| Thermal state of soils is a composite product of energy and mass exchange between the atmosphere and land surface. Soil temperature is a sensitive indicator to the changing climate, and plays a critical role in all geophysical and geochemistrical processes. Understanding soil thermal conditions and their relationships to climatic factors have received substantial attentions during the several decades. Recent studies show that global temperature changes occurred associating with significantly regional and seasonal diversities. Key questions are what is the impact of these differences on the thermal states of the soils? For this purpose, this study employed synthetic gridded climate datasets, meteorological stations data and numerical model to explore the regional climate changes from a novel perspective. We then investigated the responses of the near-surface soil freeze-thaw cycles and thermal states of permafrost to the climate change. Main conclusions are summarized as follows:At global scales, climatology(1961-1990) of mean annual air temperature can depict better the spatial characteristics of the evolution in global land surface air temperature from 1951 through 2013. Land with climatology of annual mean air temperature between-15 ℃ and 26 ℃ occupied about 81% of the global land surface, where is represented as a good negative correlation between warming magnitudes and climatology of mean annual air temperature. In general, warming magnitude decreases 0.028 ℃ associating with an increase of 1 ℃ in the climatology of mean annual air temperature. At a monthly scale, the results showed that cold-season(December~ February) mean air temperature increased by more than 2.8 ℃ in the regions with climatology of mean annual air temperature between-20 ℃ and 10 ℃. These regions experienced a long period warming of up to eight months(from October through May). In addition, we defined the cold regions as the climatology of mean annual air temperature below 0 ℃ and evaluated the regional climate change. It revealed that cold regions experienced intensified warming of 1.68±0.19 ℃, while the rest regions experienced relatively weakened warming of 1.04±0.09 ℃.At regional scales, using 0 cm soil temperature observations in 636 meteorological stations across China, this study discusses the climatology of the timing and duration in the near-surface soil freeze-thaw from July, 1961 through June, 1991 and their changes for the period from 1956 through 2006. Meanwhile, we discussed their correlations to climate change. Climatology of the timing and duration in the nearsurface soil freeze-thaw indicated that: duration of the near-surface soil freeze was the longest on the Qinghai-Tibetan Plateau, i.e., it can freeze up in almost everyday over a year due mainly to its high-altitude and unique plateau environment. In high-latitude regions, near-surface soil freeze started about in September and ended in June the following year. Climatology of the near-surface soil freeze represented as a significant latitude zonal distribution in eastern China, and controlled mainly by terrain in western China. Number of freeze day is correlated to permafrost regions map while the threshold is different for high-latitude and high-altitude permafrost regions. Boundaries of high-altitude permafrost regions are agreed with the contour of the number of freeze day of 260 days, and boundaries of high-latitude permafrost regions are agreed with the contour of 220 days. By comparative analysis of remote sensing data and ground-based observations, we found that microwave remote sensing inversion of near-surface soil freeze days is a big difference in the ground-based data. Using ground-based observations as the benchmark, remote sensing data overestimated when long-term average number of freeze day below 66 days, while the rest regions represented as underestimation. The most significant differences are found on or around the QinghaiTibet Plateau.The first date of the near-surface soil freeze was delayed by about 5 days, or at a rate of 0.10 ± 0.03 day/year, and the last date was advanced by about 7 days, or at a rate of 0.15 ± 0.02 day/year. The duration of the near-surface soil freeze decreased by about 12 days or at a rate of 0.25 ± 0.04 day/year, while the actual number of the near-surface soil freeze days was decreased by about 10 days or at a rate of 0.20±0.03 day/year. Warming in spring is more obvious than that in fall. Changes in near-surface soil freeze time in western China are enhanced with the increasing altitude, i.e., more rapid changes occurred in high-altitude regions. Magnitude of the trend in the first date of the near-surface soil freeze increased by 0.095 day/year with a rise of 1000 m in altitude, meanwhile magnitude of the trend in the number of freeze days increased by 0.035 day/year. Changes in timing and duration of the near-surface soil freeze have a negative correlation with latitude in eastern China. Associating with an increase of 1° in latitude, magnitude of the trend in the last date of the near-surface soil freeze decreased 0.005 day/year, magnitude of the trend in duration decreased by 0.006 day/year, and magnitude of the trend