1. Introduction
Permafrost, defined as ground materials at or below 0°C for two or more consecutive years, underlies about 17% of the global exposed land surface (Gruber 2012). Long-term observational records indicate that mean permafrost temperature increased by 0.29 ± 0.12°C (mean ± standard deviation) during the period 2007–16 (Biskaborn et al. 2019). The resulting permafrost degradation has significantly affected landscapes (Jorgenson and Grosse 2016; Holloway et al. 2020), hydrological processes (Streletskiy et al. 2015; Jin et al. 2022), biogeochemical cycles (Jorgenson et al. 2001; Miner et al. 2021), and infrastructure (Melvin et al. 2017; Hjort et al. 2018). The thaw of ice-rich permafrost is causing widespread subsidence due to the melting of excess ice (O’Neill et al. 2023). Climate change has put a large amount of permafrost carbon at high risk. The release of this buried carbon, which is about twice the amount currently in Earth’s atmosphere, may drive a strong positive feedback loop with climate warming (Schuur et al. 2015; Miner et al. 2022). The thermal state of permafrost is crucial for understanding the decomposition of permafrost carbon and its contribution to future climate change.
Permafrost is a subsurface phenomenon that is difficult to investigate directly through surface features (Cao et al. 2017, 2019b). Therefore, permafrost studies commonly rely on in situ observations (e.g., temperature sensors in boreholes) and numerical simulations (Biskaborn et al. 2015; Burke et al. 2020; Sun et al. 2023). Despite their importance, observations in permafrost regions are sparse due to the difficulty and cost of monitoring in the remote and harsh environments where permafrost is found. This includes both in situ measurements of permafrost and the meteorological data needed as boundary conditions for simulations. Climate reanalyses, with long temporal coverage, are one solution to this problem of data scarcity and are valuable resources for permafrost research. Reanalysis datasets are created by assimilating a broad range of observations and satellite data into coupled process-based transient models. However, most studies have used only the atmospheric variables (e.g., air temperature, solar radiation) of reanalysis as model forcing for stand-alone permafrost simulations (e.g., Fiddes et al. 2015; Guo et al. 2018; Tao et al. 2019; Cao et al. 2019a). The use of reanalysis-derived soil temperature data for permafrost studies is limited by the significant bias and coarse spatial resolution (50–150 km) (Hu et al. 2019; Cao et al. 2020).
The Japanese Reanalysis for three quarters of a century (JRA-3Q) is the latest-generation reanalysis produced by the Japan Meteorological Agency (JMA). Early results show that JRA-3Q represents atmospheric variables significantly better than its predecessor, JRA-55 (Harada et al. 2021). These improvements have been attributed to better and more frequent observations used for assimilation, as well as more realistic physical processes in the model, including revised processes for radiation budgets, evaporation, excessive sensible/latent heat fluxes, and the implementation of a new multilayer snow and soil model (Kosaka et al. 2024). These improvements should be expected to make JRA-3Q more suitable for use in permafrost regions, but no evaluations have yet been conducted.
Here, we evaluate the performance of JRA-3Q in permafrost regions. We compare the representation of key variables, active layer thickness (ALT) and permafrost area, using referenced estimates from different methodologies. Additionally, we conduct a detailed simulation example to demonstrate the drawbacks of the land surface model employed in JRA-3Q. More specifically, the main objectives of this study are to
-
evaluate the suitability of JRA-3Q for permafrost research in comparison to its predecessor (JRA-55) and other reanalyses;
-
assess the capability of JRA-3Q to replicate soil freezing phenomena and highlight avenues for further development of the land surface models employed in reanalysis in light of the revealed bias and its potential causes.
2. JRA-3Q reanalysis
a. JRA-55 and its land surface scheme
To explicitly highlight improvements to JRA-3Q that are relevant for permafrost applications, key permafrost-related processes in the land surface scheme of its predecessor, JRA-55, are summarized here. JRA-55 employed the second-generation version of the Simple Biosphere Model, version 2 (SiB2), with revised land surface parameterization (Sellers et al. 1996a,b; Randall et al. 1996). Unlike many other state-of-the-art reanalyses, snow depth in JRA-55 was derived separately rather than being calculated as part of the land surface model. This was done using an optimized univariate two-dimensional interpolation based on snow depth observations; snow water equivalent (SWE) dynamics were provided separately by the SiB2 model (Kobayashi et al. 2015). The snow processes were heavily simplified: density was parameterized as a linear function of SWE following Verseghy (1991), snow albedo was treated as a constant of 0.8—except during snowmelt periods when it was reduced by 40% (Sellers et al. 1986)—and the snow temperature was fixed as the lesser value of the skin temperature and 0°C.
The soil scheme in JRA-55 used three layers to treat the hydrological processes and vegetation roots, but soil temperature was treated as a single layer across the entire soil column (Sellers et al. 1996b). The soil temperature was determined empirically, and the influence of snow scaled linearly with the snow-covered area (Deardorff 1978). Not surprisingly, the simplified snow scheme and soil temperature dynamics of JRA-55 limited the suitability of the reanalysis for applications relating to soil temperature (Li et al. 2020; Herrington et al. 2024). The root-mean-square error (RMSE) of JRA-55 soil temperature in summer and winter was found to be greater than 8° and 5°C, respectively (Herrington et al. 2024). The soil temperature bias of JRA-55 was even more pronounced in permafrost regions, with biases (and RMSE values) in near-surface (0–0.3 m) soil temperature of up to about −9.6 (10)°C in summer and about 1.5 (5.5)°C in winter. In deeper soil (0.3–3.0 m), the corresponding values were about −4.2 (5.6)°C in summer and about −1.5 (4.1)°C in winter.
b. JRA-3Q
JRA-3Q represents a significant upgrade of JRA-55 and is produced with an improved four-dimensional variational data assimilation (4D-Var). JRA-3Q (specifications of JRA-55 are shown in brackets for comparison) has a higher spatial resolution of 0.375° (0.5625°), a longer coverage period from 1947 (1958) onward, and a finer vertical resolution with 45 (37) pressure levels between 1000 (1000) and 0.01 (1.0) hPa (Naoe et al. 2021; Harada et al. 2021). The temporal resolution of JRA-3Q is 6 hourly for surface as well as pressure level variables and hourly (3 hourly) for surface forecast analysis. Note that, at the time of writing only JRA-3Q data between 1991 and 2012 had been released to the public, and so this evaluation is conducted using data between this period.
c. The improved snow and soil scheme of JRA-3Q
An improved SiB2 model is constructed in JRA-3Q. A more realistic snow scheme is employed in the land surface model of JRA-3Q which allows for up to four layers (Harada et al. 2021). The thicknesses of snow layers are dynamically discretized (cf. Oleson et al. 2010), and snow density takes into account the effects of compaction, metamorphism (crystal breakdown due to wind or thermodynamic stress), and melt (Anderson 1976; Oleson et al. 2010). The thermal conductivity of snow is determined by snow density following Jordan (1991). Snow albedo in JRA-3Q is no longer a constant but is determined by the snow age.
