[go: up one dir, main page]
Next Article in Journal
Ecological Function Zoning Framework for Small Watershed Ecosystem Services Based on Multivariate Analysis from a Scale Perspective
Previous Article in Journal
Analyzing Spatial Distribution and Influencing Factors of Soil Organic Matter in Cultivated Land of Northeast China: Implications for Black Soil Protection
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Land
Volume 13
Issue 7
10.3390/land13071029
land-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Assessment of Soil Temperature and Its Change Trends in the Permafrost Regions of the Northern Hemisphere

by
Yifan Wu
1,2,
Guojie Hu
1,2,*,
Lin Zhao
3,
Defu Zou
1,
Xiaofan Zhu
1,
Yao Xiao
1,
Tonghua Wu
1,
Xiaodong Wu
1,
Youqi Su
4 and
Rui Zhang
5
1
Cryosphere Research Station on Qinghai-Tibet Plateau, Key Laboratory of Cryospheric Science and Frozen Soil Engineering, Northwest Institute of Eco-Environment and Resources, Chinese Academy of Sciences, Lanzhou 730000, China
2
University of Chinese Academy of Sciences, Beijing 100049, China
3
School of Geographical Sciences, Nanjing University of Information Science & Technology, Nanjing 210044, China
4
School of Atmospheric Sciences, Chengdu University of Information Technology, Chengdu 610000, China
5
College of Water Conservancy and Hydropower Engineering, Gansu Agricultural University, Lanzhou 730000, China
*
Author to whom correspondence should be addressed.
Land 2024, 13(7), 1029; https://doi.org/10.3390/land13071029
Submission received: 9 May 2024 / Revised: 28 June 2024 / Accepted: 6 July 2024 / Published: 9 July 2024
Browse Figures
Figure 1
<p>Spatial distribution diagram of temperature observation stations in permafrost areas in the Northern Hemisphere.</p> "> Figure 2
<p>Spatial distribution of annual average soil temperature among observation stations in permafrost areas of the Northern Hemisphere. (<b>a</b>), (<b>b</b>), (<b>c</b>) and (<b>d</b>) represent 0–50 cm, 50–100 cm, 100–200 cm, and 200–300 cm, respectively.</p> "> Figure 3
<p>Taylor plot of soil temperature simulated and observed for different CMIP 6 modes from 2003 to 2012. (<b>a</b>), (<b>b</b>), (<b>c</b>) and (<b>d</b>) represent 0–50 cm, 50–100 cm, 100–200 cm, and 200–300 cm, respectively.</p> "> Figure 4
<p>Observed and simulated soil temperature trends in the Northern Hemisphere from 2003 to 2012. (<b>a</b>), (<b>b</b>), (<b>c</b>) and (<b>d</b>) represent 0–50 cm, 50–100 cm, 100–200 cm, 200–300 cm, respectively.</p> "> Figure 5
<p>Rate of soil temperature change in the permafrost region of the Northern Hemisphere from 1900 to 2014. (<b>a</b>), (<b>b</b>), (<b>c</b>) and (<b>d</b>) represent 0–50 cm, 50–100 cm, 100–200 cm, 200–300 cm, respectively.</p> "> Figure 5 Cont.
<p>Rate of soil temperature change in the permafrost region of the Northern Hemisphere from 1900 to 2014. (<b>a</b>), (<b>b</b>), (<b>c</b>) and (<b>d</b>) represent 0–50 cm, 50–100 cm, 100–200 cm, 200–300 cm, respectively.</p> "> Figure 6
<p>Trend of soil temperature in Northern Hemisphere permafrost from 1900 to 2014 and 1950 to 2014. (<b>a</b>), (<b>b</b>), (<b>c</b>) and (<b>d</b>) represent 0–50 cm, 50–100 cm, 100–200 cm, 200–300 cm, respectively.</p> ">
Versions Notes

Abstract

:
In this paper, we used data from 42 soil temperature observation sites in permafrost regions throughout the Northern Hemisphere to analyze the characteristics and variability in soil temperature. The observation data were used to evaluate soil temperature simulations at different depths from 10 CMIP6 models in the permafrost region of the Northern Hemisphere. The results showed that the annual average soil temperature in the permafrost regions in the Northern Hemisphere gradually decreased with increasing latitude, and the soil temperature gradually decreased with depth. The average soil temperatures at different depths were mainly concentrated around 0 °C. The 10 CMIP6 models performed well in simulating soil temperature, but most models tended to underestimate temperatures compared to the measured values. Overall, the CESM2 model yielded the best simulation results, whereas the CNRM-CM6-1 model performed the worst. The change trends in annual average soil temperature across the 42 sites ranged from −0.17 °C/10a to 0.41 °C/10a from 1900 to 2014, the closer to the Arctic, the faster the soil warming rate. The rate of soil temperature change also varied at different depths between 1900–2014 and 1980–2014. The rate of soil temperature change from 1980 to 2014 was approximately three times greater than that from 1900 to 2014.

