1 MARCH 2011
YANG ET AL.
1525
On the Climatology and Trend of the Atmospheric Heat Source over the
Tibetan Plateau: An Experiments-Supported Revisit
KUN YANG AND XIAOFENG GUO
Laboratory of Tibetan Environment Changes and Land Surface Processes, Institute of Tibetan Plateau Research,
Chinese Academy of Sciences, Beijing, China
JIE HE
Laboratory of Tibetan Environment Changes and Land Surface Processes, Institute of Tibetan Plateau Research,
Chinese Academy of Sciences, and Graduate University of Chinese Academy of Sciences, Beijing, China
JUN QIN
Laboratory of Tibetan Environment Changes and Land Surface Processes, Institute of Tibetan Plateau Research,
Chinese Academy of Sciences, Beijing, China
TOSHIO KOIKE
Department of Civil Engineering, University of Tokyo, Tokyo, Japan
(Manuscript received 14 May 2010, in final form 27 October 2010)
ABSTRACT
Atmospheric heating over the Tibetan Plateau (TP) enhances the Asian summer monsoon. This study
presents a state-of-the-art estimate of the heating components and their total over the TP, with the aid of highaccuracy experimental data, an updated land surface model, and carefully selected satellite data.
The new estimate differs from previous estimates in three aspects: 1) different seasonality—the new estimation shows the maximum total heat source occurs in July (the mature period of the monsoon), rather than
in the previously reported month of May or June (around the onset of the monsoon), because previous studies
greatly overestimated radiative cooling during the monsoon season [June–August (JJA)]; 2) different regional pattern—the eastern TP exhibits stronger heating than the western TP in summer, whereas previous
studies gave either an opposite spatial pattern because of overestimated sensible heat flux over the western TP
or an overall weaker heat source because of overestimated radiative cooling; and 3) different trend—sensible
heat, radiative convergence, and the total heat source have decreased since the 1980s, but their weakening
trends were overestimated in a recent study. These biases in previous studies are due to fairly empirical
methods and data that were not evaluated against experimental data.
1. Introduction
The Tibetan Plateau (TP) exerts a huge heat source on
the atmosphere in the Northern Hemisphere during boreal spring and summer (Flohn 1957; Ye and Gao 1979).
Because the average elevation of the TP is more than
4000 m MSL and thus the total mass of the air column
over the TP is much less than that over its surroundings,
Corresponding author address: Kun Yang, Professor, Institute of
Tibetan Plateau Research, Chinese Academy of Sciences, P.O.
Box 2871, Beijing 100085, China.
E-mail: [email protected]
DOI: 10.1175/2010JCLI3848.1
Ó 2011 American Meteorological Society
the atmospheric heating is more efficient in this region.
Furthermore, the sensible heat from the TP ground can
directly warm the midtroposphere. It has been indicated
that the heating over the TP not only enhances the Asian
monsoon circulation (He et al. 1987; Yanai et al. 1992; Wu
and Zhang 1998; Qian et al. 2004; Sato and Kimura 2007)
but also triggers vigorous deep convections over the TP
(Yang et al. 2004) that greatly enhance the troposphere–
stratosphere exchanges of water vapor and air pollutants
(Fu et al. 2006). Zhou et al. (2009) even suggested that the
thermodynamic processes over the TP may impact the
climate at a hemispheric scale and even a global scale
through modulating atmosphere–ocean interactions.
1526
JOURNAL OF CLIMATE
Meanwhile, the TP exhibits a striking warming since the
1980s, which may have influenced rainfall in east China
(Wang et al. 2008) and interacted with the weakening of
the Asian monsoon in recent decades (Duan and Wu
2008, hereafter DW08). Recent field experiments on the
TP have advanced our understanding of the heating processes from the atmospheric boundary layer up to the free
atmosphere (Yang et al. 2004; Taniguchi and Koike 2007;
Tamura et al. 2010), but how to quantify the heat source
and its trend is still a major topic of TP meteorology.
A high-accuracy estimate of the heat source over the TP
is a basis for testing the capability of a climate model or
a reanalysis system to reproduce the atmospheric energy
cycle in the Asian monsoon system. Such an estimate also
contributes to setting reasonable conditions (e.g., surface
sensible heat flux, diabatic heating) for process studies by
numerical modeling (e.g., Kuwagata et al. 2001; Liu et al.
2007; Wu et al. 2007).
The total heat source (TH) comprises three components, that is, surface sensible heat (SH), latent heat release of condensation (LH), and radiative convergence
(RC). In the literature, there are two approaches to estimate the heat source.
The first approach is to vertically integrate the local
change rate and advection of energy derived from a reanalysis dataset. The accuracy of this approach highly
depends on the amount of observations assimilated into
an analysis model. The TP has an area of about 2.5 3
106 km2, but there are only a few operational radiosonde
stations whose data have been assimilated into current
reanalysis products. Therefore, a high-accuracy reanalysis dataset is still absent, except for several episodes,
such as the First Global Atmospheric Research Program (GARP) Global Experiment (FGGE) during the
summer of 1979 and the Global Energy and Water
Cycle Experiment (GEWEX) Asian Monsoon Experiment–Tibet (GAME–Tibet) during the summer of 1998.
Heat source estimations using these episode data provided insights into the relationship between the atmospheric heating over the TP and the summer monsoon’s
seasonal march (Luo and Yanai 1984; Yanai et al. 1992;
Ueda et al. 2003); however, they are not applicable of
long-term investigations, such as interannual variability
and climatic change of the heat source.
The second approach to estimate the heat source is
based on surface data and/or satellite products. This type
of estimation was first presented in Ye and Gao (1979)
and followed by many studies (e.g., Chen et al. 1985;
Zhao and Chen 2000b; Li and Chen 2003). Among various components, SH and LH were estimated from the
China Meteorological Administration (CMA) station
data. However, shortcomings are obvious in these studies.
First, SH was estimated by the bulk method with a
VOLUME 24
constant heat transfer coefficient or a simple parameterization. Yang et al. (2011) pointed out these empirical
methods may produce opposite results in SH trend, as
they do not follow fundamental micrometeorological
theory. Second, the term of radiative convergence was
estimated from satellite products without recognizing
their biases. Yang et al. (2006a) called attention to the
accuracy of current satellite estimates of radiation budget.
Particularly, they may have considerable biases for the TP
region because of its extreme climatic conditions. Third,
the CMA stations are irregularly distributed, but their
representativeness for use of estimating the heat source
has not been clarified. Therefore, large uncertainties may
exist in previous estimates because of limitations in both
data availability and methods. These issues will be addressed in this study.
As shown later, previous estimates of the heat source
gave rather mixed results. Meanwhile, several hydrometeorological experiments have been implemented on the TP
since the end of the 1990s. The obtained data have significantly advanced our understanding of land–atmosphere
interaction processes, and they have been applied to land
model development (Yang et al. 2009a; van der Velde
et al. 2009; Chen et al. 2010). Although these experimental
data have been available for nearly 10 years, they were not
used for estimating the heat source in previous studies.
The objective of this study is to provide a state-of-theart estimate of the TP heat source terms and their trends
during recent decades. In contrast to previous estimates,
the new estimate relies on recent advances in field data
analyses, land surface modeling, and satellite products’
evaluation. The heat source was estimated for 3 TP subregions with distinct climate: dry western TP or W–TP
(west of 858E), transitional central TP or C–TP (858–
958E), and wet eastern TP or E–TP (east of 958E).