3. Reference data for evaluation
a. In situ observations
Observations of near-surface air temperature Ta, soil temperature Ts, and snow depth (HS) from 1173 sites in permafrost regions were used here to evaluate JRA-3Q in permafrost regions (Table 1). The data sources are generally the same as those in Cao et al. (2020) with the additional inclusion of the Copernicus Climate Change Service Global Land and Marine Observations Database (fourth version) from the European Centre for Medium-Range Weather Forecasts (ECMWF), the Yukon Permafrost Database from the Yukon Geological Survey (Yukon), and the International Soil Moisture Network (ISMN) from the International Centre for Water Resources and Global Change. The same evaluations were conducted at 30 213 sites in nonpermafrost regions as reference. The presence or absence of permafrost for each station is derived from the global permafrost zonation index map (hereafter referred to as the PZI map) (Gruber 2012). Measured ALT from Peng et al. (2018) is used to evaluate ALT derived from JRA-3Q. Note that meteorological variables from some sites are included in the JRA-3Q data assimilation. However, the exact details of in situ observations assimilated by JRA-3Q are not available here (i.e., the evaluation may not be completely stand-alone). Observations were screened by removing obvious outliers through visual inspection. The observations were aggregated if the completeness of the record was greater than 95%. Observed soil temperatures were grouped according to depth based on the JRA-3Q soil layers. The evaluation of snow depth was conducted for the consistent snow-covered period (October–March for the Northern Hemisphere, and April–September for the Southern Hemisphere) by following Slater et al. (2017) and Cao et al. (2020).
Summary of near-surface air temperature Ta (°C), soil temperature Ts (°C), and HS (cm) observations in permafrost regions, including the total number of stations (N), the temporal coverage (coverage) and measurement range (given in brackets), the corresponding JRA-3Q soil layers and depth range in meters (SL), and a reference for each dataset when available. CMA = China Meteorological Administration; WDCs = World Data Centers in Russia and Ukraine; ECMWF = European Centre for Medium-Range Weather Forecasts; USGS = U.S. Geological Survey; NPS = National Park Service in Alaska; GI-UAF = Geophysical Institute, University of Alaska Fairbanks; Tibet-OBS = regional-scale soil moisture and soil temperature measurements at multiple depths on the Tibetan Plateau; CTP-SMTMN = multiscale Soil Moisture and Temperature Monitoring Network in the Central Tibetan Plateau; GTN-P = Global Terrestrial Network for Permafrost; Yukon = Yukon Permafrost Database; ISMN = International Soil Moisture Network. Note that references of the Nordicana D dataset are given for each site in the online supplemental material.
b. Permafrost maps
We use four permafrost maps as reference datasets to evaluate permafrost area derived from JRA-3Q. These are 1) the International Permafrost Association map (IPA map) derived mainly based on observations and mean annual air temperature (MAAT) (Brown et al. 1997); 2) the PZI map compiled based on the heuristic–empirical relationship between MAAT and permafrost zonation (Gruber 2012); 3) the Northern Hemisphere permafrost map derived via the semiphysical “temperature at the top of permafrost” (TTOP) model (TTOP map) (Obu et al. 2019); and 4) the circumpolar permafrost map (CP map) derived from a statistical model (Karjalainen et al. 2019). The permafrost extent from these maps is estimated for different time periods using different modeling paradigms and provides near-global coverage (Cao et al. 2020). While these maps are not perfect or a “source of truth,” we treat them as an ensemble and use their range and mean as the “best available” reference for the evaluation of the permafrost area against the JRA-3Q output.
4. Methods
a. Evaluation
b. Comparison with permafrost-specific simulation
To address the potential influences of the DECP algorithm used in the land surface scheme of JRA-3Q, we performed a stand-alone simulation using GEOtop 2.0 (Endrizzi et al. 2014), a model with a more sophisticated treatment of soil freezing behavior which parameterizes an apparent heat capacity, taking into consideration instantaneous latent heat effects (Dall’Amico et al. 2011). GEOtop is a process-based numerical model that has been widely used for permafrost simulation (Gubler et al. 2013; Fiddes et al. 2015; Cao et al. 2019b). We simulated the soil temperature for two sites: 1) the Ch04 site (latitude: 31.82°N, longitude: 91.74°E, elevation: 4808 m) over the Tibetan Plateau and 2) the Lac de Gras (LdG) site (latitude: 64.7°N, longitude: 110.6°W, elevation: 478 m) in the Northwest Territories, Canada (Fig. 2a). They represent the surface condition of alpine swamp meadow and a sedge fen. Ground temperature records and detailed site descriptions can be found from Zhao et al. (2021) for Ch04 and Cao et al. (2019a) for LdG, respectively. To conduct a like-to-like simulation, the model was driven by meteorological variables from JRA-3Q, and the required soil texture information used the same datasets as JRA-3Q to reduce confounding factors in comparison. The simulated soil column was defined to be 3.5 m, and the same zero heat flux bottom boundary as JRA-3Q was used.
5. Results and discussion
a. Near-surface air temperature and HS
In permafrost regions, the MAAT of JRA-3Q has an overall bias of −0.55°C, and about 83.4% of evaluated grid cells have a wRMSE of less than 2°C (Table 2; Fig. 1a). This outperforms JRA-55 on most permafrost-distributed mainland (e.g., Zhang et al. 2021). Especially in Siberia, an overall warm-biased MAAT of +2.39°C was significantly reduced to −0.20°C. While the Ta of JRA-3Q is slightly cold biased, the bias is generally close for all seasons (Fig. 3a and Fig. A1 in the appendix), indicating JRA-3Q has similar performance throughout the year. The overall MAAT in permafrost regions was improved by 0.8°C compared to the ERA5-Land (Cao et al. 2020), especially over the Tibetan Plateau where the remarkably cold bias was significantly reduced by 4.7°C.
Comparison of JRA-3Q with observations and published data products for MAAT (°C), daily HS (cm), MAGT (°C) of different soil layers (center depth; m), and mean annual SO (°C). N is the total number of observation sites. The number of unique grid cells is given in brackets. SL1 through SL7 correspond to each JRA-3Q soil layer, while “Overall” represents the mean over the entire soil column.