1. Introduction

Permafrost is defined as ground whose temperature remains at or below 0 °C for at least two consecutive years [1]. Permafrost is mainly distributed in high-latitude and high-altitude areas and accounts for about 22% of the land area of the Northern Hemisphere [2]. With climate warming, permafrost has undergone extensive degradation, mainly manifesting as soil temperature increases, active layer thickening, and ground ice melting [3]. With permafrost degradation, the decomposition of soil organic matter increases, releasing carbon dioxide and methane into the atmosphere and forming a positive feedback loop with climate change [4]. In addition, permafrost degradation also affects ecosystems and surface and underground hydrological processes [5,6] and can even cause ground subsidence, affecting the stability of ground infrastructure [7,8]. Therefore, the study of permafrost change in the context of climate warming has received much attention.
Soil temperature is one of the key indicators used to study the thermal conditions of permafrost. Research has shown that permafrost temperatures in the Northern Hemisphere range between −15 °C and 0 °C. In terms of spatial distribution, the permafrost temperatures are low in the Arctic region and gradually increase moving south, which means the average soil temperatures in most areas of the Qinghai-Tibet Plateau are relatively high [9]. Since the 1980s, the soil temperatures in the permafrost areas of the Northern Hemisphere have increased, and with permafrost degradation, the thickness and range of permafrost have been greatly reduced [10,11]. However, the amplitudes and rates of soil temperature changes vary among different regions. In terms of warming rate, those in the permafrost areas in northern Alaska and western Russia (northern Europe) were the most pronounced, with permafrost warming rates of 0.6~3 °C and 0.3~2 °C in northern Alaska and western Russia, respectively [12]. The temperature of the permafrost in northern Alaska has been increasing at a rate of 0.1 °C/a at the 20 m depth since 1984 [13]. These changes were characterized by sharp increases in surface soil temperature in winter and by shallow and deep soil warming at faster rates in spring and summer than the surface soil [14,15]. At the same time, studies have predicted that the soil temperature in the Northern Hemisphere will increase in the future with climate warming [16,17]. However, how soil temperature change will manifest in the Northern Hemisphere remains unknown. In permafrost areas, due to the geography, extreme climate, and other challenging factors, long-term, wide-scale soil temperature observation data are limited and patchy, so reanalysis data have been widely utilized because they have the advantages of being long-term and spatially continuous, which help make up for the deficiencies in the observation data [18,19]. Some scholars have used the reanalysis data and model simulation results to evaluate simulated changes in permafrost [20]. Furthermore, scholars have used observation data to test the applicability of surface temperature reanalysis data in the Qinghai-Tibet Plateau permafrost area, with the results showing that the data accurately reflected the basic surface temperature characteristics of the plateau permafrost area but with some overestimates of temperatures [21]. Although reanalysis data uses the real-time global data assimilation system and considers surface, radio, air, and satellite observation data, there will inevitably be errors in the reanalysis data when compared with measured data, especially in areas with complex terrain [22]. Therefore, reanalysis data may not accurately reflect the real situation on the ground [23,24,25,26].
The Coupled Model Intercomparison Project (CMIP) has built the most extensive climate modeling database, which provides a scientific basis for the study of the role of permafrost and climate change [27,28,29,30]. Compared with previous projects, CMIP6 contains more models, more perfectly designed scientific tests, and includes more data [31]. At present, many scholars have used the CMIP6 model data to evaluate the temperature and precipitation in different regions. Generally, the results have produced overestimates of precipitation for most patterns and underestimates of air temperatures [32,33]. The CMIP6 model has been shown to have a reasonable ability to simulate the spatial distribution of annual precipitation in the Northern Hemisphere and permafrost areas, but compared with the observed data, the CMIP6 model overestimated precipitation in those areas by 11% and 42%, respectively [34]. Compared to precipitation, the CMIP6 model is better at simulating temperature [35]. The simulation results showed that most models from CMIP6 were able to accurately simulate temperature change in the Northern Hemisphere [36]. A comparison of the ERA 5 reanalysis data with 22 CMIP6 models using average Arctic near-surface temperature showed that most CMIP6 models underestimated average Arctic temperatures from 1979 to 2014 [37]. Currently, there are no reports on the applicability of the CMIP6 model for simulating soil temperatures in permafrost areas.
In summary, many previous studies have estimated the patterns and changes in permafrost, but most have focused on simulations of temperature and precipitation, and many have used simulated patterns based on reanalysis data. However, relatively few studies have evaluated soil temperature patterns in the Northern Hemisphere. Based on this, this paper established three primary objectives: first, to reveal the regional differences in soil temperature in the permafrost in the Northern Hemisphere using data from 10 CMIP 6 models and observation data from 42 sites to analyze and characterize the soil temperature of the permafrost; second, to evaluate the simulation results of each model in the Northern Hemisphere; and finally, to analyze the trends in soil temperatures in Northern Hemisphere permafrost using the results of the optimal model.

2. Data and Methods

2.1. Data Sources

The data used in this study are mainly divided into two parts. The first part covers the observed soil temperature data in the Northern Hemisphere, and the other part covers the model simulation data from CMIP6. The measured data were selected from temperature observation stations located in the permafrost region of the Northern Hemisphere [2]. To ensure the data quality, sites with missing data for 12 consecutive months were excluded. Finally, soil temperature data from 42 observatories from 2003 to 2012 were used (Figure 1). The data from the 32 daily soil temperature stations in Russia, Alaska, and Svalbard were downloaded from the Global Permafrost Land Network (GTN-P) database (http://gtnpdatabase.org, accessed on 9 October 2023). The data from the seven daily soil temperature stations in the permafrost region of Northeast China were obtained from the Genhe ground temperature observation dataset of the China Tibetan Plateau Data Center [38,39]. The soil temperature data of the three stations on the Qinghai-Tibet Plateau are the soil temperature observation data of the Tanggula station and the soil temperature and humidity monitoring data of the China04 and China06 stations, respectively [40].
The historical simulation data (1900–2014) of the CMIP6 model were downloaded from https://esgf-node.llnl.gov/search/cmip6/ (accessed on 5 May 2024). To facilitate the comparison with the measured data and evaluate the simulation ability of the CMIP6 model, 10 models were selected with the same time series to coincide with the observed data time series (2003–2012) and with observation depths and stratification patterns that were similar. Metadata about the datasets, including land surface model, institute, soil layer, depth, and stratification, are provided in Table 1. To evaluate the simulation capability of the models, it was necessary to compare model data from the same segment and the same depth as the measured data. Based on the model depth and observation depth, overlapping or similar depths were selected for analysis. Meanwhile, to analyze soil temperature changes at different depths, soil depths were divided into four layers, namely 0–50 cm, 50–100 cm, 100–200 cm, and 200–300 cm.

2.2. Taylor Diagrams Analysis Method

The abilities of the 10 models to simulate soil temperature were assessed using Taylor diagrams analysis. The Taylor diagram is a polar map composed of the correlation coefficient and mean–variance ratio of the spatial distribution of simulated temperature and observed temperature. Four evaluation indicators were used to evaluate the simulation quality of the 10 CMIP6 models, namely mean absolute error (MAEs), root mean square error (RMSEs), standard deviation (SD), and the correlation coefficient (R) between the model and observed data. Integrated model rank (MR) was used to assess the accuracy of each model simulation. The MR is defined as follows:
M R = 1 − 1 1 × m × n ∑ i = 1 n r a n k i
where m is the number of models; n is the number of evaluation indicators, and ranki is the ranking of each index based on the model. The closer the MR is to 1, the better the ability of the model.

