal that the average area of the cold regions in China from 1961 to 2019 was about 427.70 × 10⁴ km², accounting for about 44.5% of the total land area. The rate of change of the area of the cold regions from 1961 to 2019 was −14.272 × 10⁴ km²/10 a, exhibiting a very significant decreasing trend. On the basis of the changes between 1991–2019 and 1961–1990, the area of China’s cold regions decreased by 49.32 × 10⁴ km². The findings of this study provide references for studying changes in the natural environment due to climate change, as well as for studying changes on a global scale.

Snow depth is a critical parameter for characterizing snow dynamics and water resources, and its accurate inversion is essential for hydrological processes, climate studies, and disaster prevention in cold regions. Based on long-term daily ground meteorological observation data from the hydrological years 1961 to 2015 at two meteorological stations in Mohe and Mishan, Heilongjiang Province, China, this study integrates physical parameters of snow density and snow albedo from the ERA5-Land reanalysis data to systematically compare the performance of three machine learning and three deep learning models in retrieving daily snow depth. Four feature combination schemes were designed to evaluate the contributions of meteorological factors, lagged snow depth terms, and snow physical parameters. The results indicate that, for both machine learning and deep learning models, the first-order lagged value of snow depth is the most important variable determining prediction accuracy. In terms of model performance, machine learning methods excelled, with XGBoost performing particularly outstandingly, achieving optimal prediction accuracy and stability under the best feature combination (coefficient of determination, R², reaching 0.989; root mean square error, RMSE, of 1.19 cm). Among deep learning methods, 1D CNN demonstrated strong local feature extraction capabilities, achieving prediction accuracy comparable to the best-performing machine learning model (R² of 0.9878, RMSE of 1.26 cm). Notably, models specifically designed for time-series data, such as LSTM (R² of 0.9848, RMSE of 1.41 cm), and the more complex 1D CNN-LSTM hybrid model (R² of 0.9803, RMSE of 1.60 cm) did not show significant advantages in this study. This indicates that model complexity and predictive performance are not simply positively correlated. Through comprehensive analysis of data from both stations, this study demonstrates that a prediction framework centered on historical snow depth as the core driving factor, combined with key meteorological elements, is highly robust. Although the inclusion of ERA5-Land snow physical parameters did not significantly improve model accuracy, it provides important insights for the future development of hybrid models that integrate physical mechanisms with data-driven approaches. The findings offer an effective solution for reconstructing long-term snow depth time series and hold significant application value for simulating cryospheric hydrological processes and studying climate change.