The impacts of vegetation on the soil surface freezing-thawing processes at permafrost southern edge simulated by an improved process-based ecosystem model
Introduction
Permafrost (the ground that remains below 0 °C for at least two consecutive years) underlies approximately 23.9% of the exposed land surface of the Northern Hemisphere and it contains twice as much carbon as the atmosphere (Grosse et al., 2016; Schädel et al., 2016). The large quantities of carbon stored in the frozen soil could be released into the atmosphere due to permafrost degradation caused by climatic warming. This degradation would potentially change the global carbon budget and accelerate global climate warming (Chaudhary et al., 2020; Hollesen et al., 2011). The most dramatic permafrost degradation mainly occurs at the southern edge of permafrost area due to the significant poleward movement of permafrost caused by climate warming (Guo et al., 2018; Yue et al., 2020). The soil surface freezing-thawing (FT) processes, which are measured by the amplitude and duration of the diurnal soil FT cycle as well as seasonal FT timing, would effectively mediate the underlying permafrost degradation (Guo et al., 2018). The soil FT processes during the transition period from cold-to-warm or warm-to-cold seasons can last from a few days to a few weeks and they are susceptible to change because of the rapid climate warming (Bao et al., 2021; Pachauri et al., 2014).
There have been many researches investigating the soil FT processes in response to climatic warming, mainly using near-surface soil temperature data from in situ measurements (Bao et al., 2021; Cao et al., 2020; Wang et al., 2020; Yoshikawa et al., 2020), land surface temperature data from satellites (Baltzer et al., 2014; Kim et al., 2014) as well as ecosystem model simulations (Foster et al., 2019; Guo and Wang, 2014). However, there are still uncertainties in these studies due to the sparse and short-term field measurements (Tao et al., 2017), coarse resolution of the available remote sensing data (Grosse et al., 2016; Tao et al., 2017) or the poor representation of soil FT processes in ecosystem models (Barman and Jain, 2016; Jan et al., 2020; Nitzbon et al., 2020). To better understand the pattern and drivers of the soil FT processes at regional scale, the more realistic representation and parameterizations of soil FT processes are needed to be incorporated into ecosystem models.
The impacts of vegetation on soil FT processes have been examined by many studies. Field measurements in south Siberia (Hu et al., 2013), Zhangbei county of Hebei province, China (Chen et al., 2020) and Qinghai-Tibetan Plateau (Hu et al., 2020a) found that vegetation could exert buffer effect on soil temperature, and consequently affect the soil FT processes. Guo et al. (2018) found that soil FT cycles would be larger in amplitude and longer in duration at steppe sites than at forest sites. Soils in the forest sites and steppe sites froze almost simultaneously, but experience a delay in thawing for forest sites (Guo et al., 2018). In addition, the hydrologic and thermal regimes of soil surface and soil FT processes could be affected by vegetation through intercepting snowfall and solar radiation (Chang et al., 2015; French, 2007; Karjalainen et al., 2019) and through buffer effect of litter layer (Chaudhary et al., 2020; Lawrence and Slater, 2008). Thus, it is essential to improve the ecosystem model for better simulating the impacts of vegetation on soil FT processes.
Three forest sites and two grassland sites investigated in this study were located along a vast latitudinal (around 10°) and longitudinal (around 20°) gradient at the south edge of the permafrost region of Mongolia and northeastern China. Soil temperature and meteorological data were recorded at these sites during 2003 to 2014. The soil-vegetation thermal and hydraulic modules of the Boreal Ecosystems Productivity Simulator (BEPS) have been improved and tested at the studied sites. Using the measured soil temperature data and those simulated by the BEPS model, we investigated the impacts of vegetation types on soil FT processes.