3. Results

3.1. Spatial Distribution Characteristics of Soil Temperature

The spatial distributions of the annual average soil temperatures at 0–50 cm, 50–100 cm, 100–200 cm, and 200–300 cm in the Northern Hemisphere are shown in Figure 2, where Figure 2a is the spatial distribution of annual average soil temperature in the 0–50 cm layer. All 42 sites had data at this depth, with an annual average soil temperature of −0.07 °C, a maximum of 11.93 °C and a minimum of −13.03 °C, and a range of 24.96 °C. As seen from the figure, the average annual soil temperature was below −6 °C at most sites near the Arctic, and the annual average soil temperature increased with decreasing latitude. The frequency distribution of annual mean soil temperature showed that 14% of sites had annual average soil temperatures of less than −7 °C, while most (67%) sites had annual average soil temperatures between −4 and 2 °C and 19% had average annual soil temperatures greater than 5 °C.
Figure 2b shows the spatial distribution of the annual average soil temperatures at 50–100 cm; 40 sites had observation depths within this range. The mean annual soil temperature at these sites was −1.31 °C, ranging from a minimum of −10.35 °C to a maximum of 7.55 °C, with an extreme range of 17.9 °C. The spatial characteristics of the annual average soil temperatures in this layer were similar to 0–50 cm but were slightly different in frequency distribution, with the average temperature at most 50–100 cm sites around −1 °C and only 27.5% less than −7 °C and 17.5% greater than 5 °C.
Figure 2c shows the spatial distribution of the annual average soil temperatures at 100–200 cm; 32 sites had data for this depth. The average annual mean soil temperature at these sites was −1.32 °C, with a maximum of 1.41 °C, a minimum of −6.55 °C, and an extreme range of 7.96 °C. Most (69%) of sites in this layer had an average annual soil temperature of around 0.5 °C, 28% of sites had an average annual low temperature of less than −2.5 °C, and only 3% of sites had averages greater than 1.5 °C.
Figure 2d shows the spatial distribution of the annual average soil temperatures at 200–300 cm; 20 stations had observation depths within this range. The average annual mean soil temperature at these sites was −1.04 °C with a maximum of 1.32 °C a minimum of −6.25 °C, and a range of 7.57 °C. Most (70%) of sites in this layer had annual average soil temperatures of about 0.5 °C, 25% of sites had annual average low temperatures of less than −2.5 °C, and only 5% of sites had average temperatures greater than 1.5 °C.
Overall, the closer the site was to the Arctic, the lower its average annual soil temperature. With depth, the amplitude of the annual mean soil temperature gradually decreased, as did the spatial differences. The spatial distributions of annual average soil temperatures at different depths did not differ much. Most of the soil temperatures of 0–50 cm and 50–100 cm were concentrated around −1 °C, and those of 100–200 cm and 200–300 cm were concentrated around −0.5 °C.

3.2. Assessment of CMIP6 Models

Ten CMIP6 models were evaluated using month-by-month soil temperature data and the coinciding time series simulation data from 42 Northern Hemisphere sites, i.e., from 2003 to 2012. Most of the models exhibited good simulation power for soil temperature. On the whole, the correlation coefficients were all greater than 0.9, except for MPI-ESM-1-2-HAM and NorESM2-MM, indicating that most models correlated well with the observed values. The RMSEs for all models ranged between 2.19 and 9.62 °C, with a mean of 6.17 °C; the highest error was observed in IPSL-CM5A2-INCA and the lowest in IPLS-CM6A-LR-INCA. The standard deviation ranged between 4.96 and 10.08 °C, with a mean of 6.7 °C; the largest standard deviation was observed in IPSL-CM5A2-INCA2 and the smallest in CESM2. The mean absolute error (MAE) shows the degree of variation between models, with the MAE of all models ranging between 1.72 °C and 8.74 °C and a mean of 4.9 °C.
Although different models differed in their ability to simulate soil temperature, all but IPSL-CM5A2-INCA and CNRM-CM6-1 produced simulations that accurately reflected the actual soil temperature (Figure 3). The models with highest accuracy at 0–50 cm were CESM2, NorESM2-LM, and CESM2-WACCM; the most accurate at 50–100 cm were IPLS-CM6A-LR-INCA, NorESM2-LM, and CESM2; the most accurate at 100–200 cm were IPLS-CM6A-LR-INCA, MPI-ESM-1-2-HAM, and CESM2; and the most accurate at 200–300 cm were CESM2, IPLS-CM6A-LR-INCA, and NorESM2-MM.
The change trends in soil temperature were analyzed using the CMIP6 simulated soil temperature data at different depths at the 42 stations in the Northern Hemisphere from 2003 to 2012 and the measured annual mean ground temperature data during the same period (Figure 4). Overall, of the ten models, IPLS-CM6A-LR-INCA was the most similar to the observed trends, while KIOST-ESM was the most different from the observed data. This indicated that the IPLS-CM6A-LR-INCA model is best suited for simulating soil temperature in the Northern Hemisphere. CESM2, CESM2-WACCM, NorESM2-LM, and NorESM2-MM also produced trends that were close to the measured values, indicating that these models have good simulation ability. From the perspective of different depths, the change trends of annual average ground temperature in each model were basically the same, but in the 0–50 cm and 50–100 cm layers, the IPLS-CM6A-LR-INCA model produced overestimates with higher average annual ground temperatures than the actual observed values.
A comprehensive ranking of the 10 simulation patterns was performed using the MR values (Table 2). There were large differences in the simulations of the different models. From the comprehensive ranking of each model, the CESM2 model was revealed as having the best simulation effect, while CNRM-CM6-1 was the worst; the top three models were CESM2, CESM2-WACCM, and IPSL-CM6A-LR-INCA, indicating that the simulated results from these models were closest to the observed data. Except for the IPSL-CM5A2-INCA model (less effective for the first three depths but better for 200–300 cm), there were minimal differences among depths for any given model.
After selecting the models that were ranked first in each layer (AME1) and the top three models overall (AME3), the simulated MAE3, R, RMSE, and MAE values were averaged and compared with MAE1 (Table 3). The correlation coefficient between the optimal model data and the observed data was much larger than many of the other models, except for the 100–200 cm layer, where the correlation coefficient of AME3 was larger than that of AME1. For 0–50 cm and 100–200 cm, the mean absolute error of AME3 was better than AME1, but in the 50–100 cm and 200–300 cm layers, AME1 had smaller RMSE and mean absolute error values. Overall, the simulation obtained after averaging the top three optimal models was better, which indicated that multiple-model averaging could effectively reduce the uncertainty of any single-model simulation in CMIP6. Therefore, AME3 was used to analyze the soil temperature variation among different sites in this study.

3.3. Changes in Soil Temperature at Different Depths between Different Sites

The average annual rate of soil temperature change in permafrost areas from 1900 to 2014 was analyzed using the optimal model (MAE3) (Figure 5). The change trends in the annual average soil temperatures of the 42 sites ranged between −0.17 and 0.41 °C/10a. Most (98%) of the sites showed significant warming trends (p < 0.05), and the rates of change were greater at sites closer to the Arctic. The warming rates in continuous permafrost areas were higher than in discontinuous, isolated, and sporadic permafrost zones. The sites with the greatest rates of change were all in Russia, while those with the smallest rates of change were concentrated in Alaska and the Tibetan Plateau.

3.4. Trends of Soil Temperature in the Northern Hemisphere

The optimal CMIP6 model was used to analyze the soil temperature in the Northern Hemisphere from 1900 to 2014 and from 1980 to 2014 (Figure 6). Clearly, the soil temperatures during these periods exhibited trends of gradual warming. From 1900 to 2014, the soil temperature warming rate was 0.17 °C/10a, but the rates of change at different depths were different, with the warming rate decreasing with increasing soil depth. From 1980 to 2014, the soil temperature warming rate was 0.57 °C/10a, with the warming rate decreasing with increasing depth. In general, the rate of soil temperature change from 1980 to 2014 was greater than from 1900 to 2014, and the difference in the warming rate was 0.4 °C/10a, indicating that the change in soil temperature has significantly accelerated in the past 35 years.