Comparison of JRA-3Q (a) MAAT (°C) and (b) daily HS (cm) with observations. Large circle markers represent sites in permafrost regions, and small square markers represent sites in nonpermafrost regions. Ice sheets and glaciers were obtained from global annual land cover maps produced by the European Space Agency Climate Change Initiative (ESA-CCI) and distributed by ECMWF.
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0267.1
The wBIAS of snow depth is about −5.5 cm based on the 1.2 × 106 daily observations from 636 grid cells in permafrost regions (Table 2; Fig. 1b). A strong bias exists in Alaska where the wRMSE is as high as 22.5 cm (or 73.5% of the observed mean snow depth). However, both thin and thick biases were found in Alaska, leading to a near-zero wBIAS of −0.4 cm. Such bias is also reported for other reanalyses (Xiao et al. 2020) and is attributed to an incorrect partitioning of precipitation between rainfall and snowfall (Broxton et al. 2016). The snow depth of JRA-3Q over the Tibetan Plateau has an overall wBIAS of +3.1 cm, which is improved by 5.9 cm or 65.6% compared to ERA5-Land (wBIAS = 9.0 cm). This, in part, explains the improved air temperature results over the Tibetan Plateau in winter. Regardless of the uncertainties in Alaska, the overall wRMSE of JRA-3Q snow depth was reduced by about 10.7 and 4.7 cm compared to the ERA5-Land with a single-layer snow scheme and optimized multilayer snow scheme (Cao et al. 2020, 2022). In nonpermafrost regions, the MAAT and snow depth are found to have better performance, with the RMSE of 1.20°C and 11.4 cm, respectively (Table 2).
b. Soil temperature and ALT
Comparison of JRA-3Q (a) MAGT (°C) and (b) SO (°C) with observations. The wBIAS for MAGT is calculated using all available MAGTs from the seven soil layers. The triangles in (a) mark the location of the Ch04 and LdG boreholes where we conduct detailed permafrost simulation using a stand-alone model.
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0267.1
Seasonal deviations of JRA-3Q soil temperature in permafrost regions. Seasonal soil temperature is first simulated for each depth and grid, and then the comparison is conducted for each season by averaging the RMSE of all grids. The numbers in black at the top of each cell are for all permafrost regions, while those in gray are the results for Alaska. The Ta is the near-surface air temperature, and SO is the surface offset, which is calculated by subtracting Ta from the priority of the second soil layer to the third one unless the second layer observations are missing. Note that the selected soil layer of observations should correspond to that of JRA-3Q.
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0267.1
Estimates of ALT in JRA-3Q are substantially overestimated with a wBIAS and wRMSE of 0.65 and 1.00 m, respectively (Figs. 4a and 5). This corresponds to the warm-biased soil temperature in summer (wBIAS = 2.52°C, Fig. 3a).
Permafrost areas derived from various sources: (a) IPA map of continuous and discontinuous permafrost zones (i.e., areas underlain by >50% permafrost), (b) 6-hourly soil temperature of JRA-3Q with a freezing temperature threshold of 0°C, (c) 6-hourly soil temperature of JRA-3Q with a freezing temperature threshold of 0.01°C, (d) MAGT of JRA-3Q at 3.49 m, and (e) the PZI calculated using the JRA-3Q MAAT from 1991 to 2012. In subfigure (a), the color of the filled circles represents the ALT wBIAS for the observed JRA-3Q grid, and their size indicates the number of ALT measurements used for evaluation.
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0267.1
Comparison of ALTs based on 1372 measurements from 155 stations located in 104 grids. The observed sites are mainly located in high latitudes, and the distribution is present in Fig. 4a. The black points represent the medians of each grid cell derived from different sites and/or different measurement times, and the gray lines represent the interquartile ranges (Q1, Q3) of the corresponding medians.
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0267.1
c. Permafrost distribution
The global permafrost area estimated from JRA-3Q using MAGT at 3.49 m was about 10.8 ± 0.3 × 106 km2 (Figs. 4 and 6). Our results align closely with the ensemble mean of reference (REF) estimates of 14.3 ± 1.7 × 106 km2. On the other hand, the area derived from the 6-hourly Ts using a 0°C threshold was significantly underestimated (1.7 ± 0.2 × 106 km2). However, when this method uses the 0.01°C threshold (corresponds to a fake permafrost definition of “ground that remains at or below 0.01°C for two or more years”), the area estimates are 11.4 ± 0.3 × 106 km2 and are within the range of literature estimates. Finally, the estimated permafrost area based on the PZI model driven by JRA-3Q MAAT (1991–2012) was about 11.5–15.8 × 106 km2, or about 2.6 × 106 km2 greater than the area derived from MAGT (Fig. 6). The difference between the permafrost area derived from soil temperature and the one derived from near-surface air temperature provides further evidence that the snow insulation in JRA-3Q is too strong.
Comparison of reference permafrost area and estimated permafrost area calculated using different JRA-3Q variables: 6-hourly soil temperature Ts using two different thresholds, the MAGT at 3.49 m, and the MAAT driven. The black horizontal line shows the REF permafrost area (14.3 ± 1.7 × 106 km2). The triangle represents the permafrost area based on JRA-3Q MAAT (1991–2012) by following Gruber (2012), and the gray line indicates the possible range of possible permafrost areas using different parameterization assumptions.
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0267.1
d. Issue of the DECP method
Both the observations and our stand-alone simulation results at two typical permafrost sites indicate that JRA-3Q overestimates soil temperature and that JRA-3Q soil temperature dynamics do not correctly represent zero-curtain behavior (Figs. 7a,b). The zero curtain refers to the period during which the effect of latent heat dampens temperature fluctuations and maintains a relatively stable temperature (Outcalt et al. 1990). During the soil freezing process, the phase transition is spread over a range of temperatures below the freezing point due to the freezing characteristic curve (Devoie et al. 2022). However, soil temperatures in JRA-3Q show significant oscillation near 0°C (Figs. 7c,d). This means partially frozen behavior takes place at temperatures slightly above 0°C (and hence above the freezing point). This is because, in DECP, the soil heat conduction is first carried throughout the soil column, and the heat due to phase change is then compensated by adjusting soil temperature (note that JRA-3Q is not capable of representing the unfrozen water). Previous studies have also reported that DECP can result in unrealistic representation of zero-curtain effects (Nicolsky et al. 2007; Zhang et al. 2008; Tubini et al. 2021). In contrast, our stand-alone simulation using GEOtop uses a parameterization of apparent heat capacity to model the phase change process more realistically.
Comparison of JRA-3Q OBS and the GEOtop simulation at the sites of (a) Ch04 during 2010–11 and (b) LdG during 2015–17. (c) The enlargement of the freezeback period for Ch04 [gray shadow in (a)], while (d) the enlargement of the freezeback period for LdG [gray shadow in (b)].
