instance, more than 5,700 dams and 5,000 floodgates have
been constructed in the Huai River Basin, with a drainage
area of 2.7 × 10
5
km
2
in the eastern China, and most rivers
are now regulated by water projects, resulting in dramatic
changes in the hydrological regimes and increased pollution
discharge (Zhang
et al.
, 2010). Consequently, the extensive
water project construction has played an important role in
affecting the lake changes in these areas.
In contrast, significant positive correlations are traced
between climatic factors (e.g., annual mean precipitation and
temperature) and the lake changes in the
TPL
(
P
<0.05). Over
the past few decades, an increase of precipitation has been
observed in the Inner Tibetan Plateau based on a high-reso-
lution regional climate simulation (Gao
et al.
, 2015), and the
cumulative net precipitation (precipitation minus evapotrans-
piration) has also shown an apparent increasing trend since
1998 (Zhang
et al.
, 2017a). The increased precipitation (snow-
fall and rainfall) is an important source of water for the lakes
on the Tibetan Plateau, and has contributed to the increase
in water level, particularly for the salt lakes in closed basins
(Zhang
et al.
, 2011). For example, Qinghai Lake, China’s
largest lake within a closed basin, experienced a continuous
increase in surface area (increased by 153.9 km
2
) from 2006
to 2015, which was primarily associated with precipitation
and evaporation (Cui
et al.
, 2017; Li
et al.
, 2007). Similarly,
the air temperature records from the available meteorological
stations in the
TPL
have indicated an obvious increase in tem-
perature (
P
<0.05, Figure 7a), which is highly correlated with
the lake area change patterns (
P
<0.01). The climate on the
Tibetan Plateau has been experiencing drastic changes during
the past three decades, with a rapid warming trend (0.3°C per
decade) at the altitudes above 4,000 m, which is twice the
rate of observed global warming (Xu
et al.,
, 2009). The climate
warming has exerted a great influence on the cryosphere
and hydrological cycle, including accelerated glacier/snow
melting and permafrost degradation (Li
et al.
, 2014, Neckel
et
al.
, 2014), which has contributed to the rapid enlargement of
lakes (Zhang
et al.
, 2017a).
According to the Second Chinese Glacier Inventory, the
glacier/snow covers of the Tibetan Plateau decreased by 9.5
to 26% between the 1970s and 2010 (Guo
et al.
, 2015). Selin
Co, the largest salt water lake in Tibet, expanded rapidly
from 1975 to 2008, with an area increase of 420 km
2
/10a on
average, which was mainly due to the increase in melt water
from the mountainous glaciers/snow under the background of
global warming (Duo
et al.,
, 2010).
North China, especially Xinjiang and Inner Mongolia,
is characterized by an arid and semiarid climate with low
annual precipitation and high annual total solar radiation
(Xie and Wang, 2007; Yang
et al.
, 2003). Air temperature and
precipitation have displayed an increasing tendency since the
1980s in the northwestern China, suggesting a climatic shift
to a warm humid pattern in this region (Shi
et al.
, 2003). A
significant positive relationship can be observed between the
annual mean precipitation and lake area changes (
P
<0.01)
in Xinjiang, which is located in the hinterland of the arid
areas in Northwest China. Moreover, anthropogenic activities,
such as increased water consumption by agricultural irriga-
tion, have also played a role in influencing the lake changes
in Xinjiang. For example, the Tarim River Basin, located in
the southern Xinjiang, experienced rapid population growth
(increased by 183.3%) and accelerated expansion of cultivat-
ed land (increased by 70%) from 1949 to 2000, posing great
threat to the natural riparian ecosystems (Chen
et al.
, 2011;
Zhang
et al.
, 2010). In addition, the decreasing trend of the to-
tal lake area in Inner Mongolia is significantly associated with
intensive human activities, including agricultural irrigation,
coal mining, and grazing (
P
<0.05). Over the past 30 years, the
area of irrigated cropland in Inner Mongolia has increased by
218.6% (from 9.7×10
3
km
2
to 30.9×10
3
km
2
), resulting in the
overexploitation of groundwater and river water in this re-
gion. The expansion of cultivation and the increasing number
of livestock (including sheep, goat, and cattle) have also led
to the degradation of the typical steppe ecosystem distributed
throughout Inner Mongolia, affecting the soil function and
water conservation of grasslands (Sasaki
et al.
, 2008; Schön-
bach
et al.
, 2011). For example, the total area of degraded
steppe due to overgrazing expanded by 498.3 km
2
or 6.9%
from 1985 to 1999 in Xilin River Basin (Tong
et al.
, 2004).
Moreover, the coal production in Inner Mongolia showed an
abrupt increase after the year 2000, accompanied by the rapid
decline in lake area. Coal industry is highly water intensive,
which needs large volumes of water for mining activities
(consuming 2.54 m
3
of water to mine every ton of coal), and
also impacts the local water balance, leading to groundwater
level declining, water and soil loss, and land desertification
(Pan
et al.
, 2012; Zhang
et al.
, 2013). Along with the rapid
progress of industrialization in China, the water demand in
the coal industry could dramatically increase, posing seri-
ous threats to the local environment and water system in the
ecologically fragile regions (Pan
et al.
, 2012).
Limitations and Future Work
In this study, we investigated the spatiotemporal changes in
the lakes across the entire China and the climatic/anthropo-
genic drivers from 1985 to 2015. To provide a quantitative
understanding of driving factors, correlation analysis between
climatic/human factors and lake changes was performed.
However, the correlation analysis may induce some potential
uncertainties in estimating the relationship between lake
changes and driving factors. First, correlation analysis was
performed at the aggregated scale over the five lake zones,
which could be subject to modifiable areal unit problem
(
MAUP
) or ecological fallacy (Fotheringham and Wong, 2015).
Ideally, correlation analysis between area change of each lake
and the associated human/climate factors can provide more
details in the analysis of driving factors. Due to the limited
data availability of human factors (i.e., irrigation area and coal
production) for an individual lake, the correlation analysis
was conducted over the five typical lake-zones based on the
climatic, geographic, and socioeconomic conditions across
China.
Moreover, since the diverse climatic/human factors have
driven profound influence on the lakes across China, the
direct causal-effect between lake changes and driving fac-
tors could be difficult to be indicated by correlation analysis.
Although most of the driving factors used in this study were
collected following the previous studies (Ma
et al.
, 2010; Tao
et al.,
, 2015), more climatic/human variables are needed to
explain the dynamic patterns of China’s lakes. The dynamic
lake systems in China have been suffering from dramatic
changes under various natural and anthropogenic impacts.
Simple correlation analysis may be not capable of modeling
the complex relationship between lake changes and driving
factors. Additionally, to avoid the lag impacts of temperature
or precipitation on lake area, more subtle monitoring of lake
changes at a one-year interval using accumulated precipita-
tion and average temperature could be performed. Therefore,
the detailed investigations of spatiotemporal pattern (intra-
year or inter-year scale) of lake changes and the exact driving
forces by using more complete remote sensing observations
and natural and socio-economic data are still required in the
targeted research. At present, the quantitative information
about the long-term evolution patterns of the nationwide
lakes and the associated driving forces still remains unclear.
In this study, the quantitative evaluation of the dramatic lake
664
October 2018
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
