An improved model for soil surface temperature from air
temperature in permafrost regions of Qinghai-Xizang (Tibet)
Plateau of China
Received: 7 December 2015 / Accepted: 31 May 2016 / Published online: 17 June 2016
Ó Springer-Verlag Wien 2016
Abstract Soil temperature plays a key role in hydro-ther-
mal processes in environments and is a critical variable
linking surface structure to soil processes. There is a need
for more accurate temperature simulation models, particu-
larly in Qinghai-Xizang (Tibet) Plateau (QXP). In this
study, a model was developed for the simulation of hourly
soil surface temperatures with air temperatures. The model
incorporated the thermal properties of the soil, vegetation
cover, solar radiation, and water ﬂux density and utilized
ﬁeld data collected from Qinghai-Xizang (Tibet) Plateau
(QXP). The model was used to simulate the thermal regime
at soil depths of 5 cm, 10 cm and 20 cm and results were
compared with those from previous models and with
experimental measurements of ground temperature at two
different locations. The analysis showed that the newly
developed model provided better estimates of observed
ﬁeld temperatures, with an average mean absolute error
(MAE), root mean square error (RMSE), and the normal-
ized standard error (NSEE) of 1.17 °C, 1.30 °C and
13.84 %, 0.41 °C, 0.49 °C and 5.45 %, 0.13 °C, 0.18 °C
and 2.23 % at 5 cm, 10 cm and 20 cm depths, respectively.
These ﬁndings provide a useful reference for simulating soil
temperature and may be incorporated into other ecosystem
models requiring soil temperature as an input variable for
modeling permafrost changes under global warming.
Soil temperature plays an important role in many physical,
chemical, and biological processes in terrestrial ecosystems
by regulating the lower boundary of mass and energy
exchange between the land and atmosphere (Bond-Lam-
berty and Thomson 2010; Canadell et al. 2007; Huang et al.
2014; Liang et al. 2014; Paul et al. 2004; Toosi et al. 2014).
Temperature is a required input in process-based land
surface models used to estimate carbon, water, or energy
balances along the soil–plant–atmosphere continuum (Niu
et al. 1997; Riveros-Iregui et al. 2011; Thornton et al.
2002). It is also an important indicator of climate change
and a critical parameter in numerical weather forecasting
and climate prediction (Holmes et al. 2008; Qian et al.
2011; Zhang et al. 2001). However, temperature mea-
surements at the land surface and at various soil depths are
spatially and temporally limited, necessitating the devel-
opment of accurate soil temperature simulations (Niu et al.
1997). More accurate predictions of the spatial and tem-
poral patterns of soil temperature can also enhance our
understanding of the dynamics of vegetation and soil
organic matter in different landscapes (Kang et al. 2000).
Soil temperature can be estimated by three different
methods: (1) physical approaches, which are based on soil
heat ﬂow and energy balance (Mihalakakou 2002; Thun-
holm 1990); (2) empirical approaches, which expand upon
regression-based relationships between air and soil tem-
perature or elevation and soil temperature (Plauborg 2002;
Zheng et al. 1993); and (3) hybrid method, which combine
physical and empirical method (Kang et al. 2000; Liang
et al. 2014). Physical approaches typically focus on bare
soil surfaces and depend strongly upon the initial and
boundary conditions (Gao 2005; Gao et al. 2003, 2008;
Shao et al. 1998). They are traditionally difﬁcult to apply in
Responsible Editor: J.-F. Miao.
& Lin Zhao
Cryosphere Research Station on Qinghai-Xizang Plateau,
State Key Laboratory of Cryospheric Sciences, Cold and Arid
Regions Environmental and Engineering Research Institute,
Chinese Academy of Sciences, Lanzhou 730000, Gansu,
Meteorol Atmos Phys (2017) 129:441–451