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0267.1
Since permafrost is defined strictly by temperature, this oscillation near the freezing point has important consequences for permafrost applications. The frequency distribution of JRA-3Q 6-hourly soil temperature is strongly asymmetric around about 0°C, with a positive modal value due to the omission of unfrozen water parameterization (Fig. 8). We would rather expect this distribution to have a modal value at or below the freezing point, as that is where the temperature change is slowed due to the effects of latent heat. Consequently, there is a strong possibility for misinterpretation of permafrost area characteristics (e.g., presence of ice). For example, in our estimates of permafrost area, the artificial positive soil temperature during the freeze/thaw period led to the remarkably underestimated permafrost area based on the 6-hourly soil temperature (Figs. 4b and 6).
The frequency histogram of JRA-3Q 6-hourly soil temperature during 1991–2012. Soil temperatures range from −1° to 1°C and are concerned at 0°C, and they were grouped with an interval of 0.05°C. The labels SL1 through SL7 correspond to each JRA-3Q soil layer.
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0267.1
As a further demonstration, we calculate permafrost area using JRA-3Q 6-hourly Ts with a 0.01°C threshold for inclusion. The estimated permafrost area increased significantly to about 11.4 ± 0.3 × 106 km2 (Figs. 4c and 6). This example highlights the limitations of the DECP method employed in the land surface model. This limitation restricts the suitability of JRA-3Q soil temperature in the interpretation of detailed permafrost characteristics, which generally require high temporal resolution (Figs. 6 and 4b,c). A few studies (e.g., Nicolsky et al. 2007; Zhang et al. 2008; Koven et al. 2013) have also reported the inaccuracies in soil temperature simulation based on the DECP method. However, the DECP method is still widely employed in many land surface models, such as CLM, CoLM, ORCHIDEE, and JSBACH (Tubini et al. 2021). The applications of such models in permafrost studies are generally restricted to coarse spatial and temporal scales (e.g., Nicolsky et al. 2007; Lawrence et al. 2012; Gouttevin et al. 2012; Porada et al. 2016). Furthermore, the depth of the soil column also contributes to the uncertainty of the estimated permafrost area (Burke et al. 2020). The deeper the soil column, the less likely it is that the deeply buried permafrost will be overlooked (Matthes et al. 2017).
6. Summary
The JRA-3Q reanalysis offers significant improvements in both the snow and soil schemes; it outperforms its predecessor (JRA-55) and other state-of-the-art reanalyses (ERA5-Land) when considering air and soil temperatures. Our results show that
-
JRA-3Q soil temperature has an overall RMSE of 2.21°C in permafrost regions, and we were able to create plausible estimates of permafrost extent from both the soil and air temperatures. This demonstrates the suitability of JRA-3Q for permafrost studies with coarse temporal resolution.
-
Active layer thickness is overestimated when derived from JRA-3Q soil temperature due to the warm bias in the thawing season.
-
JRA-3Q could not correctly represent the zero curtain or the effect of latent heat in freezing soil. This is due to the use of the decoupled energy conservation parameterization (DECP) method in the numerical solver for the soil heat conduction equations. This issue restricts the application of JRA-3Q soil temperature for simulating detailed permafrost phenomena requiring high temporal resolution.
-
Because DECP is present in many other land surface models (e.g., CLM, CoLM, ORCHIDEE, and JSBACH), our results suggest that these models may share the same limitation.
-
The omission of unfrozen water parameterization and simplified soil organic matter effects introduce additional uncertainties in JRA-3Q soil temperature.
The next generation of reanalyses and land surface models is expected to have improved spatiotemporal resolution. Therefore, we recommend that the active participation of the permafrost community in the development of future land surface models can be beneficial through the use of state-of-the-art numerical solvers to achieve more realistic heat transfer with phase change and better suitability for permafrost-specific applications.
Acknowledgments.
This research has been supported by the National Natural Science Foundation of China (42422608), the Youth Innovation Promotion Association CAS (2023075) to B. Cao, the Xinjiang Key Laboratory of Water Cycle and Utilization in Arid Zone (XJYS0907-2023-12), the State Key Laboratory of Geodesy and Earth’s Dynamics (SKLGED2023-5-1), and the Science and Technology Department of Tibet Program (XZ202301ZY0035G). The authors thank Hitoshi Yonehara, Yayoi Harada, and Hiroaki Naoe for their detailed introduction of the land surface scheme in JRA-3Q. We thank Stephan Gruber for his helpful comments. We also thank the three anonymous reviewers and the subject editor for their constructive comments that improved substantially the quality of the manuscript.
Data availability statement.
The JRA-3Q dataset can be downloaded from the Data Integration and Analysis System (DIAS, https://search.diasjp.net/en/dataset/JRA3Q). The in situ observations collected for the assessment have been summarized in Table 1. The CMA dataset from the National Meteorological Information Center is not publicly open but is available to registered researchers at http://data.cma.cn/en. The WDC dataset from the World Data Centers in Russia and Ukraine is available in the Supporting Information. All other datasets are open access (last access: 7 March 2025). The ECMWF dataset is available from the Climate Data Store (CDS, https://cds.climate.copernicus.eu/datasets/insitu-observations-surface-land), the USGS dataset is available from https://pubs.er.usgs.gov/publication/ds1092, the NPS datasets are available from https://irma.nps.gov/DataStore/Reference/Profile/2294387 and https://irma.nps.gov/DataStore/Reference/Profile/2287413 (NPS 2023), and the Nordicana D data are available from https://nordicana.cen.ulaval.ca/en_liste.aspx (Nordicana D 2023). GI-UAF is available from the Permafrost Laboratory of the University of Alaska (https://permafrost.gi.alaska.edu/sites_list) (Permafrost Laboratory 2023), the Tibet-OBS dataset is available from https://data.tpdc.ac.cn/zh-hans/data/805a6b93-201c-48ea-a131-27dd639d477a (TPDC 2023), the CTP-SMTMN datasets are available from https://data.tpdc.ac.cn/en/data/ef949bb0-26d4-4cb6-acc2-3385413b91ee/ and https://data.tpdc.ac.cn/zh-hans/data/39dcce47-c127-4ed3-9957-a2a7584a1ae3 (TPDC 2023), the GTN-P dataset is available from http://gtnpdatabase.org/boreholes (GTN-P 2023), the Yukon dataset is available from the Yukon Permafrost Database (https://service.yukon.ca/permafrost/Temperature.html) (Yukon 2024), the ISMN dataset is available from https://ismn.earth/en/dataviewer/ (ISMN 2024), and the datasets from Julia Boike are available from https://doi.pangaea.de/10.1594/PANGAEA.947032 (Boike et al. 2022a) and https://doi.pangaea.de/10.1594/PANGAEA.948951 (Boike et al. 2022b). Observed active layer thicknesses data are available from Peng et al. (2018). The IPA map is available from the National Snow and Ice Data Center (https://nsidc.org/data/GGD318/versions/2). The PZI map is available from https://doi.org/10.5194/tc-6-221-2012 (Gruber (2012). The TTOP map is available from https://doi.org/10.1016/j.earscirev.2019.04.023 (Obu et al. (2019), and the CP map is available from https://doi.org/10.1038/sdata.2019.37 (Karjalainen et al. (2019). The global annual land cover maps providing glacier information are available from the European Space Agency Climate Change Initiative (https://cds.climate.copernicus.eu/cdsapp#!/dataset/satellite-land-cover) (C3S, CDS 2019).