The paper is organized as follows. Three types of data
are first introduced (section 2) followed by a new method
and a conventional method to quantify the components of
the heat source (section 3). Then, the climatology of the
estimated heat source terms and their trends are presented
and compared with a conventional empirical estimate
(sections 4 and 5, respectively). The representativeness of
the heat source terms estimated from sparsely distributed
stations are discussed in section 6, and concluding remarks are given in section 7.
2. Data
a. Satellite data
Two satellite products of surface radiation budget and
top of the atmosphere (TOA) radiation budget were used
in this study. They are the GEWEX Surface Radiation
Budget (GEWEX SRB; Stackhouse et al. 2004) and the
1 MARCH 2011
YANG ET AL.
1527
FIG. 1. Distribution of CMA stations on the TP. The four stations marked by a circle within
a circle are collocated with Tibetan experimental sites. They are SQH and Gerze on the
W–TP and Naqu and Anduo (Amdo) on the C–TP. Annual precipitation contours (mm)
during 1984–2006 are shown, after correction by the algorithm in Ye et al. (2004).
International Satellite Cloud Climatology Project Flux
Data (ISCCP–FD; Zhang et al. 2004). The former was
estimated with the cloud cover and radiances from the
ISCCP stage DX (ISCCP–DX) nominal 30-km pixels
within each 18 3 18 cell. The latter was estimated with the
ISCCP, stage D1 (ISCCP-D1) cloud cover, cloud-top
temperature, optical thickness, and cloud phase based on
15 cloud types in a 280-km (2.58) equal-area grid with
climatologies for cloud particle size and vertical structure.
The GEWEX SRB product has several versions, and this
study used version 3.0 for shortwave radiation (SW) and
version 2.0 for longwave radiation (LW).
Another satellite product involved in this study is the
Tropical Rainfall Measuring Mission (TRMM) 3B42
(V6) precipitation, with a 3-h temporal resolution and a
0.258 spatial resolution in a global belt spanning 508S–
508N in latitude. This product is a fusion of microwave
precipitation estimates and infrared precipitation estimates, which was then scaled to match monthly rain
gauge analyses (Huffman et al. 2007).
The radiation products are available since 1984, and
they are used to estimate RC and evaluate the spatial representativeness of station data. The TRMM precipitation data are available since 1998, which are used to
upscale LH from stations to a regional scale.
b. CMA routine data
CMA provides long-term station data for this study.
Figure 1 shows the distribution of CMA stations on the TP
(marked by solid circles in the figure). There are a total of
85 stations. Each station started operations in different
years, and some measurements are not available at the
early stage. There are a total of 21 stations that have
‘‘good data records’’ (defined as data records that are
available for more than 300 days of a year) since 1961;
however, the number of the stations was increased to 78
during 1984–2006, of which there are 3, 20, and 55 on the
W-TP, C-TP, and E-TP, respectively. The simultaneous
availability of both satellite data and the CMA dataset
makes this study focus on the heat source estimation
during 1984–2006.
At each CMA station, measured meteorological parameters are 6-hourly [0200, 0800, 1400, 2000 Beijing
standard time (BST); BST 5 UTC 1 8 h] wind speed, air
temperature, and specific humidity; maximum, minimum,
and daily mean air pressure; daily precipitation; and daily
sunshine duration. Instead of being directly used to estimate sensible heat flux in many previous studies, these
observed data are used herein to drive a land surface
model (LSM) that produces the surface net radiation and
sensible heat flux.
Figure 2 shows the observed trends in annual-mean
wind speed and air temperature in the subregions. It is
shown that the wind speed has significantly decreased,
which is a common phenomenon in China (Xu et al.
2006), while the air temperature increased during recent
decades. More details of the temperature trend were analyzed with satellite data (Oku et al. 2006; Qin et al. 2009)
and reanalysis data (You et al. 2010). The response of SH
to the TP climate changes was investigated in this study.
c. Tibetan experimental data
The GAME–Tibet in 1998 and the follow-on postGAME and Coordinated Enhanced Observation Period
1528
JOURNAL OF CLIMATE
VOLUME 24
FIG. 2. Trends in annual-mean (a) wind speed and (b) air temperature during 1984–2006 at
CMA stations on the western, central, and eastern TP.
(CEOP) experiments during 1999–2004 collected highaccuracy data for TP hydrometeorological studies (Koike
et al. 1999; Ueno et al. 2001; Ma et al. 2002b; Tanaka et al.
2003). Among these experimental stations, four automatic weather stations (AWSs) were collocated with
CMA stations, and thus they were selected to evaluate
land surface modeling results and satellite products in
this study. The AWSs are marked by a circle within a
circle in Fig. 1. Two stations were deployed on the W–TP
[Shiquanhe (SQH) and Gerze (or Gaize)] and the other
two stations on the C–TP (Naqu and Anduo). The measuring periods at the experimental sites are different:
May–September 1998 for SQH, May–June 1998 and
October 2002–December 2004 for Gerze, June–September
1998 and October 2000–August 2004 for Naqu, and June
1998–May 2003 and May–December 2004 for Anduo.
The experimental data are used to evaluate heat source
estimates.
3. Method description
A new estimation and a typical conventional method
were presented and compared to each other. The conventional method is very similar to the one presented in
DW08. Herein, the phrase ‘‘conventional method’’ means
that the used data and method are either empirical or not
evaluated against experimental data. Table 1 summarizes
the differences between the two methods and the details
are presented below.
a. Estimating sensible heat flux
1) CONVENTIONAL METHOD
Conventionally, SH was estimated from CMA station
data by the following bulk method:
SH 5 rcp CH u(T g
T a ),
(1)
where r (kg m23) is the air density, cp (51004 J kg21 K21)
is the specific heat of air at constant pressure, u (m s21) is
the observed wind speed at level zm (m), Tg (K) is the
observed ground temperature, Ta (K) is the observed air
temperature at level zh (m), and CH is the bulk heat transfer
coefficient.
There are a number of studies on how to determine CH
value for the TP. A brief review of these studies is given
in Yang et al. (2011). Similar to DW08, CH was set to be
0.004 for the C–TP and the E–TP stations, while it was
set to 0.00475 for the W-TP stations. The CMA 6-hourly
data are used as input for Eq. (1). DW08 assumed r 5
0.8 kg m23 in Eq. (1), but it was calculated from observed
air temperature and air pressure in this study. This difference does not fundamentally alter the climatology of
the sensible heat flux and its trend.
TABLE 1. Differences between the conventional estimation and the new estimation.
Heat source component
Conventional
New
SH
LH
RC
Empirical bulk equation
Estimated from CMA precipitation data
Estimated from ISCCP–FD data
Improved LSM simulation
Estimated from corrected CMA precipitation data
Estimated from selected satellite data and LSM simulation
1 MARCH 2011
YANG ET AL.
1529
FIG. 3. Comparisons of observed monthly-mean downward SW and LW with estimates from CMA data (‘‘Yang06 SW’’ and ‘‘CD99
LW’’) and two satellite estimates (GEWEX SRB and ISCCP–FD) at the four Tibetan experimental sites. Indices in each panel are the
mean bias error (MBE) (W m22), the root-mean-square error (RMSE) (W m22), and the coefficient of determination (R2).