4. Discussion

This study showed that the soil temperature in the permafrost region of the Northern Hemisphere increased from 1900 to 2014. The rate of soil temperature change at the 42 sites in the Northern Hemisphere ranged from −0.17 to 0.41 °C/10a, and the annual average soil temperatures of sites gradually decreased with increasing latitude. Compared with previous studies, the rate of soil temperature change was greater [36], which may have been due to the different time scales and data used. There were regional differences in the rates of soil temperature change in the permafrost regions of the Northern Hemisphere. In the high latitudes of Eurasia, the soil temperatures at the 20 cm, 40 cm, 80 cm, 160 cm, and 320 cm depths increased between 1963 and 2012 at rates of 0.2 °C/10a, 0.28 °C/10a, 0.28 °C/10a, 0.28 °C/10a, and 0.27 °C/10a, respectively [41]. The soil temperature warming rate in the Northern Hemisphere from 1950 to 2014 was 0.30 °C/10a, which was basically consistent with the rate of soil temperature in Eurasia. In the Tibetan Plateau, the spatial variation in soil temperature warming trends at different depths was relatively consistent, and the soil temperature warming trends from 1960 to 2021 were less pronounced than those from 1980 to 2021 [42]. Our study has also reached a similar conclusion. The soil temperature of the 0–10 cm layer increased at rates between 0.1–0.3 °C/10a from 2000 to 2016, and when compared with the temperature change rate from 1900 to 2014 (0.07 °C/10a, based on the simulation by the optimal model in this paper), the more recent rate of change on the Tibetan Plateau is larger, indicating that the soil is warming faster in recent years [43]. Differences in solar radiation received at different elevations, snow cover durations, and vegetation types may have contributed to the differences in soil temperature change rate among the different regions in the Northern Hemisphere [44,45,46,47,48,49].
All 10 CMIP6 models examined in this paper were able to simulate the change in soil temperature in the Northern Hemisphere permafrost area, but the simulation results of different models were different or even very different. The simulation from individual models was also quite different from the actual observations, which may have been due to models often relying on simplifications and assumptions about complex natural processes. Sometimes, it cannot fully reflect reality. The selection and setting of parametric schemes with different spatial and temporal resolutions are also the reasons for the differences in observation and model results [50,51]. In past studies, multi-model ensemble averages have commonly been used to eliminate single-model “noise”, effectively reducing the uncertainty of estimated results [52,53,54,55,56]. In this study, the three models with the best simulation effect (optimal model) were selected for ensemble averaging, and the correlation coefficient between the optimal model and the observed data was greatly improved compared with any single model, which was consistent with the results of previous studies. In terms of simulation ability ranking, the top three models in this study were CESM2, CESM2-WACCM, and IPSL-CM6A-LR-INCA, while CNRM-CM6-1 had the worst simulation effect. CESM2-WACCM also had relatively high effectiveness in a previous study because the CESM2-WACCM integrates multiple highly complex physical and chemical processes and external forcing and has a high spatiotemporal resolution. Hence, the simulation of the local or regional temperature is more accurate [33,57]. Interestingly, CESM2 produced the best simulation in this study, while it was not ranked highly in previous studies [35,36,37]. The differences in the ranking may have been produced by differences in the data used and the research objects. The previous studies used reanalysis data, and their research objects were temperature [35,36,58], while in this paper, measured data were used, and the research objects were soil temperature at different depths.
By comparing the change trends of the simulated and measured data, it was found that the 10 CMIP6 simulation models exhibited certain degrees of systematic model deviation (Figure 4). Most of the simulated soil temperatures were lower than the observed values, i.e., there was cold deviation. Previous studies have made similar observations [52,59,60,61]. In northern permafrost, studies have shown that the cold deviation is greater in winter than in other seasons, which may be due to the higher albedo due to snow cover, resulting in the value of the simulation being lower than the observed value. Furthermore, cloud volume will also affect simulations, potentially resulting in cold deviation [58]. Therefore, simulation parameters need to be appropriately optimized when simulating these unique and variable ground surfaces [37]. Further determination of the drivers underlying the patterns in soil temperature deviation in the Northern Hemisphere permafrost regions is needed to reduce cold deviation and improve the accuracy of simulations. In terms of trends in soil temperature, while previous studies have produced underestimates (cold deviation), this study overestimated the change trends, possibly due to the different time scales and regions of the studies.

5. Conclusions

Using 10 CMIP6 model data and observation data from 42 sites in the Northern Hemisphere, the changes in soil temperature at different depths in northern permafrost regions were simulated and analyzed, and the simulation effect of each model was evaluated. The main conclusions are summarized as follows:
(1)
The annual mean soil temperature in the permafrost region of the Northern Hemisphere gradually decreased with increasing latitude, with the lower soil temperatures near the Arctic. The spatial distribution of annual average soil temperatures did not vary much among different depths, and most of the soil temperatures in the 0–50 cm and 50–100 cm layers were concentrated around −1 °C. Those of the 100–200 cm and 200–300 cm layers were mainly concentrated around −0.5 °C.
(2)
All 10 examined CMIP6 models exhibited good soil temperature simulation abilities. Comparing the simulated and observed data, all models had correlation coefficients greater than 0.9, RMS error between 2.19 °C and 9.62 °C, standard deviation between 4.96 °C and 10.08 °C, and MAE from 1.72 °C to 8.74 °C, with a mean value of 4.9 °C. Most of the models underestimated the soil temperature. Among the 10 models, the IPLS-CM6A-LR-INCA simulated data and change trends were closest to the measured values, while the values from the KIOST-ESM model were furthest from the measured data. There were large differences in the simulations of different models. The comprehensive ranking of each model showed that the CESM2 model had the best simulation effect, CNRM-CM6-1 had the worst, and the average of the top three optimal models produces even better simulation results, indicating that multi-model averaging can effectively reduce the uncertainty of a single-model simulation in CMIP6.
(3)
The change trends in annual mean soil temperature at the 42 sites ranged between −0.17 °C/10a and 0.41 °C/10a, with most (98%) sites showing significant warming trends (p < 0.05) and greater rates of change at sites closer to the Arctic Circle. In 1900–2014 and 1980–2014, soil temperatures in the Northern Hemisphere all showed warming trends. The rate of soil temperature change in the past 35 years was about three times greater than that in the past 115 years.

Author Contributions

Conceptualization, G.H. and Y.W.; methodology, D.Z.; software, X.Z.; validation, Y.X. and L.Z.; formal analysis, Y.W.; investigation, G.H.; resources, Y.S.; data curation, R.Z.; writing—original draft preparation, Y.W.; writing—review and editing, G.H.; visualization, D.Z; supervision, T.W.; project administration, X.W.; funding acquisition, G.H. and L.Z. All authors have read and agreed to the published version of the manuscript.