in duration decreased by 0.007 day/year. Changes in the nearsurface soil freeze/thaw status were primarily controlled by changes in air temperature, but urbanization may also play an important role. The trend of the number of freeze days was about-0.27 day/year in the regions with a high rate of urban expansion from 1956 through 2006, while it showed a statistically non-significant change since 1990. Further analysis indicated that after 1990, the number of freeze days in the regions with the lower rate of urban expansion decreased at a rate of about 0.85 day/year.In cold regions, permafrost thermal states also play an important role in indicating the changing climate. This study presents a statistical method to diagnose the existence of permafrost in the near-surface soil based on soil temperatures observations. It was applied to hydrometeorological stations data, then diagnosed 28 stations with permafrost in the near-surface soil. We analyzed the spatial characteristics of long-term average states of the soil temperatures downward to 3.20 m. Climatology of soil temperatures at five standard depths(0.20 m, 0.40 m, 0.80 m, 1.60 m and 3.20 m) represented better statistical relationships with altitudes and latitudes. Generally, latitude and elevation can jointly explain 50~70% of variations in spatial characteristics of soil temperatures, while latitude plays a leading role. Magnitudes of warming in soil temperatures decreased generally with depth, which were 0.38 ℃ per decade, 0.34 ℃ per decade, 0.28 ℃ per decade, 0.21 ℃ per decade and 0.24 ℃ per decade, respectively at five standard depths. At a monthly scale, there is a significant seasonal difference in soil temperatures. There was enhanced warming in the cold months(January-March), while no statistically significant trend in the warm months(JulySeptember). However, the season difference was small in deep soils. Especially at the depth of 3.20 m, monthly trends ranged from 0.16 to 0.32 ℃ per decade. Meanwhile, there were eight months with a trend above 0.20 ℃ per decade.Using Monte Carlo stochastic modeling method, GIPL model was successfully applied in the sample site(Isit station) without soil property in the Siberian permafrost regions, and achieved good results. Taking Russian Isit station for an example, the rootmean-square errors between simulated and observed soil temperatures were up to 1 ℃ over the calibration period. Sensitivity analysis showed that the annual mean temperature variation influences soil thermal states most strongly. In addition, changes in air temperature during the cold-season play a significant role in soil temperatures changes, followed by changes in snow depth and warm-season air temperature. Influences of changes of ±1 standard deviation in cold-season air temperature, snow depth and warm-season air temperature on the soil temperature at the depth of 15 m are about 3.31, 2.44, and 1.52 ℃, respectively. Active layer thickness is mostly influenced by warm-season air temperature changes, i.e., change of ±1 standard deviation in warm-season air temperature will lead to a change of 0.43 m in active layer thickness. Meanwhile, the influences of snow depth and air temperature on active layer thickness represent a strongly non-linear characteristics. Generally, magnitudes of change in active layer thickness caused by reducing air temperature are smaller than that caused by increasing air temperature. Combining variations of snow depth and cold-season air temperature make a strongly non-linear variations in active layer thickness. When snow cover is thin, cold-season air temperature influences less on active layer thickness. Changes of ±1 standard deviation in cold-season air temperature lead to a change of 0.19 m and 0.66 m in active layer thickness under the scenarios of thin snow cover and thick snow cover, respectively. Similarly, changes of ±1 standard deviation in snow depth lead to a change of 0.10 m and 0.57 m in active layer thickness under the scenarios of low and high cold-season air temperature, respectively.In conclusion, more rapid warming occurs in cold regions and cold season, which is critical to the sensitive cryosphere. Changes in soil thermal states closing to the ground surface present a significant seasonal diversities, which is consistent with the characteristics of air temperature changes, i.e., non-significant warming in summer while significant warming in winter. This may result in lengthening of the growth season and lead to alteration in ecosystem structures, then influence the energy and water exchanges between the ground and atmosphere. Snow variations are not only an important factor to predict the soil thermal conditions, and their feedback to climatic system is also a critical aspect in understanding the physical processes. It should be comprehensive considering the effects of air temperature, snow cover and soil thermal conditions in predicting ecological, agricultural, hydrological processes under a changing climate. However, their interactions and feedback mechanisms are lacking in detailed investigation, and are not described well in land surface models and global climate models. |
PDF Full Text Request