2) NEW METHOD
Xu and Haginoya (2001) showed that it is possible to
produce reasonable SH values on the TP region by land
surface modeling. In the new estimate, SH was simulated
by the second-generation Simple Biosphere model (SiB2;
Sellers et al. 1996), which has been improved according to
experimental data analyses and TP surface conditions
(Yang et al. 2009a). The improvements include the implementation of an aerodynamic canopy model suitable
for short and sparse vegetation (Watanabe and Kondo
1990), a new surface flux parameterization scheme developed from the experimental data (Yang et al. 2008), a
high-accuracy soil water flow scheme (Ross 2003) with
additional considerations of soil freezing and thawing,
and a new parameterization of soil surface evaporation
resistance. The improved model performs better than the
original one in simulating surface temperature, sensible
heat flux, and net radiation (Yang et al. 2009a).
The simulation requires input of high-resolution data of
wind speed, air temperature, specific humidity, downward
radiation, and precipitation. They were downscaled from
CMA 6-hourly or daily data. A statistical method to
disaggregate wind speed and air temperature from CMA
6-hourly data to half-hourly data was developed in Yang
et al. (2009b, hereafter Yang09), and it was adopted
herein. Specific humidity was linearly interpolated from
CMA 6-hourly data. Daily precipitation was uniformly
distributed for the hours from late afternoon to early
morning, as suggested from high-resolution experimental
data (Yang et al. 2007). Downward SW was not observed
at most of the CMA stations and LW was totally not observed. In this study, SW was estimated from CMA sunshine data by the model in Yang et al. (2006b, hereafter
Yang06) and LW estimated by the scheme presented in
Crawford and Duchon (1999, hereafter CD99). This
CMA data-based estimate of downward SW and LW is
more accurate than the two satellite estimates (Fig. 3),
and thus it was used to drive the land surface modeling.
Relevant soil parameters and vegetation parameters
(classification and coverage) in the model were derived
from 18 3 18 International Satellite Land Surface Climatology Project Initiative II (ISLSCP II) soil data
(Global Soil Data Task Group 2000) and vegetation data
(Loveland et al. 2001). Leaf area index data were sourced
from Moderate Resolution Imaging Spectroradiometer
1530
JOURNAL OF CLIMATE
VOLUME 24
FIG. 4. Annual mean net radiation estimated by GEWEX SRB, ISCCP–FD, and the LSM.
The data are averaged over all CMA Tibetan stations. (a) TOA net radiation; (b) surface net
radiation.
(MODIS) 0.258 3 0.258 gridded 8-day leaf area index
products (Knyazikhin et al. 1999).
To achieve a sufficient modeling spinup for dry stations,
the simulated period started from 1952. Climatologic data
were used to drive the model until the period for which
observations are available.
b. Estimating latent heat release of condensation
Latent heat release due to water vapor condensation
can be estimated as follows:
LH 5 Lrw P,
(2)
where rw is the density of water (kg m23), P (mm s21) is
the surface precipitation, and L is the specific heat of
evaporation or sublimation (J kg21).
At CMA stations, precipitation was measured with the
Chinese standard precipitation gauge since the late 1950s.
The gauge is a cylinder of galvanized iron, 65 cm high and
20 cm in diameter. It was placed 0.7 m above the ground
without a windshield. According to Ye et al. (2004), a correction to CMA gauge-measured precipitation was made
to account for wind-induced error, wetting loss, and trace
precipitation. This correction was made once a day and
the corrected amount is about 15% of the gauged precipitation for the wetter C–TP and E–TP and about 30%
for the drier W–TP. As this correction depends on wind
speed and air temperature, which has changed in recent
decades, the trend in precipitation is slightly altered by the
correction (Ding et al. 2007). Nevertheless, this change is
not very significant and not concerned further.
In Eq. (2), the conventional method uses the CMA
gauge-measured precipitation, whereas the new method
uses the corrected precipitation.
c. Estimating radiative convergence
Radiative convergence was calculated as follows:
RC 5 Rntoa
Rnsfc ,
(3)
where Rntoa and Rnsfc are net radiation fluxes at TOA
and at the ground surface, respectively.
Both GEWEX SRB and ISCCP–FD provide the data
for Rntoa and Rnsfc. The land surface modeling also yields
Rnsfc. Zhao and Chen (2000a) presented an empirical
method to estimate Rntoa and Rnsfc. Ma et al. (2002a)
developed another remote sensing approach to estimate
Rnsfc, but it is not applicable to cloudy conditions. Attention needs to be paid to the use of these datasets. For instance, Fig. 3 shows that ISCCP–FD SW and LW data are
quite biased, though it was used in DW08. In Zhao and
Chen (2000a), the estimated wintertime Rnsfc is comparable to observations over the W–TP, but the summertime
one is about 20 W m22 higher than the observations (not
shown). Therefore, a dataset should be evaluated before
being used for the RC estimation.
Figure 4a shows the comparison between annual mean
Rntoa of GEWEX SRB and ISCCP-FD after averaged
over three subregions. It is seen that the two satellite estimates give comparable Rntoa trends. By contrast, the
two satellite estimates and the LSM estimate of Rnsfc are
very different; their discrepancies in annual-mean values
can be up to 30 W m22 (Fig. 4b). The significant negative
trend in GEWEX SRB Rnsfc observed in Fig. 4b is mainly
due to a strong negative trend in outgoing (reflected)
shortwave radiation (0.55 W m22 yr21), while the large
interannual variability in ISCCP–FD Rnsfc is due to the
large interannual variability of all radiation components,
except for incoming shortwave radiation (not shown).
The comparable Rntoa trends and the mixed Rnsfc trends
are not surprising, as it is more difficult to retrieve surface
radiation budget than to retrieve TOA radiation budget.
To select the least error-prone Rnsfc, Fig. 5 compares the
observed and the estimated monthly-mean Rnsfc at the
four Tibetan experimental sites. Clearly, the two satellite
1 MARCH 2011
YANG ET AL.
1531
FIG. 5. Comparisons of observed monthly-mean surface net radiation with satellite estimates (GEWEX SRB and ISCCP–FD) and
LSM at the four Tibetan experimental sites (SQH, Gerze, Naqu, and Anduo).
products overestimated Rnsfc, but the LSM-simulated Rnsfc
is closer to the observations.
To reduce the uncertainties, RC in the new estimation
was calculated with Rnsfc from the land surface modeling
and Rntoa averaged over the two satellite products. By
contrast, the conventional estimate was estimated from
ISCCP-FD radiation data, as adopted in DW08.
d. Estimating total heat source
The total heat source is the arithmetic sum of the above
three components as follows:
TH 5 SH 1 LH 1 RC.
(4)
4. Heat source climatology over the Tibetan
Plateau
Figure 6 compares the climatology of monthly-mean
heat source terms between the conventional estimate and
the new estimate in each subregion. Figure 7 compares the
new estimate with several previous estimates for the
entire plateau. Attention needs to paid to the periods and
averaging methods used for Fig. 7. The study periods are
different: before 1979 in Ye and Gao (1979), a single year
of 1979 in Yanai et al. (1992), 1961–95 in Zhao and Chen
(2001), and 1984–2006 in this study. The methods to average the heat source terms from the subregions to the
entire plateau are also different among these studies.