Funding

This work was financially supported by the National Natural Science Foundation of China (42322608, 41931180, 42071094), and Youth Innovation Promotion Association of the Chinese Academy of Sciences (2022430).

Data Availability Statement

The 32 daily soil temperature stations in Russia, Alaska, and Svalbard were downloaded from http://gtnpdatabase.org (accessed on 9 October 2023). Soil temperature observation data from the Northeast and the Qinghai-Tibet Plateau can be downloaded from the National Tibetan Plateau/Third Pole Environment Data Center (https://data.tpdc.ac.cn/en/disallow/789e838e-16ac-4539-bb7e-906217305a1d/, accessed on 11 October 2023). The historical simulation data (1900–2014) of the CMIP6 model were downloaded from https://esgf-node.llnl.gov/search/cmip6/ (accessed on 5 May 2024).

Acknowledgments

We thank our colleagues for their insightful comments on an earlier version of this manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Subcommittee, P. Glossary of Permafrost and Related Ground-Ice Terms; Associate Committee on Geotechnical Research, National Research Council of Canada: Ottawa, ON, Canada, 1988; Volume 156, pp. 63–64. [Google Scholar]
  2. Obu, J.; Westermann, S.; Bartsch, A.; Berdnikov, N.; Christiansen, H.H.; Dashtseren, A.; Delaloye, R.; Elberling, B.; Etzelmueller, B.; Kholodov, A.; et al. Northern Hemisphere permafrost map based on TTOP modelling for 2000–2016 at 1 km2 scale. Earth-Sci. Rev. 2019, 193, 299–316. [Google Scholar] [CrossRef]
  3. Cheng, G.; Zhao, L.; Li, R.; Wu, X.; Sheng, Y.; Hu, G.; Zou, D.; Jin, H.; Li, X.; Wu, Q. Characteristic, changes and impacts of permafrost on Qinghai-Tibet Plateau. Chin. Sci. Bull. 2019, 64, 2783–2795. [Google Scholar]
  4. Schuur, E.A.G.; McGuire, A.D.; Schädel, C.; Grosse, G.; Harden, J.W.; Hayes, D.J.; Hugelius, G.; Koven, C.D.; Kuhry, P.; Lawrence, D.M.; et al. Climate change and the permafrost carbon feedback. Nature 2015, 520, 171–179. [Google Scholar] [CrossRef] [PubMed]
  5. Wrona, F.J.; Johansson, M.; Culp, J.M.; Jenkins, A.; Mård, J.; Myers-Smith, I.H.; Prowse, T.D.; Vincent, W.F.; Wookey, P.A. Transitions in Arctic ecosystems: Ecological implications of a changing hydrological regime. J. Geophys. Res.-Biogeosci. 2016, 121, 650–674. [Google Scholar] [CrossRef]
  6. Zhao, L.; Hu, G.; Zou, D.; Wu, X.; Ma, L.; Sun, Z.; Yuan, L.; Zhou, H.; Liu, S. Permafrost Changes and Its Effects on Hydrological Processes on Qinghai-Tibet Plateau. Bull. Chin. Acad. Sci. 2019, 34, 1233–1246. [Google Scholar]
  7. Ding, Y.J.; Zhang, S.Q.; Zhao, L.; Li, Z.Q.; Kang, S.C. Global warming weakening the inherent stability of glaciers and permafrost. Sci. Bull. 2019, 64, 245–253. [Google Scholar] [CrossRef] [PubMed]
  8. Gruber, S. Ground subsidence and heave over permafrost: Hourly time series reveal interannual, seasonal and shorter-term movement caused by freezing, thawing and water movement. Cryosphere 2020, 14, 1437–1447. [Google Scholar] [CrossRef]
  9. Romanovsky, V.E.; Drozdov, D.S.; Oberman, N.G.; Malkova, G.V.; Kholodov, A.L.; Marchenko, S.S.; Moskalenko, N.G.; Sergeev, D.O.; Ukraintseva, N.G.; Abramov, A.A.; et al. Thermal State of Permafrost in Russia. Permafr. Periglac. Process. 2010, 21, 136–155. [Google Scholar] [CrossRef]
  10. Qin, D. Climate change science and sustainable development. Prog. Geogr. 2014, 33, 874–883. [Google Scholar]
  11. Smith, S.L.; Romanovsky, V.E.; Lewkowicz, A.G.; Burn, C.R.; Allard, M.; Clow, G.D.; Yoshikawa, K.; Throop, J. Thermal State of Permafrost in North America: A Contribution to the International Polar Year. Permafr. Periglac. Process. 2010, 21, 117–135. [Google Scholar] [CrossRef]
  12. Romanovsky, V.E.; Smith, S.L.; Christiansen, H.H. Permafrost Thermal State in the Polar Northern Hemisphere during the International Polar Year 2007-2009: A Synthesis. Permafr. Periglac. Process. 2010, 21, 106–116. [Google Scholar] [CrossRef]
  13. Osterkamp, T.E. Establishing long-term permafrost observatories for active-layer and permafrost investigations in Alaska: 1977–2002. Permafr. Periglac. Process. 2003, 14, 331–342. [Google Scholar] [CrossRef]
  14. Fang, X.W.; Luo, S.Q.; Lyu, S.H. Observed soil temperature trends associated with climate change in the Tibetan Plateau, 1960–2014. Theor. Appl. Climatol. 2019, 135, 169–181. [Google Scholar] [CrossRef]
  15. Wu, Q.; Zhang, T.; Liu, Y. Thermal state of the active layer and permafrost along the Qinghai-Xizang (Tibet) Railway from 2006 to 2010. Cryosphere 2012, 6, 607–612. [Google Scholar] [CrossRef]
  16. Aalto, J.; Karjalainen, O.; Hjort, J.; Luoto, M. Statistical Forecasting of Current and Future Circum-Arctic Ground Temperatures and Active Layer Thickness. Geophys. Res. Lett. 2018, 45, 4889–4898. [Google Scholar] [CrossRef]
  17. Ni, J.; Wu, T.H.; Zhu, X.F.; Hu, G.J.; Zou, D.F.; Wu, X.D.; Li, R.; Xie, C.W.; Qiao, Y.P.; Pang, Q.Q.; et al. Simulation of the Present and Future Projection of Permafrost on the Qinghai-Tibet Plateau with Statistical and Machine Learning Models. J. Geophys. Res.-Atmos. 2021, 126, 20. [Google Scholar] [CrossRef]