Section snippets
Site description
Three forest sites and two grassland sites at the southern edge of a permafrost region in Mongolia and northeastern China were investigated in this study (Fig. 1 and Table 1). Three sites (SKT, DXF, and KBU) are in the permafrost region, and two sites (CHB and CNG) belong to the region of seasonally frozen ground. These sites were selected fulfilling the following conditions in data quality and availability: (1) soil data including soil texture and soil organic carbon (SOC) content are
Simulation of soil temperature
The hourly soil temperature simulated by the original and improved BEPS model were compared with measured data to evaluate the performance of the original and improved BEPS model in simulating soil temperature at the studied sites (Supplementary materials, Fig. S1). The regression statistics are summarized in Table 2. The original and improved BEPS models are both able to explain >90% of the variance in soil temperature. The root mean square error (RMSE) of soil temperature simulated by the
The impacts of different improvement schemes on the simulations of soil temperature
The Gaussian kernel density estimation (KDE) of the differences between the simulated soil temperatures from the experimental model runs and the observations of the five studied sites was conducted to compare the contribution of the four modifications of BEPS model in simulating soil temperature. Considering the impacts of canopy and wind speed on the snowpack density reduces the MAE of simulated soil temperature from 3.23 to 2.90 °C (Fig. 7). The incorporation of puddled water into the
Conclusions
In this study, the BEPS model has been improved through more careful treatments of the impacts of thermal “buffer zone” (i.e. snowpack, puddled water and SOM) on subsurface heat transfer. The different patterns of soil surface FT processes between forest and grassland and the regulation of vegetation to soil FT processes were analyzed based on the measured soil temperature data and those simulated by the BEPS model. The major conclusions are drawn as follows:
- (1)
Compared to the original BEPS model,
CRediT
Zhenhai Liu: Conceptualization, Software, Writing - Original draft preparation.
Bin Chen: Conceptualization, Methodology, Writing - Review and Editing.
Shaoqiang Wang: Writing - Review and Editing, Supervision, Project administration.
Qinyi Wang: Validation, Data Curation.
Jinghua Chen: Software.
Weibo Shi: Software.
Xiaobo Wang: Software.
Yuanyuan Liu: Visualization.
Yongkai Tu: Visualization.
Mei Huang: Supervision.
Junbang Wang: Supervision.
Zhaosheng Wang: Supervision.
Hui Li: Data Curation.
Tongtong
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
The authors sincerely appreciate Prof. Jingming Chen for sharing the codes of the BEPS model. This study used eddy covariance data acquired and shared by ChinaFLUX, AsiaFlux and FLUXNET. The GLOBMAP-V2 LAI datasets and SoilGrids dataset are provided by Yang Liu and Tomislav Hengl. We greatly offer our profound appreciation to all the providers of the freely available data. This research was supported by the Science and Technology Strategic Pilot of the Chinese Academy of Sciences (XDA20030203).
References (67)
- et al.
Remote sensing-based ecosystem-atmosphere simulation scheme (eass)-model formulation and test with multiple-year data
Ecol. Modell.
(2007) - et al.
Assessment of foliage clumping effects on evapotranspiration estimates in forested ecosystems
Agric. For. Meteorol.
(2016) - et al.
Importance of tree- and species-level interactions with wildfire, climate, and soils in interior alaska: implications for forest change under a warming climate
Ecol. Modell.
(2019) - et al.
A comparison of digital and film fisheye photography for analysis of forest canopy structure and gap light transmission
Agric. For. Meteorol.
(2001) - et al.
CO2 fluxes over an old, temperate mixed forest in northeastern china
Agric. For. Meteorol.
(2006) - et al.
Vegetation can strongly regulate permafrost degradation at its southern edge through changing surface freeze-thaw processes
Agric. For. Meteorol.
(2018) - et al.
Global clumping index map derived from the modis brdf product
Rem. Sens. Environ.
(2012) - et al.
Forest buffers soil temperature and postpones soil thaw as indicated by a three-year large-scale soil temperature monitoring in the forest-steppe ecotone in inner asia
Glob. Planet. Change
(2013) - et al.
Review of algorithms and parameterizations to determine unfrozen water content in frozen soil
Geoderm.
(2020) - et al.
Modelling multi-year coupled carbon and water fluxes in a boreal aspen forest
Agric. For. Meteorol.
(2006)