These differences may contaminate the comparison but
do not fundamentally alter our conclusions.
a. Sensible heat flux
Figure 6a shows that some basic features of the two
SH estimates in this study are similar to each other. The
maximum SH in each subregion roughly occurs in May
and the minimum SH occurs in December. SH decreases
from the W–TP to the E–TP, and this regional contrast is
associated with regional differences in rainfall and the
surface feature (Fig. 1). Nevertheless, the new SH’s
magnitude in Fig. 6a is generally less than the conventional one, and their differences are particularly large over
the W–TP. The higher accuracy of the new estimate can
be justified by the following analysis of the surface energy
budget over the W–TP at both annual and monthly scales.
According to experimental data at Gerze station on the
W–TP, the annual-mean surface net radiation averaged
over 2003/04 is 76 W m22. At annual scale, ground heat
flux (G0) is near 0; thus, the net radiation (Rnsfc) is primarily partitioned into sensible heat (SH) and latent heat
(LE). The W–TP is dry, with annual precipitation of
about 160 mm; thus, precipitated water is almost all lost
through evaporation because of very strong solar radiation. This leads to annual-mean LE of 13 W m22; thus,
annual-mean SH should be about 63 W m22. Based on
experimental data in 1998, Li et al. (2000) obtained SH
of 67 W m22 for this station. These SH values are close
to the new value (62 W m22) but less than the conventional one (76 W m22).
The maximum SH occurred in May and June. According to observations at Gerze station, the monthly-mean
Rnsfc averaged over 2003/04 is 110 W m22 in May and
129 W m22 in June. At the Shiquanhe experimental site,
the net radiation is even less. As the monthly mean of G0
and LE are about 10 W m22, it is generally not expected
that monthly-mean SH on the W–TP is larger than
100 W m22. Zhao and Chen (2001), according to the surface energy budget and observational data, gave a SH
value of 100 W m22 for May and June. This value is close
to our new estimate (98 W m22), while it is much less than
the conventional estimate for the two months (;140
W m22). In reality, most of previous studies, using either
reanalysis data or CMA data, greatly overestimated SH
in May and June (ranging from 120 to 220 W m22 on the
W-TP), as summarized in Table 2. On the relatively wet
C–TP and E–TP, maximum SH should be less than the
one on the W-TP; therefore, the values for the C–TP and
the E–TP in Table 2 are also too high to trust.
1532
JOURNAL OF CLIMATE
VOLUME 24
FIG. 6. Comparison of the climatology of the monthly-mean heat source terms between the conventional estimate and the new
estimate, averaged over all CMA stations in (left to right) W–TP, C–TP, and E–TP for the period 1984–2006: (a) SH, (b) LH, (c) RC,
and (d) TH.
Figure 7a compares SH between the new estimate and
the two estimates in the literature for the entire plateau. It
is observed that the estimate of Zhao and Chen (2001),
which is approximately constrained by the surface energy
budget, is close to the new estimate, whereas an earlier
effort by Ye and Gao (1979) greatly overestimated SH
because of a large value (0.008) of CH being used.
In brief, SH estimates in most previous studies were too
high for May and June, a period of major concern in
monsoon studies, because of their omission of the physical constraint by the surface energy budget.
b. Latent heat release of condensation
Figure 6b shows the monthly-mean LH climatology.
The two methods produced a clear seasonal variation of
LH. The maximum LH in each subregion occurs about
July–August, and the minimum LH occurs in December–
January. In contrast to SH, LH drastically increases from
1 MARCH 2011
1533
YANG ET AL.
FIG. 7. Comparison of the estimated climatology of the heat source terms averaged over the whole TP
between the new estimate in this study and estimates in previous studies: (a) SH, (b) LH, (c) RC, and (d) TH.
the W–TP to the E–TP, as the plateau monsoon migrates
from the southeast to the northwest. The correction to
CMA precipitation gives slightly higher LH values over
the W–TP and during the non monsoon seasons as well,
whereas the corrected rainfall become relatively large over
the E–TP and during the monsoon season. The relative
differences are almost more than 15% for all months and
all subregions. Therefore, this correction is not negligible.
Nevertheless, Fig. 7b shows that the magnitude of the new
estimate for the entire plateau is only slightly greater than
two previous estimates. This is probably related to the
different periods and spatial averaging methods concerned
in these studies, in addition to the precipitation correction
in the present study.
c. Radiative convergence
As shown in Fig. 6c, RC in all months has negative
values, which means that the so-called radiative convergence is actually radiative cooling. Surprisingly, the conventional and the new RC estimates are so different in
both seasonality and magnitude. The RC estimate by the
conventional method (using ISCCP–FD data) is generally
smaller than the new estimate, except during several
months over the W–TP. The former is less than the latter
by 25% over the C–TP and the E–TP at annual scale. In
an annual cycle, the former estimate nearly gives the
minimum radiative convergence (or maximum radiative cooling) during the monsoon season [June–August
(JJA)], when, by contrast, the new estimate gives the
maximum radiative convergence (or minimum radiative cooling).
Figure 7c compares RC of the new estimate and two
estimates in the literature for the entire plateau. Similar to
the above conventional estimate, the previous estimates
are generally significantly lower than the new one, particularly for summer or for the rainy season. These large
differences highlight the primary importance to conscientiously evaluate the radiation data used for RC estimation. It is observed in Fig. 7c that the result in Zhao and
Chen (2001) is close to the present one during winter.
However, their difference in RC is as high as 20 W m22
TABLE 2. Maximum monthly-mean of sensible heat flux estimated in previous studies.
Reference
Ye and Gao (1979)
Luo and Yanai (1984)
Wu and Zhang (1998)
Chen et al. (2003)
Region
Input
Period
Value (W m22)
W–TP
E–TP
W–TP
E–TP
Entire TP
TP
CMA data
CMA data
Reanalysis data
Reanalysis data
Reanalysis data
CMA data
June
May
May–early June
May–June
Since May
May–June
;220
;120
;170
;105
;(120–150)
;120
1534
JOURNAL OF CLIMATE
VOLUME 24
FIG. 8. The spatial distribution of the total heat source over the TP during the summer monsoon
(JJA) averaged over 1984–2006.
during summer (JJA), when the former overestimated
the surface net radiation.
d. Total heat source
Figure 6d shows there are fundamental differences in
the total heat source estimated between the two methods.
Regarding the seasonality, the conventional estimate
shows the maximum TH occurs in May (over the W–TP)
or June (over the C–TP and the E–TP), but the new estimate indicates that the maximum TH occurs in July for
all subregions. The maximum TH in the conventional
estimate is comparable to the new one over the W-TP,
but it is significantly less than the latter over C–TP and
E–TP. Figure 7d shows that the maximum value of the
plateau-averaged TH occurs in May or June in three
previous estimates, while it occurs in July in the new estimate. Zhao and Chen (2001) produced lower values of
TH climatology, because of its lower RC and lower LH
values. Of some interest is the fact that the climatology in
Ye and Gao (1979), the earliest estimate to our knowledge, is closest to the new one. This is a coincidence, because its large positive biases in SH greatly cancel out its
large negative biases in RC. The TH values in Yanai et al.