  18. Kim, J.-E.; Alexander, M.J. Tropical precipitation variability and convectively coupled equatorial waves on submonthly time scales in reanalyses and TRMM. J. Clim. 2013, 26, 3013–3030. [Google Scholar] [CrossRef]
  19. Zhang, J.; Wang, W.C.; Wei, J. Assessing land-atmosphere coupling using soil moisture from the Global Land Data Assimilation System and observational precipitation. J. Geophys. Res. Atmos. 2008, 113, D17. [Google Scholar] [CrossRef]
  20. Chang, Y.; Lyu, S.; Luo, S.; Wu, J.; Li, R.; Li, S. Evaluation and Projections of Permafrost on the Qinghai-Xizang Plateau by CMIP5 Coupled Climate Models. Plateau Meteorol. 2016, 35, 1157–1168. [Google Scholar]
  21. Qin, Y.; Wu, T.; Li, R.; Xie, C.; Qiao, Y.; Chen, H.; Zou, D.; Zhang, L. Application of ERA Product of Land Surface Temperature in Permafrost Regions of Qinghai-Xizang Plateau. Plateau Meteorol. 2015, 34, 666–675. [Google Scholar]
  22. Albergel, C.; Dutra, E.; Muñoz-Sabater, J.; Haiden, T.; Balsamo, G.; Beljaars, A.; Isaksen, L.; de Rosnay, P.; Sandu, I.; Wedi, N. Soil temperature at ECMWF: An assessment using ground-based observations. J. Geophys. Res. Atmos. 2015, 120, 1361–1373. [Google Scholar] [CrossRef]
  23. Ding, X.; Lai, X.; Fan, G.; Wen, J.; Yuan, Y.; Wang, X.; Wang, Z.; Zhu, L.; Zhang, Y.; Wang, B. Analysis on the Applicability of Reanalysis Soil Temperature and Moisture Datasets over Qinghai-Tibetan Plateau. Plateau Meteorol. 2018, 37, 626–641. [Google Scholar]
  24. Li, R.; Shihua, L.U.; Han, b.; Gao, Y. Preliminary Comparison and Analyses of Air Temperature at 2 m Height between Three Reanalysis Data-Sets and Observation in the East of Qinghai-Xiang Plateau. Plateau Meteorol. 2012, 31, 1488–1502. [Google Scholar]
  25. Liu, C.; Yu, Y.; Xie, J.; Zhou, X.; Li, J.; Ge, J. Applicability of Soil Temperature and Moisture in Several Datasets over Qinghai-Xizang Plateau. Plateau Meteorol. 2015, 34, 653–665. [Google Scholar]
  26. Zhu, Z.; Shi, C.; Zhang, T.; Wang, J. Applicability Analysis of Four Soil Moisture Reanalysis Datasets in China. Plateau Meteorol. 2018, 37, 240–252. [Google Scholar]
  27. Chenghai, W.; Yongpingg, W.U.; Yang, C.U.I. Evaluating the Progress of the CMIP and Its Application Prospect in China. Adv. Earth Sci. 2009, 24, 461–468. [Google Scholar]
  28. Covey, C.; AchutaRao, K.M.; Cubasch, U.; Jones, P.; Lambert, S.J.; Mann, M.E.; Phillips, T.J.; Taylor, K.E. An overview of results from the Coupled Model Intercomparison Project. Glob. Planet. Change 2003, 37, 103–133. [Google Scholar] [CrossRef]
  29. Fan, X.; Miao, C.; Gou, J.; Wu, Y.; Zhang, Q. Coupled Model Intercomparison Project and its research progress in simulation performance. Prog. Geogr. 2023, 42, 1204–1215. [Google Scholar] [CrossRef]
  30. Xing, Z.P.; Zhao, L.; Fan, L.; Hu, G.-J.; Zou, D.-F.; Wang, C.; Liu, S.-C.; Du, E.-J.; Xiao, Y.; Li, R.; et al. Changes in the ground surface temperature in permafrost regions along the Qinghai-Tibet engineering corridor from 1900 to 2014: A modified assessment of CMIP6. Adv. Clim. Chang. Res. 2023, 14, 85–96. [Google Scholar] [CrossRef]
  31. Zhou, T.; Zou, L.; Chen, X. Commentary on the Coupled Model Intercomparison Project Phase 6 (CMIP6). Progress. Inquisitiones Mutat. Clim. 2019, 15, 445–456. [Google Scholar]
  32. Hu, Q.; Jiang, D.; Fan, G. Evaluation of CMIP5 Models over the Qinghai-Tibetan Plateau. Chin. J. Atmos. Sci. 2014, 38, 924–938. [Google Scholar]
  33. Yang, M.; Xiao, T.; Li, Y.; Hu, T. Evaluation and Projection of Climate Change in Southwest China Using CMIP6 Models. Plateau Meteorol. 2022, 41, 1557–1571. [Google Scholar]
  34. Jin, H.; Xiaoqing, P.; Cong, C.; Yaohua, Z.; Ran, D.; Sijia, P. Comparative study on precipitation changes in the Northern Hemisphere and permafrost regions based on CMIP6 models. J. Glaciol. Geocryol. 2023, 45, 641–654. [Google Scholar]
  35. Yang, X.L.; Zhou, B.T.; Xu, Y.; Han, Z.Y. CMIP6 Evaluation and Projection of Temperature and Precipitation over China. Adv. Atmos. Sci. 2021, 38, 817–830. [Google Scholar] [CrossRef]
  36. Hu, G.J.; Zhao, L.; Wu, T.H.; Wu, X.D.; Park, H.; Li, R.; Zhu, X.F.; Ni, J.; Zou, D.F.; Hao, J.M.; et al. Continued Warming of the Permafrost Regions Over the Northern Hemisphere Under Future Climate Change. Earth Future 2022, 10, 21. [Google Scholar] [CrossRef]
  37. Cai, Z.Y.; You, Q.L.; Wu, F.Y.; Chen, H.W.; Chen, D.L.; Cohen, J.D. Arctic Warming Revealed by Multiple CMIP6 Models: Evaluation of Historical Simulations and Quantification of Future Projection Uncertainties. J. Clim. 2021, 34, 4871–4892. [Google Scholar] [CrossRef]
  38. Chang, X. Geotemperature observation data set of Genhe River (2012–2019). 2021. [Google Scholar] [CrossRef]
  39. Chang, X.; Jin, H.; He, R.; Zhang, Y.; Li, X.; Jin, X.; Li, G. Permafrost changes in the northwestern Da Xing’anling Mountains, Northeast China, in the past decade. Earth Syst. Sci. Data 2022, 14, 3947–3959. [Google Scholar] [CrossRef]
  40. Zhao, L.; Zou, D.F.; Hu, G.J.; Wu, T.H.; Du, E.J.; Liu, G.Y.; Xiao, Y.; Li, R.; Pang, Q.Q.; Qiao, Y.P.; et al. A synthesis dataset of permafrost thermal state for the Qinghai-Tibet (Xizang) Plateau, China. Earth Syst. Sci. Data 2021, 13, 4207–4218. [Google Scholar] [CrossRef]