APPENDIX
Seasonal and Annual Performance of JRA-3Q
Figure A1 shows the seasonal and annual performance of JRA-3Q in the near-surface air temperature and soil temperature at different depths.
Taylor diagram of the seasonal and annual performance of JRA-3Q (without equal weight). The Ta is the near-surface air temperature, and SL1 through SL7 represent the soil temperature corresponding to each JRA-3Q soil layer, while “overall” represents the mean over the entire soil column. Circle markers represent winter (DJF), triangle markers represent summer (JJA), and square markers represent the entire year (annual). The gray dashed concentric rings refer to the centralized RMSE (CRMSE). Observations have a CRMSE of zero, along with a normalized standard deviation and a correlation of one, indicating that the closer a point is to this value, the better the performance.
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0267.1
REFERENCES
Anderson, E. A., 1976: A point energy and mass balance model of a snow cover. National Weather Service NOAA Tech. Rep. NWS 19, 150 pp., https://repository.library.noaa.gov/view/noaa/6392/noaa_6392_DS1.pdf.
Biskaborn, B. K., J.-P. Lanckman, H. Lantuit, K. Elger, D. A. Streletskiy, W. L. Cable, and V. E. Romanovsky, 2015: The new database of the Global Terrestrial Network for Permafrost (GTN-P). Earth Syst. Sci. Data, 7, 245–259, https://doi.org/10.5194/essd-7-245-2015.
Biskaborn, B. K., and Coauthors, 2019: Permafrost is warming at a global scale. Nat. Commun., 10, 264, https://doi.org/10.1038/s41467-018-08240-4.
Bob, S., and K. Yang, 2019: Time-lapse observation dataset of soil temperature and humidity on the Tibetan Plateau (2008–2016). National Tibetan Plateau/Third Pole Environment Data Center, accessed 7 March 2025, https://doi.org/10.11888/Soil.tpdc.270110.
Boike, J., and Coauthors, 2018: A 20-year record (1998–2017) of permafrost, active layer and meteorological conditions at a high Arctic permafrost research site (Bayelva, Spitsbergen). Earth Syst. Sci. Data, 10, 355–390, https://doi.org/10.5194/essd-10-355-2018.
Boike, J., and Coauthors, 2019: A 16-year record (2002–2017) of permafrost, active-layer, and meteorological conditions at the Samoylov Island Arctic permafrost research site, Lena River Delta, northern Siberia: An opportunity to validate remote-sensing data and land surface, snow, and permafrost models. Earth Syst. Sci. Data, 11, 261–299, https://doi.org/10.5194/essd-11-261-2019.
Boike, J., W. L. Cable, D. Y. Bolshiyanov, N. Bornemann, M. N. Grigoriev, I. Grunberg, and F. Miesner, 2022a: Continuous measurements in soil and air at the permafrost long-term observatory at Samoylov Station (2002 et seq). Alfred Wegener Institute - Research Unit Potsdam, PANGAEA, accessed 7 March 2025, https://doi.org/10.1594/PANGAEA.947032.
Boike, J., I. Grunberg, F. Miesner, N. Bornemann, and W. L. Cable, 2022b: Continuous measurements in soil and air at the permafrost long-term observatory at the bayelva station near ny-Ålesund (2018 et seq). PANGAEA, accessed 7 March 2025, https://doi.org/10.1594/PANGAEA.948951.
Boone, A., V. Masson, T. Meyers, and J. Noilhan, 2000: The influence of the inclusion of soil freezing on simulations by a soil–vegetation–atmosphere transfer scheme. J. Appl. Meteor., 39, 1544–1569, https://doi.org/10.1175/1520-0450(2000)039%3C1544:TIOTIO%3E2.0.CO;2.
Brown, J., O. Ferrians, J. Heginbottom, and E. Melnikov, 1997: Circum-Arctic map of permafrost and ground-ice conditions, circum-Pacific map series CP-45, scale 1: 10,000,000. U.S. Geological Survey in Cooperation with the Circum-Pacific Council for Energy and Mineral Resources Tech. Rep., 1 pp.
Broxton, P. D., X. Zeng, and N. Dawson, 2016: Why do global reanalyses and land data assimilation products underestimate snow water equivalent? J. Hydrometeor., 17, 2743–2761, https://doi.org/10.1175/JHM-D-16-0056.1.
Burke, E. J., Y. Zhang, and G. Krinner, 2020: Evaluating permafrost physics in the Coupled Model Intercomparison Project 6 (CMIP6) models and their sensitivity to climate change. Cryosphere, 14, 3155–3174, https://doi.org/10.5194/tc-14-3155-2020.
C3S, CDS, 2019: Land cover classification gridded maps from 1992 to present derived from satellite observations. C3S, CDS, accessed 7 March 2025, https://doi.org/10.24381/cds.006f2c9a.
C3S, CDS, 2021: Global land surface atmospheric variables from 1755 to 2020 from comprehensive in-situ observations. C3S, CDS, accessed 7 March 2025, https://doi.org/10.24381/cds.cf5f3bac.
Cao, B., and Coauthors, 2017: Spatial variability of active layer thickness detected by ground-penetrating radar in the Qilian Mountains, Western China. J. Geophys. Res. Earth Surf., 122, 574–591, https://doi.org/10.1002/2016JF004018.
Cao, B., X. Quan, N. Brown, E. Stewart-Jones, and S. Gruber, 2019a: GlobSim (v1.0): Deriving meteorological time series for point locations from multiple global reanalyses. Geosci. Model Dev., 12, 4661–4679, https://doi.org/10.5194/gmd-12-4661-2019.
Cao, B., T. Zhang, Q. Wu, Y. Sheng, L. Zhao, and D. Zou, 2019b: Brief communication: Evaluation and inter-comparisons of Qinghai–Tibet Plateau permafrost maps based on a new inventory of field evidence. Cryosphere, 13, 511–519, https://doi.org/10.5194/tc-13-511-2019.