(1992) are close to those in Zhao and Chen (2001) during
JJA. Nevertheless, we note that there is an abrupt change
of heat source from May to June of 1979 in Yanai et al.
(1992), but this change is not seen in Zhao and Chen
(2001) nor in other climatologic studies.
Two fundamentally different spatial patterns of TH in
summer (JJA) were reported in the literature. One pattern is ‘‘TH decreases from the W–TP to the E–TP.’’ For
example, Ye and Gao (1979) reported the value of TH
climatology is 116 W m22 over the W–TP and 94 W m22
over the E–TP. In the summer of 1979, Yanai et al. (1992)
gave a TH value of 78 W m22 over the W–TP and
71 W m22 over the E–TP. The other pattern is ‘‘TH increases from the W–TP to the E–TP,’’ which was concluded by studies using the data from First Qinghai-Xizang
(Tibet) Plateau Meteorological Experiment (TIPMEX).
For example, Chen et al. (1985) gave 46 W m22 of TH
over the W–TP and 77 W m22 over the E–TP for the
summer of 1979. Zhao and Chen (2001) produced TH
climatology of 59 W m22 over the W–TP and 71 W m22
over the E–TP. In this study, we showed that the conventional method reproduced the former pattern (76
W m22 over the W–TP and 61 W m22 over the E–TP
during JJA), whereas the new estimate yielded the latter
pattern (82 W m22 over the W–TP and 111 W m22 over
the E–TP during JJA).
A more detailed spatial distribution of TH during JJA is
shown in Fig. 8. High TH (.100 W m22) is observed over
the southern and eastern TP, while low TH (,100 W m22)
over the northern and western TP. Another noteworthy
result is that the newly estimated summertime TH values
over the E–TP are higher than most previous estimates,
mainly because of the much smaller radiative cooling in
this study (Fig. 7c).
5. Heat source trend over the Tibetan Plateau
Figure 9 shows the linear trends in SH, RC, and TH
for all subregions and all seasons, together with 95%
1 MARCH 2011
YANG ET AL.
1535
FIG. 9. The estimated trends (W m22 decade21) in seasonal-mean and annual-mean heat source terms between the new estimate
(‘‘present’’) and the conventional method (‘‘conventional’’) for (left to right) the subregions and the entire plateau during 1984–2006.
confidence intervals of the estimated trends that were
calculated following Wigley (2006). The trends in LH are
not evident and thus omitted. The trends in the heat
source terms over the W–TP are small and most of them
exhibit large uncertainties and cannot pass the significance test ( p , 0.05). By contrast, the trends over the
C–TP and the E–TP are larger in magnitude and most of
them pass the significance test. In general, both the conventional method and the new method produced negative trends in SH, RC, and TH; however, the new method
gave weaker negative trends than the conventional one.
Regarding the trends in SH, the new ones are about half
to one-third of the conventional ones. The latter are believed overly estimated because of two causes. First, the
bulk heat transfer coefficient has strong diurnal variations
(small in the nighttime and large in the daytime) over the
TP; thus, it is theoretically questionable to take the coefficient as a constant value throughout a day, as assumed
in the conventional method and many previous studies.
Second, because of climate change, the atmospheric instability actually increased over the TP because the wind
speed steadily weakened and ground air temperature increased during the concerned period. Accordingly, the
bulk heat transfer (CH) may have increased with the enhancement of the atmospheric instability. This change
would suppress the decreasing trend in SH, but this effect is not taken into account in the conventional method
(Yang et al. 2011).
Based on Tibetan experimental data, Yang09 developed
a new algorithm to calculate sensible heat flux from CMA
data. This algorithm was applied by Yang et al. (2011) to
estimate the SH trend over the TP. Figure 10 shows that
the relative trends (% decade21) in the newly estimated
SH are comparable to those given by Yang et al. (2011) but
weaker than the conventional ones. As the new approach
used herein and Yang09 are independent methods, their
agreement in the SH trends indirectly supports each other.
Regarding the RC trends (Fig. 9b), they are dominated
by large negative trends in the TOA radiation budget
(Fig. 4a). The new RC trends are comparable to the conventional ones over the W–TP; however, the two become
quite different over the E–TP, where the new ones are
about half of the conventional ones. This difference is
mainly caused by the use of different data sources of the
surface net radiation. The former used LSM-simulated
net radiation, which does not show a strong trend. The
1536
JOURNAL OF CLIMATE
VOLUME 24
FIG. 10. The relative trends (% decade21) in seasonal-mean and annual-mean sensible heat flux for (left to right) the subregions and
the entire plateau during 1984–2006, estimated by the LSM in this study (‘‘present’’), the flux parameterization in Yang09, and the
conventional method.
conventional one used ISCCP–FD data, which yields
a significant positive trend on the E–TP (not shown).
The TH trends are shown in Fig. 9c. The new TH trends
are comparable to the conventional ones over the W–TP,
but they become nearly half of the latter over the C–TP
and the E–TP.
In summary, the conventional method may have notably overestimated the decreasing trends in the TP heat
source terms.
6. Upscaling from stations to regional scale
The above estimates are based on surface data and satellite data at CMA stations. Given the TP’s vast area, the
number of CMA stations is too small to be fully meaningful. Most of the stations were located on the C–TP and
the E–TP, and there are only three stations on the W–TP.
So, it is worth analyzing the spatial representativeness of
the above estimate. In reality, upscaling the heat source
terms from stations to the regional scale has not been
addressed in the literature. In this study, a simple spatial
upscaling with the aid of satellite products was given to
evaluate the representativeness of the estimated climatology of the heat source terms and their trends. The
spatial upscaling was applied to monthly-mean data, and
the regional mean defined below is the value averaged over
all grids with an elevation of more than 3000 m in each of
the three subregions.
The regional mean of RC was upscaled with the aid of
the GEWEX SRB data, as this satellite product has better
accuracy than ISCCP–FD for the TP region (Fig. 3), as
follows:
RCreg 5 RCsta
srb
RCsta
srb
,
(5)
where RCreg is the radiative convergence averaged over all
grids in a region, RCsta is the corresponding value averaged
over all stations in the region (i.e., the estimate in section
3), RCreg_srb is the RC value of GEWEX SRB averaged
over all grids in the region, and RCsta_srb is the corresponding value averaged over all stations in the region.
The spatial upscaling for LH is similar, as follows:
LHreg 5 LHsta
Preg
trmm
Psta
trmm
,
(6)
where LHreg is the latent heat release averaged over a
region, LHsta is the value averaged over all stations in
the region (i.e., the estimate in section 3), Psta_trmm is the
TRMM precipitation averaged over all stations in the
region, and Preg_trmm is the corresponding value averaged
over all grids in the region.
SH is assumed to be proportional to surface net radiation at each month and each region. Its spatial upscaling
is defined as follows:
a. Upscaling approach
A hypothesis for the upscaling is that the spatial variability represented by relevant satellite products is reliable
though their regional-mean values may be biased. This
hypothesis seems realistic, as radiation and precipitation
are mainly influenced by clouds, whose spatial distribution can be directly observed by satellites.