  41. Qiao, L. Changes of Soil Temperature in High Latitudes of Eurasia and Its Response to Climate. Master’s Thesis, Lanzhou University, Lanzhou, China, 2022. [Google Scholar]
  42. Guojie, H.; Lin, Z.; Zhe, S.; Defu, Z.; Yao, X.; Guangyue, L.; Erji, D.; Chong, W.; Yuanwei, W.; Xiaodong, W. Spatiotemporal characteristics and variability in the thermal state of permafrost on the Qinghai–Tibet Plateau. Permafr. Periglac. Process. 2024, 35, 143–156. [Google Scholar] [CrossRef]
  43. Yang, N.; Fan, G. Analysis of Soil Temperature Variation Over the Qinghai-Xizang Plateau from 2000 to 2016. J. Southwest Univ. Nat. Sci. Ed. 2019, 41, 40–51. [Google Scholar]
  44. Jing, L.I.; Yu, S.; Ji, C.; Bo, Z.; Jichun, W.U.; Xiumin, Z. Characteristics of Ground Temperatures and Influencing Factors of Permafrost Development and Distribution in the Source Region of Datong River. Prog. Geogr. 2011, 30, 827–836. [Google Scholar]
  45. Wang, X.; Chen, R.; Han, C.; Yang, Y.; Liu, J.; Liu, Z.; Guo, S.; Song, Y. Soil temperature change and its regional differences under different vegetation regions across China. Int. J. Climatol. 2021, 41, E2310–E2320. [Google Scholar] [CrossRef]
  46. Sherstiukov, A.B.; Anisimov, O.A. Assessment of the Snow Cover Effect on Soil Surface Temperature from Observational Data. Russ. Meteorol. Hydrol. 2018, 43, 72–78. [Google Scholar] [CrossRef]
  47. Zhang, T.J. Influence of the seasonal snow cover on the ground thermal regime: An overview. Rev. Geophys. 2005, 43, 4. [Google Scholar] [CrossRef]
  48. Hu, G.; Zhao, L.; Li, R.; Park, H.; Wu, X.; Su, Y.; Guggenberger, G.; Wu, T.; Zou, D.; Zhu, X. Water and heat coupling processes and its simulation in frozen soils: Current status and future research directions. CATENA 2023, 222, 106844. [Google Scholar] [CrossRef]
  49. Hu, G.; Zhao, L.; Zhu, X.; Wu, X.; Wu, T.; Li, R.; Xie, C.; Hao, J. Review of algorithms and parameterizations to determine unfrozen water content in frozen soil. Geoderma 2020, 368, 114277. [Google Scholar] [CrossRef]
  50. Tett, S.F.B.; Gregory, J.M.; Freychet, N.; Cartis, C.; Mineter, M.J.; Roberts, L. Does Model Calibration Reduce Uncertainty in Climate Projections? J. Clim. 2022, 35, 2585–2602. [Google Scholar] [CrossRef]
  51. Hu, G.; Zhao, L.; Wu, X.; Li, R.; Wu, T.; Xie, C.; Qiao, Y.; Shi, J.; Li, W.; Cheng, G. New Fourier-series-based analytical solution to the conduction–convection equation to calculate soil temperature, determine soil thermal properties, or estimate water flux. Int. J. Heat Mass Transf. 2016, 95, 815–823. [Google Scholar] [CrossRef]
  52. He, J.; Zhan, F.; Qi, L.; Wang, D. Evolution of Surface Temperature during Global Warming Hiatus Based on Observations and CMIP5 Simulations. Chin. J. Atmos. Sci. 2016, 40, 33–45. [Google Scholar]
  53. Jiang, J.; Jiang, D.; Lin, Y. Projection of Monsoon Area and Precipitation in China under the RCP4.5 Scenario. Chin. J. Atmos. Sci. 2015, 39, 901–910. [Google Scholar]
  54. Knutti, R. Should we believe model predictions of future climate change? Philos. Trans. R. Soc. A Math. Phys. Eng. Sci. 2008, 366, 4647–4664. [Google Scholar] [CrossRef] [PubMed]
  55. Tao, C.W.; Jiang, C.; Sun, J.X. Projection of future changes in climate in Northeast China using a CMIP5 multi-model ensemble. Chin. J. Geophys. Chin. Ed. 2016, 59, 3580–3591. [Google Scholar] [CrossRef]
  56. Zhang, W.; Zhang, J.; Fan, G. Evaluation and Projection of Dry- and Wet-Season Precipitation in Southwestern China Using CMIP5 Models. Chin. J. Atmos. Sci. 2015, 39, 559–570. [Google Scholar]
  57. Danabasoglu, G.; Lamarque, J.F.; Bacmeister, J.; Bailey, D.; DuVivier, A.; Edwards, J.; Emmons, L.; Fasullo, J.; Garcia, R.; Gettelman, A. The community earth system model version 2 (CESM2). J. Adv. Model. Earth Syst. 2020, 12, e2019MS001916. [Google Scholar] [CrossRef]
  58. Zhu, Y.Y.; Yang, S.N. Evaluation of CMIP6 for historical temperature and precipitation over the Tibetan Plateau and its comparison with CMIP5. Adv. Clim. Chang. Res. 2020, 11, 239–251. [Google Scholar] [CrossRef]
  59. Chapman, W.L.; Walsh, J.E. Simulations of Arctic temperature and pressure by global coupled models. J. Clim. 2007, 20, 609–632. [Google Scholar] [CrossRef]
  60. Davy, R.; Outten, S. The Arctic Surface Climate in CMIP6: Status and Developments since CMIP5. J. Clim. 2020, 33, 8047–8068. [Google Scholar] [CrossRef]
  61. Pei, Y.; Song, M.; Ma, X.; Wu, T.; Zhang, S. Simulation assessment and prediction of future temperatures in Northwest China from BCC-CSM Model. Sci. Cold Arid Reg. 2022, 14, 138–150. [Google Scholar]
Land 13 01029 g001
Figure 1. Spatial distribution diagram of temperature observation stations in permafrost areas in the Northern Hemisphere.
Figure 1. Spatial distribution diagram of temperature observation stations in permafrost areas in the Northern Hemisphere.
Land 13 01029 g001
Land 13 01029 g002
Figure 2. Spatial distribution of annual average soil temperature among observation stations in permafrost areas of the Northern Hemisphere. (a), (b), (c) and (d) represent 0–50 cm, 50–100 cm, 100–200 cm, and 200–300 cm, respectively.