Cao, B., S. Gruber, D. Zheng, and X. Li, 2020: The ERA5-Land soil temperature bias in permafrost regions. Cryosphere, 14, 2581–2595, https://doi.org/10.5194/tc-14-2581-2020.
Cao, B., G. Arduini, and E. Zsoter, 2022: Brief communication: Improving ERA5-Land soil temperature in permafrost regions using an optimized multi-layer snow scheme. Cryosphere, 16, 2701–2708, https://doi.org/10.5194/tc-16-2701-2022.
Chadburn, S., E. Burke, R. Essery, J. Boike, M. Langer, M. Heikenfeld, P. Cox, and P. Friedlingstein, 2015: An improved representation of physical permafrost dynamics in the JULES land-surface model. Geosci. Model Dev., 8, 1493–1508, https://doi.org/10.5194/gmd-8-1493-2015.
Dai, Y., and Coauthors, 2003: The common land model. Bull. Amer. Meteor. Soc., 84, 1013–1024, https://doi.org/10.1175/BAMS-84-8-1013.
Dall’Amico, M., S. Endrizzi, S. Gruber, and R. Rigon, 2011: A robust and energy-conserving model of freezing variably-saturated soil. Cryosphere, 5, 469–484, https://doi.org/10.5194/tc-5-469-2011.
Deardorff, J. W., 1978: Efficient prediction of ground surface temperature and moisture, with inclusion of a layer of vegetation. J. Geophys. Res., 83, 1889–1903, https://doi.org/10.1029/JC083iC04p01889.
Devoie, É. G., S. Gruber, and J. M. McKenzie, 2022: A repository of measured soil freezing characteristic curves: 1921 to 2021. Earth Syst. Sci. Data, 14, 3365–3377, https://doi.org/10.5194/essd-14-3365-2022.
Dorigo, W., and Coauthors, 2021: The international soil moisture network: Serving earth system science for over a decade. Hydrol. Earth Syst. Sci., 25, 5749–5804, https://doi.org/10.5194/hess-25-5749-2021.
Dorigo, W. A., and Coauthors, 2011: The International Soil Moisture Network: A data hosting facility for global in situ soil moisture measurements. Hydrol. Earth Syst. Sci., 15, 1675–1698, https://doi.org/10.5194/hess-15-1675-2011.
Dorigo, W. A., and Coauthors, 2013: Global automated quality control of in situ soil moisture data from the International Soil Moisture Network. Vadose Zone J., 12 (3), 1–21, https://doi.org/10.2136/vzj2012.0097.
Endrizzi, S., S. Gruber, M. Dall’Amico, and R. Rigon, 2014: Geotop 2.0: Simulating the combined energy and water balance at and below the land surface accounting for soil freezing, snow cover and terrain effects. Geosci. Model Dev., 7, 2831–2857, https://doi.org/10.5194/gmd-7-2831-2014.
FAO, IIASA, ISRIC, ISSCAS, and JRC, 2012: Harmonized world soil database (version 1.2). Accessed 7 March 2025, https://www.fao.org/soils-portal/data-hub/soil-maps-and-databases/harmonized-world-soil-database-v12/en/.
Farouki, O. T., 1981: The thermal properties of soils in cold regions. Cold Reg. Sci. Technol., 5, 67–75, https://doi.org/10.1016/0165-232X(81)90041-0.
Fiddes, J., S. Endrizzi, and S. Gruber, 2015: Large-area land surface simulations in heterogeneous terrain driven by global data sets: Application to mountain permafrost. Cryosphere, 9, 411–426, https://doi.org/10.5194/tc-9-411-2015.
Gouttevin, I., G. Krinner, P. Ciais, J. Polcher, and C. Legout, 2012: Multi-scale validation of a new soil freezing scheme for a land-surface model with physically-based hydrology. Cryosphere, 6, 407–430, https://doi.org/10.5194/tc-6-407-2012.
Gruber, S., 2012: Derivation and analysis of a high-resolution estimate of global permafrost zonation. Cryosphere, 6, 221–233, https://doi.org/10.5194/tc-6-221-2012.
GTN-P, 2023: GTN-P database. Accessed 7 March 2025, https://gtnp.arcticportal.org/.
Gubler, S., S. Endrizzi, S. Gruber, and R. S. Purves, 2013: Sensitivities and uncertainties of modeled ground temperatures in mountain environments. Geosci. Model Dev., 6, 1319–1336, https://doi.org/10.5194/gmd-6-1319-2013.
Guo, D., A. Wang, D. Li, and W. Hua, 2018: Simulation of changes in the near-surface soil freeze/thaw cycle using CLM4.5 with four atmospheric forcing data sets. J. Geophys. Res. Atmos., 123, 2509–2523, https://doi.org/10.1002/2017JD028097.
Harada, Y., S. Kobayashi, Y. Kosaka, J. Chiba, and T. Tokuhiro, 2021: Early results of the evaluation of the JRA-3Q reanalysis. EGU General Assembly 2021, Online, European Geosciences Union, EGU21-3762, https://doi.org/10.5194/egusphere-egu21-3762.
Herrington, T. C., C. G. Fletcher, and H. Kropp, 2024: Validation of pan-Arctic soil temperatures in modern reanalysis and data assimilation systems. Cryosphere, 18, 1835–1861, https://doi.org/10.5194/tc-18-1835-2024.
Hjort, J., O. Karjalainen, J. Aalto, S. Westermann, V. E. Romanovsky, F. E. Nelson, B. Etzelmüller, and M. Luoto, 2018: Degrading permafrost puts Arctic infrastructure at risk by mid-century. Nat. Commun., 9, 5147, https://doi.org/10.1038/s41467-018-07557-4.
Holloway, J. E., A. G. Lewkowicz, T. A. Douglas, X. Li, M. R. Turetsky, J. L. Baltzer, and H. Jin, 2020: Impact of wildfire on permafrost landscapes: A review of recent advances and future prospects. Permafrost Periglacial Processes, 31, 371–382, https://doi.org/10.1002/ppp.2048.
Hu, G., L. Zhao, R. Li, X. Wu, T. Wu, C. Xie, X. Zhu, and Y. Su, 2019: Variations in soil temperature from 1980 to 2015 in permafrost regions on the Qinghai-Tibetan Plateau based on observed and reanalysis products. Geoderma, 337, 893–905, https://doi.org/10.1016/j.geoderma.2018.10.044.
ISMN, 2024: The data hosting facility of the international soil moisture network. Accessed 7 March 2025, https://ismn.earth/en/.
Jin, H., and Coauthors, 2022: Permafrost degradation and its hydrogeological impacts. Water, 14, 372, https://doi.org/10.3390/w14030372.
Jordan, R. E., 1991: A one-dimensional temperature model for a snow cover: Technical Documentation for SNTHERM 89. U.S. Cold Regions Research and Engineering Laboratory Special Rep., 49 pp.
Jorgenson, M. T., and G. Grosse, 2016: Remote sensing of landscape change in permafrost regions. Permafrost Periglacial Processes, 27, 324–338, https://doi.org/10.1002/ppp.1914.
Jorgenson, M. T., C. H. Racine, J. C. Walters, and T. E. Osterkamp, 2001: Permafrost degradation and ecological changes associated with a warming climate in central Alaska. Climatic Change, 48, 551–579, https://doi.org/10.1023/A:1005667424292.
Karjalainen, O., J. Aalto, M. Luoto, S. Westermann, V. E. Romanovsky, F. E. Nelson, B. Etzelmüller, and J. Hjort, 2019: Circumpolar permafrost maps and geohazard indices for near-future infrastructure risk assessments. Sci. Data, 6, 190037, https://doi.org/10.1038/sdata.2019.37.
Kobayashi, S., and Coauthors, 2015: The JRA-55 reanalysis: General specifications and basic characteristics. J. Meteor. Soc. Japan, 93, 5–48, https://doi.org/10.2151/jmsj.2015-001.
Kosaka, Y., and Coauthors, 2024: The JRA-3Q reanalysis. J. Meteor. Soc. Japan, 102, 49–109, https://doi.org/10.2151/jmsj.2024-004.
Koven, C. D., W. J. Riley, and A. Stern, 2013: Analysis of permafrost thermal dynamics and response to climate change in the CMIP5 Earth system models. J. Climate, 26, 1877–1900, https://doi.org/10.1175/JCLI-D-12-00228.1.
Lawrence, D. M., A. G. Slater, and S. C. Swenson, 2012: Simulation of present-day and future permafrost and seasonally frozen ground conditions in CCSM4. J. Climate, 25, 2207–2225, https://doi.org/10.1175/JCLI-D-11-00334.1.
Li, M., P. Wu, and Z. Ma, 2020: A comprehensive evaluation of soil moisture and soil temperature from third-generation atmospheric and land reanalysis data sets. Int. J. Climatol., 40, 5744–5766, https://doi.org/10.1002/joc.6549.
Liu, S., Y. Wei, W. M. Post, R. B. Cook, K. Schaefer, and M. M. Thornton, 2013: The unified North American soil map and its implication on the soil organic carbon stock in North America. Biogeosciences, 10, 2915–2930, https://doi.org/10.5194/bg-10-2915-2013.
Matthes, H., A. Rinke, X. Zhou, and K. Dethloff, 2017: Uncertainties in coupled regional Arctic climate simulations associated with the used land surface model. J. Geophys. Res. Atmos., 122, 7755–7771, https://doi.org/10.1002/2016JD026213.
Melvin, A. M., and Coauthors, 2017: Climate change damages to Alaska public infrastructure and the economics of proactive adaptation. Proc. Natl. Acad. Sci. USA, 114, E122–E131, https://doi.org/10.1073/pnas.1611056113.
Miner, K. R., J. D’Andrilli, R. Mackelprang, A. Edwards, M. J. Malaska, M. P. Waldrop, and C. E. Miller, 2021: Emergent biogeochemical risks from Arctic permafrost degradation. Nat. Climate Change, 11, 809–819, https://doi.org/10.1038/s41558-021-01162-y.
Miner, K. R., and Coauthors, 2022: Permafrost carbon emissions in a changing Arctic. Nat. Rev. Earth Environ., 3, 55–67, https://doi.org/10.1038/s43017-021-00230-3.
Muñoz-Sabater, J., and Coauthors, 2021: ERA5-Land: A state-of-the-art global reanalysis dataset for land applications. Earth Syst. Sci. Data, 13, 4349–4383, https://doi.org/10.5194/essd-13-4349-2021.
Naoe, H., S. Kobayashi, Y. Kosaka, J. Chiba, T. Tokuhiro, and Y. Harada, 2021: Evaluation of a new Japanese reanalysis (JRA-3Q) in a pre-satellite era. EGU General Assembly 2021, Online, European Geosciences Union, EGU21-6881, https://doi.org/10.5194/egusphere-egu21-6881.
Nicolsky, D. J., V. E. Romanovsky, V. A. Alexeev, and D. M. Lawrence, 2007: Improved modeling of permafrost dynamics in a GCM land-surface scheme. Geophys. Res. Lett., 34, L08501, https://doi.org/10.1029/2007GL029525.
Nordicana D, 2023: The Nordicana D collection. Accessed 7 March 2025, https://nordicana.cen.ulaval.ca/en_liste.aspx.
NPS, 2023: NPS DataStore home. Accessed 7 March 2025, https://irma.nps.gov/DataStore/.
Obu, J., and Coauthors, 2019: Northern hemisphere permafrost map based on TTOP modelling for 2000–2016 at 1 km2 scale. Earth-Sci. Rev., 193, 299–316, https://doi.org/10.1016/j.earscirev.2019.04.023.
Oleson, K. W., and Coauthors, 2010: Technical description of version 4.0 of the Community Land Model (CLM). NCAR Tech. Note NCAR/TN-478+STR, 257 pp., https://doi.org/10.5065/D6FB50WZ.
O’Neill, H. B., S. L. Smith, C. R. Burn, C. Duchesne, and Y. Zhang, 2023: Widespread permafrost degradation and thaw subsidence in northwest Canada. J. Geophys. Res. Earth Surf., 128, e2023JF007262, https://doi.org/10.1029/2023JF007262.
Outcalt, S. I., F. E. Nelson, and K. M. Hinkel, 1990: The zero-curtain effect: Heat and mass transfer across an isothermal region in freezing soil. Water Resour. Res., 26, 1509–1516.
Peng, X., and Coauthors, 2018: Spatiotemporal changes in active layer thickness under contemporary and projected climate in the Northern Hemisphere. J. Climate, 31, 251–266, https://doi.org/10.1175/JCLI-D-16-0721.1.
Permafrost Laboratory, 2023: Data and Maps. Accessed 7 March 2025, https://permafrost.gi.alaska.edu/content/data-and-maps.
Porada, P., A. Ekici, and C. Beer, 2016: Effects of bryophyte and lichen cover on permafrost soil temperature at large scale. Cryosphere, 10, 2291–2315, https://doi.org/10.5194/tc-10-2291-2016.
Randall, D. A., and Coauthors, 1996: A revised land surface parameterization (SiB2) for GCMS. Part III: The greening of the Colorado State University general circulation model. J. Climate, 9, 738–763, https://doi.org/10.1175/1520-0442(1996)009<0738:ARLSPF>2.0.CO;2.
Schuur, E. A. G., and Coauthors, 2015: Climate change and the permafrost carbon feedback. Nature, 520, 171–179, https://doi.org/10.1038/nature14338.
Sellers, P. J., Y. Mintz, Y. C. Sud, and A. Dalcher, 1986: A Simple Biosphere Model (SIB) for use within general circulation models. J. Atmos. Sci., 43, 505–531, https://doi.org/10.1175/1520-0469(1986)043<0505:ASBMFU>2.0.CO;2.
Sellers, P. J., and Coauthors, 1996a: A revised land surface parameterization (SiB2) for atmospheric GCMS. Part I: Model formulation. J. Climate, 9, 676–705, https://doi.org/10.1175/1520-0442(1996)009<0676:ARLSPF>2.0.CO;2.
Sellers, P. J., C. J. Tucker, G. J. Collatz, S. O. Los, C. O. Justice, D. A. Dazlich, and D. A. Randall, 1996b: A revised land surface parameterization (SiB2) for atmospheric GCMS. Part II: The generation of global fields of terrestrial biophysical parameters from satellite data. J. Climate, 9, 706–737, https://doi.org/10.1175/1520-0442(1996)009<0706:ARLSPF>2.0.CO;2.
Slater, A. G., D. M. Lawrence, and C. D. Koven, 2017: Process-level model evaluation: A snow and heat transfer metric. Cryosphere, 11, 989–996, https://doi.org/10.5194/tc-11-989-2017.
Sousanes, P., K. Hill, and S. Miller, 2021: 2021 CAKN weather and climate data deliverables. Central Alaska Inventory and Monitoring Network, accessed 7 March 2025, https://irma.nps.gov/DataStore/Reference/Profile/2287413.
Sousanes, P., K. Hill, and S. Miller, 2022: 2022 ARCN weather and climate data deliverables. Arctic Inventory and Monitoring Network, accessed 7 March 2025, https://catalog.data.gov/dataset/2022-arcn-weather-and-climate-data-deliverables.
Streletskiy, D., O. Anisimov, and A. Vasiliev, 2015: Permafrost degradation. Snow and Ice-Related Hazards, Risks, and Disasters, Academic Press, 303–344, https://doi.org/10.1016/B978-0-12-394849-6.00010-X.
Sun, W., and Coauthors, 2023: Two-dimensional simulation of island permafrost degradation in northeastern Tibetan Plateau. Geoderma, 430, 116330, https://doi.org/10.1016/j.geoderma.2023.116330.
Tao, J., R. D. Koster, R. H. Reichle, B. A. Forman, Y. Xue, R. H. Chen, and M. Moghaddam, 2019: Permafrost variability over the Northern Hemisphere based on the MERRA-2 reanalysis. Cryosphere, 13, 2087–2110, https://doi.org/10.5194/tc-13-2087-2019.
TPDC, 2023: National Tibetan Plateau Data Center Third Pole Environment Data Center. Accessed 7 March 2025, https://data.tpdc.ac.cn/.
Tubini, N., S. Gruber, and R. Rigon, 2021: A method for solving heat transfer with phase change in ice or soil that allows for large time steps while guaranteeing energy conservation. Cryosphere, 15, 2541–2568, https://doi.org/10.5194/tc-15-2541-2021.
Urban, F. E., and G. D. Clow, 2018: DOI/GTN-P climate and active-layer data acquired in the National Petroleum Reserve–Alaska and the Arctic National Wildlife Refuge, 1998–2019 (ver. 1.2, June 2021). U.S. Geological Survey Data Series 1092, 71 pp., https://doi.org/10.3133/ds1092.
Verseghy, D. L., 1991: Class—A Canadian land surface scheme for GCMS. I. Soil model. Int. J. Climatol., 11, 111–133, https://doi.org/10.1002/joc.3370110202.
Wang, K., T. Zhang, and X. Zhong, 2015: Changes in the timing and duration of the near-surface soil freeze/thaw status from 1956 to 2006 across China. Cryosphere, 9, 1321–1331, https://doi.org/10.5194/tc-9-1321-2015.
Wang, K., and Coauthors, 2018: A synthesis dataset of permafrost-affected soil thermal conditions for Alaska, USA. Earth Syst. Sci. Data, 10, 2311–2328, https://doi.org/10.5194/essd-10-2311-2018.
Xiao, L., T. Che, and L. Dai, 2020: Evaluation of remote sensing and reanalysis snow depth datasets over the Northern Hemisphere during 1980–2016. Remote Sens., 12, 3253, https://doi.org/10.3390/rs12193253.
Yang, K., Y. Chen, L. Zhao, J. Qin, Z. La, X. Zhou, Y. Jiang, and J. Tian, 2022: Network of soil temperature and moisture on the Pali (2015–2021). National Tibetan Plateau Data Center, accessed 7 March 2025, https://doi.org/10.11888/Terre.tpdc.271918.
Yukon, 2024: Yukon Permafrost Database. Accessed 7 March 2025, https://service.yukon.ca/permafrost/.
Yukon Geological Survey, 2021: Yukon permafrost reports data. Yukon Permafrost Database, Government of Yukon, accessed 7 March 2025, https://service.yukon.ca/permafrost/Reports.html.
Zhang, P., and Coauthors, 2022: A dataset of 10-year regional-scale soil moisture and soil temperature measurements at multiple depths on the Tibetan plateau. Earth Syst. Sci. Data, 14, 5513–5542, https://doi.org/10.5194/essd-14-5513-2022.
Zhang, P., and Coauthors, 2023: A dataset of 10-year regional-scale soil moisture and soil temperature measurements at multiple depths on the Tibetan Plateau. National Tibetan Plateau Data Center, accessed 7 March 2025, https://doi.org/10.4121/20141567.v1.
Zhang, S.-Q., G.-Y. Ren, Y.-Y. Ren, Y.-X. Zhang, and X.-Y. Xue, 2021: Comprehensive evaluation of surface air temperature reanalysis over china against urbanization-bias-adjusted observations. Adv. Climate Change Res., 12, 783–794, https://doi.org/10.1016/j.accre.2021.09.010.
Zhang, Y., S. K. Carey, and W. L. Quinton, 2008: Evaluation of the algorithms and parameterizations for ground thawing and freezing simulation in permafrost regions. J. Geophys. Res., 113, D17116, https://doi.org/10.1029/2007JD009343.
Zhao, L., and Coauthors, 2021: A synthesis dataset of permafrost thermal state for the Qinghai–Tibet (Xizang) Plateau, China. Earth Syst. Sci. Data, 13, 4207–4218, https://doi.org/10.5194/essd-13-4207-2021.