RCreg
SHreg 5 SHsta
Rnsfc
reg srb
Rnsfc
sta srb
,
(7)
where SHreg is the SH value averaged over a region, SHsta
is the LSM-simulated SH value averaged over all stations
in the region (i.e., the estimate in section 3), Rnsfc_reg_srb is
the surface net radiation from GEWEX SRB averaged
1 MARCH 2011
1537
YANG ET AL.
TABLE 3. Monthly-mean heat source terms (W m22) estimated in this study after spatial upscaling with the aid of GEWEX SRB radiation
data and TRMM precipitation data, averaged over the period of 1998–2004.
Month
1
2
3
4
5
6
7
8
9
10
11
12
Annual
81
52
34
56
68
43
28
46
67
36
22
42
44
31
16
30
21
23
12
19
11
17
9
12
54
44
28
42
33
84
122
80
31
85
118
78
13
44
88
48
2
14
44
20
3
2
9
5
2
2
3
2
12
30
54
32
236
245
250
244
251
252
259
254
270
275
276
274
269
287
287
281
280
292
292
288
286
286
287
286
260
264
266
263
78
91
107
92
47
76
87
70
10
6
33
16
223
242
228
231
256
266
271
264
273
268
275
272
6
10
17
11
SH
W–TP
C–TP
E–TP
Avg
20
21
13
18
35
37
25
32
56
58
40
51
73
73
52
66
87
77
50
71
87
66
39
64
W–TP
C–TP
E–TP
Avg
7
4
7
6
8
4
9
7
8
5
20
11
8
12
37
19
9
32
75
39
19
66
118
68
W–TP
C–TP
E–TP
Avg
288
281
282
284
276
268
268
271
261
253
255
256
242
244
243
243
229
241
242
237
236
247
249
244
W–TP
C–TP
E–TP
Avg
261
257
262
260
233
227
235
232
4
10
6
7
39
40
45
41
68
68
84
73
69
85
108
87
LH
RC
TH
over all grids in the region, and Rnsfc_sta_srb is the corresponding value averaged over all stations on the region.
b. Upscaling effects on the heat source estimate
The GEWEX SRB radiation data and the TRMM
precipitation data required in the upscaling are available
for the period 1984–2004 and the period since 1998, respectively. So, the results for the common period (1998–
2004) are presented below.
The heat source estimate after upscaling is shown in
Table 3. Figure 11 compares the climatology of monthlymean heat source terms before and after upscaling. The
values after upscaling are generally less than before upscaling. This is related to the location of these stations. As
seen in Fig. 1, there are more stations on the southern TP
than on the northern TP; therefore, the estimated values
before upscaling better represent the case in the southern
TP. Because more precipitation and radiation take place
on the southern TP than on the northern TP, the heat
source is stronger over the southern TP, as can be seen
in Fig. 8. As a result, the TH values after upscaling are
mostly less than the ones before upscaling and should be
more representative at the regional scale. For instance,
their differences can be up to 20 W m22 in a region with
sparse station distribution (e.g., the W–TP). This experience suggests that the uncertainty in the estimated heat
source due to this scale effect may be comparable to that
due to the methods for heat source calculation in section 3
and should be addressed. In addition, it is worth noting
that the TH values decrease more over the W–TP, where
CMA stations are sparse, than over the E-TP, where
CMA station distribution is much denser. Therefore, the
regional contrast of TH between the W–TP and the E–
TP is enhanced by the spatial upscaling.
On the other hand, the effects of the spatial upscaling
seem rather insignificant on the trend estimation. As
shown in Fig. 12, the interannual variability of JJA-mean
and annual-mean TH values before upscaling is very similar to the counterpart after upscaling. This suggests that
the trends in the heat source terms are generally not sensitive to the number of stations. In other words, the trends
can be more reliably quantified from a limited number of
stations than the climatology of the heat source.
7. Concluding remarks
Atmospheric heating over the TP plays a profound role
in the Asian monsoon system. Although many studies in
the literature have focused on estimating the heating components and the total heat source, most of them lack substantial supports of observational data; therefore, existing
estimates give a mixed picture, and some of them seem
suspicious. In this study, recent advances in field data analyses, land model development, and satellite estimates of
radiation and precipitation are utilized to produce a stateof-the-art estimate of the TP atmospheric heat source.
By comparing with the new estimate, we uncovered
major biases in previous estimates. First, most previous
studies overestimated the sensible heat flux in May and
June. Some estimates are even higher than the observed
surface net radiation, which violates the constraint of the
surface energy balance at a monthly or annual scale. An
1538
JOURNAL OF CLIMATE
VOLUME 24
FIG. 12. Interannual variability of summer (‘‘JJA’’) and annual
(‘‘ANN’’) mean total heat source estimated at CMA stations
(‘‘sta’’) and estimated at the regional scale (‘‘reg’’) on (top to
bottom) the three subregions, with the aid of GEWEX SRB radiation data and TRMM precipitation data for the period 1998–2004.
FIG. 11. The annual cycle of the heat source estimates before and
after upscaling from stations to the regional scale during 1998–2004:
(a) SH, (b) LH, (c) RC, and (d) TH. The case for C–TP is not shown.
exception is the sensible heat flux presented in Zhao and
Chen (2001), which is similar to the present one, as both
estimates are constrained by the surface energy budget.
Second, previous studies also greatly underestimated
radiative convergence (or overestimated the radiative
cooling) during the summer monsoon (or rainy) period.
Nevertheless, the wintertime radiative convergence presented in Zhao and Chen (2001) is close to the present
one. Third, the latent heat release was estimated based on
CMA gauge-measured precipitation data in both the present study and many previous ones; therefore, their results
1 MARCH 2011
1539
YANG ET AL.
are similar to each other. The present study considered a
correction to account for the precipitation measurement
loss and thus gives slightly higher latent heat release values
than previous estimates.
Accordingly, the seasonality and regional pattern of the
total heat source in the new estimate are different from
previous studies. The total heat source in each subregion
or the plateau reaches maximum during the mature stage
of the monsoon season (July), whereas previous studies
gave the maximum in the premonsoon season (May) or
the beginning of the monsoon season (June), because
of greatly overestimated radiative cooling in the monsoon
season. The new estimate shows the maximum heat source
occurred over the E–TP rather than over the W–TP,
whereas most previous studies show an opposite regional
pattern, because of the overestimated sensible heat flux
on the W–TP before the onset of the monsoon. A spatial
upscaling with the aid of satellite data further enhances
the spatial pattern of the heat source distribution given by
the new estimate.
This study also investigated the trends in the heat
source terms. A recent estimate by DW08 showed strong
negative trends in sensible heat and radiative convergence as well as the total heat source. This study confirms
the negative trends, but the magnitudes in their work
seem greatly overestimated. Specifically, their sensible
heat trend is 2–3 times the new one averaged over the
plateau, because atmospheric stability has been weakening in recent decades but their estimate omitted its effect
on the heat exchange (Yang et al. 2011). The trend in
radiative convergence given by DW08 is similar to the
new one over the W–TP, but it becomes twice the new
one over the C–TP and the E–TP, because of biases of the
surface net radiation used in their study. As a result, the
trend in the total heat source given in their study is about
212 W m22 decade21 for the entire plateau, while the
value given herein is 27 W m22 decade21.
We believe that this observation-supported study may
reduce uncertainties in the heat source estimate over the
TP, which may contribute to relevant model evaluations,
processes analyses, and climate change studies. Nevertheless, further experiments and studies are encouraged
to reach consensus on the estimated heating components
and the total heat source. Particularly, we recommend
more investigations on the radiative forcing, as its accuracy significantly affects the heat source estimation.
Acknowledgments. This work was supported by the
‘‘100 Talents Program’’ of the Chinese Academy of Sciences and by grants from National Natural Science Foundation of China (Grants 40810059006 and 40875009). In
situ data used were obtained under the GAME-Tibet and
follow-on projects. The authors thank the Climate Data
Center at the CMA National Meteorological Information
Center for providing the station data used in this study.
Xiaofeng Guo gratefully acknowledges the support of the
K. C. Wong Education Foundation, Hong Kong. We appreciate the reviewers’ valuable comments and suggestions, which helped improve the quality of this paper.
REFERENCES
Chen, L., E. R. Reiter, and Z. Feng, 1985: The atmospheric heat
source over the Tibetan Plateau: May–August 1979. Mon.
Wea. Rev., 113, 1771–1790.
——, F. Schmidt, and W. Li, 2003: Characteristics of the atmospheric heat source and moisture sinks over the QinghaiTibetan Plateau during the second TIPEX of summer 1998
and their impact on surrounding monsoon. Meteor. Atmos.
Phys., 83, 1–18.
Chen, Y., K. Yang, D. Zhou, J. Qin, and X. Guo, 2010: Improving
Noah land surface model in arid regions with an appropriate
parameterization of the thermal roughness length. J. Hydrometeor., 11, 995–1006.
Crawford, T. M., and C. E. Duchon, 1999: An improved parameterization for estimating effective atmospheric emissivity for use
in calculating daytime downwelling longwave radiation. J. Appl.
Meteor., 38, 474–480.
Ding, Y., D. Yang, B. Ye, and N. Wang, 2007: Effects of bias correction on precipitation trend over China. J. Geophys. Res.,
112, D13116, doi:10.1029/2006JD007938.
Duan, A., and G. Wu, 2008: Weakening trend in the atmospheric
heat source over the Tibetan Plateau during recent decades.
Part I: Observations. J. Climate, 21, 3149–3164.
Flohn, H., 1957: Large-scale aspects of the ‘‘summer monsoon’’ in
South and East Asia. J. Meteor. Soc. Japan, 35, 180–186.
Fu, R., and Coauthors, 2006: Short circuit of water vapor and
polluted air to the global stratosphere by convective transport
over the Tibetan Plateau. Proc. Natl. Acad. Sci. USA, 103,
5664–5669.
Global Soil Data Task Group, 2000: Global gridded surfaces of
selected soil characteristics (IGBP-DIS). ORNL DAAC, Oak
Ridge, TN. [Available online at http://daac.ornl.gov/SOILS/
guides/igbp-surfaces.html.]
He, H., J. W. McGinnis, Z. Song, and M. Yanai, 1987: Onset of the
Asian summer monsoon in 1979 and the effect of the Tibetan
Plateau. Mon. Wea. Rev., 115, 1966–1994.
Huffman, G. J., and Coauthors, 2007: The TRMM Multisatellite Precipitation Analysis (TMPA): Quasi-global, multiyear, combinedsensor precipitation estimates at fine scales. J. Hydrometeor,
8, 38–55.
Knyazikhin, Y., and Coauthors, 1999: MODIS leaf area index (LAI)
and fraction of photosynthetically active radiation absorbed by
vegetation (FPAR) product, (MOD15). Algorithm Theoretical
Basis Document version 4.0, 130 pp. [Available online at http://
modis.gsfc.nasa.gov/data/atbd/atbd_mod15.pdf.]
Koike, T., T. Yasunari, J. Wang, and T. Yao, 1999: GAME-Tibet
IOP summary report. Proceedings of the First International
Workshop on GAME-Tibet, A. Numaguti, L. Liu and L. Tian,
Eds., Chinese Academy of Sciences and Japan National Committee for GAME, 1–2.
Kuwagata, T., A. Numaguti, and N. Endo, 2001: Diurnal variation of
water vapor over the central Tibetan Plateau during summer.
J. Meteor. Soc. Japan, 79, 401–418.
1540
JOURNAL OF CLIMATE
Li, G., T. Duan, and Y. Gong, 2000: The bulk transfer coefficients
and surface fluxes on the western Tibetan Plateau. Chin. Sci.
Bull., 45, 1221–1226.
Li, W., and L. Chen, 2003: Characteristics of the seasonal variation
of the surface total heating over the Tibetan Plateau and its
surrounding area in summer 1998 and its relationship with the
convection over the subtropical area of the western Pacific.
Adv. Atmos. Sci., 20, 343–348.
Liu, Y., B. Hoskins, and M. Blackburn, 2007: Impact of Tibetan
orography and heating on the summer flow over Asia. J. Meteor.
Soc. Japan, 85, 1–19.
Loveland, T. R., B. C. Reed, J. F. Brown, D. O. Ohlen, Z. Zhu,
L. Yang, and J. W. Merchant, 2001: Development of a global
land cover characteristics database and IGBP DISCover from
1 km AVHRR data. Int. J. Remote Sens., 21, 1303–1330.
Luo, H., and M. Yanai, 1984: The large-scale circulation and heat
sources over the Tibetan and surrounding areas during the
early summer of 1979. Part II: Heat and moisture budgets.
Mon. Wea. Rev., 112, 966–989.
Ma, Y., Z. Su, Z. Li, T. Koike, and M. Menenti, 2002a: Determination
of regional net radiation and soil heat flux densities over a heterogeneous landscape of the Tibetan Plateau. Hydrol. Processes, 16, 2963–2971.
——, O. Tsukamoto, J. Wang, and H. Ishikawa, 2002b: Analysis of
aerodynamic and thermodynamic parameters on the grassy
marshland surface of Tibetan Plateau. Prog. Nat. Sci., 12, 36–40.
Oku, Y., H. Ishikawa, S. Haginoya, and Y. Ma, 2006: Recent trends
in land surface temperature on the Tibetan Plateau. J. Climate,
19, 2995–3003.
Qian, Y., Y. Zhang, Y. Huang, Y. Huang, and Y. Yao, 2004: The
effects of the thermal anomalies over the Tibetan Plateau and
its vicinities on climate variability in China. Adv. Atmos. Sci.,
21, 369–381.
Qin, J., K. Yang, S. Liang, and X. Guo, 2009: The altitudinal dependence of recent rapid warming over the Tibetan Plateau.
Climatic Change, 97, 321–327.
Ross, P. J., 2003: Modeling soil water and solute transport—Fast,
simplified numerical solutions. Agron. J., 95, 1352–1361.
Sato, T., and F. Kimura, 2007: How does the Tibetan Plateau affect
the transition of Indian monsoon rainfall? Mon. Wea. Rev.,
135, 2006–2015.
Sellers, P. J., and Coauthors, 1996: A revised land surface parameterization (SiB2) for atmospheric GCMs. Part I: Model formulation. J. Climate, 9, 676–705.
Stackhouse, P. W., S. K. Gupta, S. J. Cox, J. C. Mikovitz, T. Zhang,
and M. Chiacchio, 2004: 12-year surface radiation budget data
set. GEWEX News, No. 14, International GEWEX Project
Office, Silver Spring, MD, 10–12.
Tamura, T., K. Taniguchi, and T. Koike, 2010: Mechanism of upper
tropospheric warming around the Tibetan Plateau at the onset
phase of the Asian summer monsoon. J. Geophys. Res., 115,
D02106, doi:10.1029/2008JD011678.
Tanaka, K., I. Tamagawa, H. Ishikawa, Y. Ma, and Z. Hu, 2003:
Surface energy and closure of the eastern Tibetan Plateau
during the GAME-Tibet IOP 1998. J. Hydrol., 283, 169–183.
Taniguchi, K., and T. Koike, 2007: Increasing atmospheric temperature in the upper troposphere and cumulus convection
over the eastern part of the Tibetan Plateau in the pre-monsoon
season of 2004. J. Meteor. Soc. Japan, 85A, 271–294.
Ueda, H., H. Kamahori, and N. Yamazaki, 2003: Seasonal contrasting features of heat and moisture budgets between the
eastern and western Tibetan Plateau during the GAME IOP.
J. Climate, 16, 2309–2324.
VOLUME 24
Ueno, K., H. Fujii, H. Yamada, and L. Liu, 2001: Weak and frequent monsoon precipitation over the Tibetan Plateau.
J. Meteor. Soc. Japan, 79, 419–434.
van der Velde, R., Z. Su, M. Ek, M. Rodell, and Y. Ma, 2009: Influence of thermodynamic soil and vegetation parameterizations on the simulation of soil temperature states and surface
fluxes by the Noah LSM over a Tibetan Plateau site. Hydrol.
Earth Syst. Sci., 13, 759–777.
Wang, B., Q. Bao, B. Hoskins, G. Wu, and Y. Liu, 2008: Tibetan
Plateau warming and precipitation changes in East Asia. Geophys. Res. Lett., 35, L14702, doi:10.1029/2008GL034330.
Watanabe, T., and J. Kondo, 1990: The influence of canopy structure and density upon the mixing length within and above
vegetation. J. Meteor. Soc. Japan, 68, 227–235.
Wigley, T. M. L., 2006: Appendix A: Statistical issues regarding
trends. Temperature Trends in the Lower Atmosphere: Steps for
Understanding and Reconciling Differences, T. R. Karl et al.,
Eds., Synthesis and Assessment Product 1.1., Climate Change
Science Program, 129–139.
Wu, G., and Y. Zhang, 1998: Tibetan Plateau forcing and timing of
the monsoon onset over South Asia and South China Sea.
Mon. Wea. Rev., 126, 913–927.
——, and Coauthors, 2007: The influence of mechanical and thermal
forcing by the Tibetan Plateau on Asian climate. J. Hydrometeor., 8, 770–789.
Xu, J., and S. Haginoya, 2001: An estimation of heat and water balances in the Tibetan Plateau. J. Meteor. Soc. Japan, 79, 485–504.
Xu, M., C.-P. Chang, C. Fu, Y. Qi, A. Robock, D. Robinson, and
H.-M. Zhang, 2006: Steady decline of East Asian monsoon
winds, 1969–2000: Evidence from direct ground measurements of wind speed. J. Geophys. Res., 111, D24111, doi:
10.1029/2006JD007337.
Yanai, M., C. Li, and Z. Song, 1992: Seasonal heating of the Tibetan
Plateau and its effects on the evolution of the Asian summer
monsoon. J. Meteor. Soc. Japan, 70, 319–351.
Yang, K., T. Koike, H. Fujii, T. Tamura, X. Xu, L. Bian, and
M. Zhou, 2004: The daytime evolution of the atmospheric
boundary layer and convection over the Tibetan Plateau:
Observations and simulations. J. Meteor. Soc. Japan, 82,
1777–1792.
——, ——, P. Stackhouse, C. Mikovitz, and S. J. Cox, 2006a: An
assessment of satellite surface radiation products for highlands
with Tibet instrumental data. Geophys. Res. Lett., 33, L22403,
doi:10.1029/2006GL027640.
——, ——, and B. Ye, 2006b: Improving estimation of hourly,
daily, and monthly solar radiation by importing global data
sets. Agric. For. Meteor., 137, 43–55.
——, and Coauthors, 2008: Turbulent flux transfer over bare-soil
surfaces: Characteristics and parameterization. J. Appl. Meteor. Climatol., 47, 276–290.
——, Y. Chen, and J. Qin, 2009a: Some practical notes on the land
surface modeling in the Tibetan Plateau. Hydrol. Earth Syst.
Sci., 13, 687–701.
——, J. Qin, X. Guo, D. Zhou, and Y. Ma, 2009b: Method development for estimating sensible heat flux over the Tibetan
Plateau from CMA data. J. Appl. Meteor. Climatol., 48, 2474–
2486.
——, X. Guo, and B. Wu, 2011: Recent trends in surface sensible
heat flux on the Tibetan Plateau. Sci. China Earth Sci., 54, 19–
28. doi:10.1007/s11430-010-4036-6.
Yang, M., T. Yao, X. Gou, H. Wang, and A. Lu, 2007: Comparison
analysis of the summer monsoon precipitation between northern
and southern slopes of Tanggula Mountains, Qinghai-Xizang
1 MARCH 2011
YANG ET AL.
(Tibetan) Plateau: A case study in summer 1998. Hydrol.
Processes, 21, 1841–1847.
Ye, B., D. Yang, Y. Ding, T. Han, and T. Koike, 2004: A bias-corrected precipitation climatology for China. J. Hydrometeor., 5,
1147–1160.
Ye, D., and Y. Gao, 1979: The Meteorology of the Qinghai-Xizang
(Tibet) Plateau (in Chinese). Science Press, 278 pp.
You, Q., S. Kang, N. Pepin, W. Flügel, Y. Yan, H. Behrawan, and
J. Huang, 2010: Relationship between temperature trend
magnitude, elevation and mean temperature in the Tibetan
Plateau from homogenized surface stations and reanalysis
data. Global Planet. Change, 71, 124–133.
Zhang, Y., W. B. Rossow, A. A. Lacis, V. Oinas, and
M. I. Mishchenko, 2004: Calculation of radiative fluxes from
the surface to top of atmosphere based on ISCCP and other
global data sets: Refinements of the radiative transfer model
1541
and the input data. J. Geophys. Res., 109, D19105, doi:
10.1029/2003JD004457.
Zhao, P., and L. Chen, 2000a: Calculation of solar albedo and radiation equilibrium over the Qinghai-Xizang Plateau and
analysis of their climate characteristics. Adv. Atmos. Sci., 17,
140–156.
——, and ——, 2000b: Study on climatic features of surface turbulent heat exchange coefficients and surface thermal sources
over the Qinghai-Xizang (Tibetan) Plateau. Acta Meteor. Sin.,
14, 13–29.
——, and ——, 2001: Climatic features of atmospheric heat source/
sink over the Qinghai-Xizang Plateau in 35 years and its relation to rainfall in China. Sci. China, 44, 858–864.
Zhou, X., P. Zhao, J. Chen, L. Chen, and W. Li, 2009: Impacts of
thermodynamic processes over the Tibetan Plateau on the
Northern Hemispheric climate. Sci. China, 52D, 1679–1693.