Figure 2. Spatial distribution of annual average soil temperature among observation stations in permafrost areas of the Northern Hemisphere. (a), (b), (c) and (d) represent 0–50 cm, 50–100 cm, 100–200 cm, and 200–300 cm, respectively.
Land 13 01029 g002
Land 13 01029 g003
Figure 3. Taylor plot of soil temperature simulated and observed for different CMIP 6 modes from 2003 to 2012. (a), (b), (c) and (d) represent 0–50 cm, 50–100 cm, 100–200 cm, and 200–300 cm, respectively.
Figure 3. Taylor plot of soil temperature simulated and observed for different CMIP 6 modes from 2003 to 2012. (a), (b), (c) and (d) represent 0–50 cm, 50–100 cm, 100–200 cm, and 200–300 cm, respectively.
Land 13 01029 g003
Land 13 01029 g004
Figure 4. Observed and simulated soil temperature trends in the Northern Hemisphere from 2003 to 2012. (a), (b), (c) and (d) represent 0–50 cm, 50–100 cm, 100–200 cm, 200–300 cm, respectively.
Figure 4. Observed and simulated soil temperature trends in the Northern Hemisphere from 2003 to 2012. (a), (b), (c) and (d) represent 0–50 cm, 50–100 cm, 100–200 cm, 200–300 cm, respectively.
Land 13 01029 g004
Land 13 01029 g005aLand 13 01029 g005b
Figure 5. Rate of soil temperature change in the permafrost region of the Northern Hemisphere from 1900 to 2014. (a), (b), (c) and (d) represent 0–50 cm, 50–100 cm, 100–200 cm, 200–300 cm, respectively.
Figure 5. Rate of soil temperature change in the permafrost region of the Northern Hemisphere from 1900 to 2014. (a), (b), (c) and (d) represent 0–50 cm, 50–100 cm, 100–200 cm, 200–300 cm, respectively.
Land 13 01029 g005aLand 13 01029 g005b
Land 13 01029 g006
Figure 6. Trend of soil temperature in Northern Hemisphere permafrost from 1900 to 2014 and 1950 to 2014. (a), (b), (c) and (d) represent 0–50 cm, 50–100 cm, 100–200 cm, 200–300 cm, respectively.
Figure 6. Trend of soil temperature in Northern Hemisphere permafrost from 1900 to 2014 and 1950 to 2014. (a), (b), (c) and (d) represent 0–50 cm, 50–100 cm, 100–200 cm, 200–300 cm, respectively.
Land 13 01029 g006
Table 1. 10 patterns in CMIP6Model.
Table 1. 10 patterns in CMIP6Model.
ModelInstituteLayersDepth (m)Stratification
CESM2NCAR2548.560.025, 0.065, 0.125, 0.21, 0.33, 0.49, 0.69, 0.93, 1.21, 1.53, 1.89, 2.29, 2.745, 3.285, 3.925, 4.665, 5.505, 6.445, 7.485, 8.9125, 11.5614, 16.4054
CNRM-CM6-1CNRM-CERFACS14120.01, 0.04, 0.1, 0.2, 0.4, 0.6, 0.8, 1, 1.5, 2, 3, 5, 8, 12
CNRM-ESM2-1CNRM-CERFACS14120.01, 0.04, 0.1, 0.2, 0.4, 0.6, 0.8, 1, 1.5, 2, 3, 5, 8, 12
NorESM2-LMNCC25490.025, 0.065, 0.125, 0.21, 0.33, 0.49, 0.69, 0.93, 1.21, 1.53, 1.89, 2.29, 2.745, 3.285, 3.925, 4.665, 5.505, 6.445, 7.485, 8.9125, 11.5614, 16.4054
NorESM2-MMNCC2548.560.025, 0.065, 0.125, 0.21, 0.33, 0.49, 0.69, 0.93, 1.21, 1.53, 1.89, 2.29, 2.745, 3.285, 3.925, 4.665, 5.505, 6.445, 7.485, 8.9125, 11.5614, 16.4054
IPSL-CM5A2-INCAIPSL73.070.014, 0.063, 0.159, 0.353, 0.74, 1.52, 3.07
IPLS-CM6A-LR-INCAIPSL1865.560.0004, 0.0019, 0.0058, 0.0137, 0.029, 0.06, 0.123, 0.24, 0.498, 0.99, 1.75, 2.5, 3.5, 5.5, 9.5, 17.5, 33.5, 65.5
CESM2-WACCMCNAR2548.560.025, 0.065, 0.125, 0.21, 0.33, 0.49, 0.69, 0.93, 1.21, 1.53, 1.89, 2.29, 2.745, 3.285, 3.925, 4.665, 5.505, 6.445, 7.485, 8.9125, 11.5614, 16.4054
MPI-ESM-1-2-HAMHAMMOZ-Consortium59.830.065, 0.319, 1.232, 4.134, 9.834
KIOST-ESMKIOST208.750.01, 0.04, 0.08, 0.125, 0.175, 0.25, 0.35, 0.5, 0.7, 0.9, 1.2, 1.6, 2, 2.4, 2.8, 3.5, 4.5, 5.5, 6.75, 8.75
Table 2. Ranking of different depth simulation results by the CMIP 6 mode in the 2003–2012.
Table 2. Ranking of different depth simulation results by the CMIP 6 mode in the 2003–2012.
Model0–50 cm50–100 cm100–200 cm200–300 cmOverall Ranking
IPSL-CM5A2-INCA10101029
IPSL-CM6A-LR-INCA35253
Noresm2-MM44575
KIOST-ESM57797
CNRM-ESM2-178888
NorESM2-LM53343
CESM211111
CNRM-CM6-18991010
CESM2-WACCM22432
MPI-ESM-1-2-HAM95666
Table 3. Comparison of evaluation indexes for different depth optimal modes.
Table 3. Comparison of evaluation indexes for different depth optimal modes.
LayerOptimal ModeRRMSEMAEBias
0–50 cmAME10.974.281.500.30
AME30.991.040.790.09
50–100 cmAME10.981.501.130.31
AME30.991.661.250.39
100–200 cmAME10.973.002.671.56
AME30.902.52.171.28
200–300 cmAME10.933.833.172.57
AME30.944.644.123.34
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Wu, Y.; Hu, G.; Zhao, L.; Zou, D.; Zhu, X.; Xiao, Y.; Wu, T.; Wu, X.; Su, Y.; Zhang, R. Assessment of Soil Temperature and Its Change Trends in the Permafrost Regions of the Northern Hemisphere. Land 2024, 13, 1029. https://doi.org/10.3390/land13071029

AMA Style

Wu Y, Hu G, Zhao L, Zou D, Zhu X, Xiao Y, Wu T, Wu X, Su Y, Zhang R. Assessment of Soil Temperature and Its Change Trends in the Permafrost Regions of the Northern Hemisphere. Land. 2024; 13(7):1029. https://doi.org/10.3390/land13071029

Chicago/Turabian Style

Wu, Yifan, Guojie Hu, Lin Zhao, Defu Zou, Xiaofan Zhu, Yao Xiao, Tonghua Wu, Xiaodong Wu, Youqi Su, and Rui Zhang. 2024. "Assessment of Soil Temperature and Its Change Trends in the Permafrost Regions of the Northern Hemisphere" Land 13, no. 7: 1029. https://doi.org/10.3390/land13071029

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop