Seasonal and diurnal variations in moisture, heat, and CO2 fluxes over grassland in the tropical monsoon region of southern China

Authors


Abstract

[1] In order to examine energy partitioning and CO2 exchange over grassland in the tropical monsoon environment of southern China, fluxes of moisture, heat, and CO2 in the near-surface layer during the period from May 2004 to July 2005 were calculated using the eddy covariance method. The study site was homogenous and approximately 400 × 300 m in size. Comparison of meteorological parameters measured at this site with historical data collected at a nearby routine observatory since 1960 shows that the measurement period is representative of the climate for this area. Seasonal and diurnal variations in radiation components, energy components, and CO2 fluxes are examined. Results show that all four radiation components changed seasonally, resulting in a seasonal variation in net radiation. The radiation components also changed diurnally, except that the downward longwave radiation showed almost no diurnal variation during the period from December to April. Summer surface albedo was higher than winter surface albedo because during winter the fraction of bare soil increases and the albedo of soil is lower than that of grass in this area. The seasonal variations in both latent heat and CO2 fluxes were stronger than those of sensible heat and soil heat fluxes. This implies that both latent heat and CO2 fluxes may be more significant climate signals than sensible heat and soil fluxes. Latent (sensible) heat flux was the main consumer of available energy in summer (winter). The energy imbalance problem was encountered and the causes are analyzed.

1. Introduction

[2] The relatively recent increase in atmospheric carbon dioxide (hereinafter, referred to as CO2) concentration has profound implications for the planet’s climate (for example, IPCC report) as well as on photosynthesis and the structure and function of plant communities. Vegetation, in turn, plays a crucial role in the global carbon balance [Woodward et al., 1998; Mielnick et al., 2001]. Surface fluxes of momentum, heat, and moisture determine to a large extent the steady state of the atmosphere [Beljaars and Holtslag, 1991]. Climate simulations are especially sensitive to the seasonal and diurnal variations in surface partitioning of available energy into sensible and latent heat fluxes [e.g., Rowntree, 1991; Dickinson et al., 1991]. In order to evaluate the long-term energy balance and evapotranspiration, a number of experimental studies have been carried out on various terrestrial land surfaces such as forest, grasslands, and paddy fields throughout the world during the past decade [e.g., Baldocchi and Vogel, 1997; Toda et al., 2002]. Previous work has reported on measurements of the seasonal and/or diurnal variations of heat, water vapor, and CO2 exchanges over different land surfaces [e.g., Hartog and Neumann, 1994; Delire and Gerard, 1995; Betts and Ball, 1995; Campbell et al., 2001; Vourlitis et al., 2001; Merquiol et al., 2002; Gao et al., 2003; Steven et al., 2005].

[3] Land surface processes in the monsoonal environment play an important role in the global energy balance and in the budget of greenhouse gases such as CO2 [Intergovernmental Panel on Climate Change, 1995]. The CO2, heat, and water vapor exchanges between the land and atmosphere in the tropical monsoon environment have received considerable attention recently because they not only play major roles in local convection, but also significantly influence the onset or decline of the monsoon [Toda et al., 2002; Xue et al., 2004]. Long-term measurements of energy exchanges and of CO2 flux in the tropical monsoon environment are therefore complementary and desirable. Several field experiments have been conducted over the land surface in recent decades aimed at studying the interaction between the lower atmosphere and underlying surface in the tropical monsoon environment. The data obtained during these long-term field experiments have enabled scientists to investigate the seasonal and diurnal cycles of surface fluxes. For example, Toda et al. [2002] estimated the surface energy flux balance and total evapotranspiration using the eddy covariance technique over a tropical monsoon environment within the framework of the global energy and Water cycle Experiment (GEWEX) Asian Monsoon Experiment (GAME). Barros et al. [2005] described the diurnal cycle of land-atmosphere interactions in Sonora, Mexico during the North American Monsoon Experiment/Soil Moisture Experiments in 2004. However, much of the data obtained so far is still not sufficiently detailed to examine the seasonal and diurnal variations of heat fluxes and CO2 exchange over the natural land surfaces in the tropical monsoon environment.

[4] To improve the current understanding of energy partitioning and CO2 exchange over the land surface in the tropical monsoon environment and to find which surface energy components show the strongest climate signals, we conducted a micrometeorological experiment over a natural grassland surface in the tropical monsoon environment in southern China from May 2004. The main objective of the present work is therefore to quantify the seasonal and diurnal variations in energy and CO2 exchanges over the above mentioned surface using eddy covariance techniques.

2. Materials and Methods

[5] The cover ratio of grassland is only about 8% in the tropical monsoon region of southern China, which is much less than that of forest. To capture the strongest surface climate signals, we originally hoped to conduct this experiment over a land surface where the human activity has been slight in the tropical Monsoon region of southern China. In this way, the forest and the grassland were optional and the forest is more important than grassland in this region. Because of the limited financial budget and because China flux investigators already set a tower for forest flux measurements in Dinghu Mountain area in the tropical monsoon region of southern China, we could not conduct the experiment in forest. We finally decided to conduct the experiment over the grassland near a meteorological station where routine meteorological measurements have been made for more than 40 years. We expected to examine whether our measurement results were representative of the climate for this area, by comparing our measured data to these historical data.

2.1. Site

[6] Measurements have been conducted at a grassland site (22.43°N, 113.23°E, 12.5 m above sea level) near Panyu county in the tropical monsoon region in southern China since 11 May 2004. This grassland is the result of people clearing forestland for cultivation. Previous farmers fell a tract of forest, burned the dead trees, and planted crops in the ashes. The field has been abandoned in recent 10 years. The soil at the site was predominantly porous terracotta soil with rapid drainage of water. It has only a thin layer of humus (the organic portion of the soil created by partial decomposition of plant or animal matters) which provides vegetation with nutrients. The site was flat, homogeneous, and approximately 300 × 400 m. Unfortunately, the leaf area index (LAI) has not been measured. Figure 1 shows the measurement mast, which was taken on 20 July 2005.

Figure 1.

Photo of the setup at the measurement site.

2.2. Micrometeorological Measurements

2.2.1. Fast Response Measurements

[7] The eddy covariance instruments were mounted on a mast (Figure 1). A three-dimensional sonic anemometer (CSAT3, Campbell Scientific Inc.) was used to measure the means and standard deviations of wind velocity components (i.e., u, v, and w) and air temperature (T) and a LI-7500 (LiCor, USA) was used to measure the mean and standard deviations of water vapor density and CO2. The sensitivity of the gas analyzer to CO2 was calibrated before the experiment using three levels of standard gases (between 300 and 400 ppmv CO2 in N2). These sensors were installed at 3.9 m above the ground surface. All signals for the sensors were recorded at a sampling rate of 10 Hz and were averaged over 30-min periods. Appropriate corrections were made for nonzero mean vertical velocity. Following Moore [1986], we corrected eddy covariance values for the effects of path length averaging of the sonic anemometer and the gas analyzer, and for the spatial separation of sensors. Corrections were made for density fluctuations in calculating the fluxes of water vapor and CO2 [Webb et al., 1980].

[8] We eliminated noise and various kinds of interference from 30-min measurements of turbulence by using a criterion of X(t) < (equation image − 4σ) or X(t) > (equation image + 4σ), where X(t) denotes the measurement (i.e., wind speed components, temperature), equation image is the mean over the interval and σ the standard deviation. Data during and after rain events was removed because the sonic anemometer would be in error in these cases. Any gaps were filled by linear interpolation.

2.2.2. Slow Response Measurements

[9] Other supporting data were collected during the experiment. Soil heat flux was measured by embedding heat flux plates at a depth of 0.01 m. Soil temperature was measured at six depths (0.01, 0.05, 0.1, 0.2, 0.4, and 0.7 m) in the soil. Upward and downward shortwave and longwave radiation components were measured with radiometers (model PSP, Eppley Inc., Newport, RI) mounted at a height of 1.5 m. The data were recorded by a Datataker (Model DT-600, DEC Inc., Australia) with a PCMCIA memory card. The data were sampled each minute and averages recorded every 10 min.

[10] The set of observational data includes the meteorological quantities: horizontal wind speed, air temperature, specific humidity, air pressure, and precipitation. Figure 2 shows the time series of (1) weekly mean wind vector (WV in m s−1), (2) weekly mean air temperature (Tair in K), (3) weekly mean specific humidity (q in g kg−1), (4) daily mean air pressure (P in hpa), and (5) daily precipitation (Prec. in mm day−1) obtained since May 2004. It is obvious that all of these meteorological quantities undergo a marked seasonal cycle. In the wet (monsoon) season, high temperature and high humidity were coincident with low air pressure and precipitation events were frequent. The reverse occurred during the dry season.

Figure 2.

Meteorological data collected at the grassland site during the period from May 2004 to June 2005 and climatic data collected at a nearby observatory during the period from January 1960 to April 2004.

[11] To examine whether our measurement results were representative of the climate for this area, the corresponding historical data obtained at a meteorological station for the period from January 1960 to April 2004, are also plotted (in solid line) in Figure 2. The meteorological station is 100 m away from the field measurement site. It can be seen that our measurements are representative of the meteorological situation over the past 45 years and are likely therefore to be representative of the climate of the region.

[12] The plots in Figures 2a–2c are derived from the fast response measurements. The composition method is applied for estimation of daily variation for each week and then a weekly average is calculated. The gaps in Figure 2 were caused by power cuts at the site. The plots in Figures 2d–2e are derived from the slow response instruments. In Figure 2, the historical data were collected at 2:00, 8:00, 14:00, and 20:00 (local time) at a routine observatory, which is at the distance of 100 m from our mast. In Figure 2d, the air pressure was a daily average and in Figure 2e precipitation is a daily sum for both our data and the historical data.

[13] Our site is located in a typical monsoon climate zone. During the winder, cold dry air frequently came from the north, influencing significantly the area, so the predominant wind direction was north-east during the period from the end of August 2004 to the beginning of March 2005. The wind direction gradually changed after March because the east wind component gradually increased and the north wind component gradually decreased. Because of the influence of the monsoon climate, the east-south wind and south wind were maintained from April to the end of September. Although the weekly mean wind speed was less than 2.2 m s−1 (shown in Figure 2a), the maximum of the hourly mean wind speed reached 5 ms−1. Figure 2a shows that the site was under the low wind regime.

[14] The seasonal variation of air temperature (Tair) was dramatic. Monthly mean air temperature reached a maximum (307.5 K) in August, and the lowest air temperature (282.5 K) occurred at the end of December. The difference between the highest air temperature and lowest air temperature was 25 K for the whole experimental period and the annual mean air temperature was 296.4 K. Compared to historical data during the period from January 1960 to April 2004, air temperature obtained during our measurements was somewhat high in summer. The 100-m distance between the two sites, not strictly consistent types of instruments used and not quite consistent measurement heights, do not warrant firmer conclusions. The historical air temperature was measured at a height of 1.5 m.

[15] Similar seasonal variation occurred in air specific humidity (q), with the correlation coefficient between q and Tair reaching 0.88. Air specific humidity varied also in response to variations in precipitation (Figures 2c and 2e). On average, air specific humidity was higher than historical values for the period from January 1960 to April 2004.

[16] Air pressure varied seasonally but in reverse phase to air temperature and air specific humidity. During the winter and spring the subtropical high-pressure zone weakened and moved eastward. After mid-October, the ridgeline even reached 15°N and the west wind trough in high altitude areas was so active as to bring northern cold air southward. Southern China therefore was either dominated by very low air temperature and a high-pressure ridge or by a stable high ridge. This explains why the air pressure maintained high values at the ground during winter and spring. In summer and autumn the subtropical high pressure jumped northward to 20°N, northern cold air weakened, and southern China was influenced by the front of the low trough. Typhoons occurred frequently in summer. This explains why the air pressure maintained low values at the ground during the summer. The air pressure measured during our experimental period was consistent with that collected during the period from January 1960 to April 2004.

[17] Almost all precipitation occurred during the period from May to September 2004 and the period from March to May 2005. The maximum daily precipitation occurred on 28 August, caused by Typhoon Rananim.

2.3. Theoretical Considerations

[18] The surface energy balance over the grass canopy can be approximated by:

equation image

where Rn is the net radiation, LE and H are the latent heat and sensible heat fluxes, respectively. G1 is the soil heat measured at a certain depth in soil and Re is the residual energy involved in various processes, such as photosynthesis, respiration, and heat storage in both grass and soil [Burba et al., 1999]. We determine Re by using the formula: Re = Rn − (H + LE + G1). Rn was measured using slow response instruments (described above). Eddy fluxes of sensible heat and latent heat were calculated as [e.g., Kaimal and Finnigan, 1994]:

equation image
equation image

where equation image, Cp and L are the density of air (kg m−3), the specific heat of air (J kg−1 K−1), and the latent heat of vaporization (J kg−1), respectively. w′, T′, and q′ are the fluctuations in the vertical wind component (m s−1), air temperature (K) and specific humidity, respectively.

[19] Since the 1980s, the development of fast response CO2 analyzers has enabled us to directly measure CO2 fluxes over rice canopies using eddy covariance methods. Ohtaki and Matsui [1982], and Ohtaki [1984] defined the following equation to reliably estimate the CO2 flux:

equation image

where Fc is CO2 flux (mg m−2 s−1) and c′ is the fluctuation in the concentration of CO2 (mg m−3) [Desai et al., 2006].

[20] The ratio of the sum of sensible and latent heat fluxes (H + LE) to available energy (the difference of net radiation and soil heat flux: RnG1) is prescribed to examine the surface heating rate ɛ.

3. Results and Discussion

3.1. Footprint Analysis

[21] The code from Schmid et al. [1994] is publicly available (http://www.indiana.edu/~climate/SAM/SAM_FSAM.html) and was used in this study. The 90% source area outlines for measurement height zm (3.9 m) is plotted for both dry and wet seasons in Figure 3, together with the respective maximum source weight locations (xm). The source areas indicate that eddy covariance measurements for both seasons are influenced mainly by the surface patch inside the field of interest. Thus they both are expected to give realistic estimates of the grassland evaporation. In Figure 3, z0 is the aerodynamic roughness length, L is the Obukhov length, σv is the standard deviation of the horizontal wind speed, and u* is the friction velocity. The roughness lengths, calculated by using the method of Martano [2000] for both dry and wet seasons, are much larger than those obtained at other grassland sites. Gao et al. [2004] modeled the surface fluxes for the Naqu flux over the Tibetan Plateau, where the roughness length was set to be 0.001 m because the height of grass canopy top was only 0.05 m. In current work, the height of grass canopy top is 1.2 m, resulting zero plane displacement is 0.8 m, and roughness length is 0.12 m and 0.17 m for wet and dry seasons, respectively.

Figure 3.

The 90% flux source areas for a measurement height of 3.9 m for both dry and wet seasons.

3.2. Seasonal Variations on a Weekly Average Basis

3.2.1. Radiation Components

[22] Figure 4 shows the seasonal variation of the four radiation components: (1) downward shortwave radiation (hereinafter, referred to as DSR), (2) upward shortwave radiation (hereinafter, referred to as USR), (3) downward longwave radiation (hereinafter, referred to as DLR), and (4) upward longwave radiation (hereinafter, referred to as ULR). The albedo of the underling surface, defined as the ratio of the maximum values of USR and DSR, is also given in Figure 4. Two gaps occurring in downward shortwave radiation were caused by instrument problems: one between 15 August and 17 September 2004 and the other between 25 February and 27 April 2005.

Figure 4.

Seasonal variations of weekly mean downward shortwave radiation (DSR), upward shortwave radiation (USR), downward longwave radiation (DLR), upward longwave radiation (ULR), and surface albedo during the period from May 2004 to June 2005 at the grassland site.

[23] Both DSR and USR maintained high values during the wet season (between May and September) and the highest weekly mean values of DSR (238.6 Wm−2) and USR (33.8 Wm−2) occurred in June. Both DSR and USR started to strongly decrease from the beginning of the dry season (i.e., the beginning of October) and kept decreasing through the dry season until reaching minimum values in February.

[24] The seasonal variations of DLR and ULR were similar. They maintained high values during the period from May to August, started to decrease at the beginning of September, maintained low values during the period from mid-November to the end of February and started to increase at the beginning of March. Both the lowest value (309.5 Wm−2) of DLR and the lowest value (364.8 Wm−2) of ULR occurred in December. Both the highest value (447.2 Wm−2) of DLR and the highest value (489.2 Wm−2) of ULR occurred at the end of June. It should be noted that the DLR depends strongly on cloud cover. The seasonal variation of surface albedo is also obvious. It maintained a relatively high value (0.13) during wet summers and a low value (0.11) during dry winters. The result (i.e. that summer surface albedo was higher than winter surface albedo) is due to the fact that in winter more bare soil is exposed and the soil albedo in this area is lower than that of grass.

[25] Figures 2 and 4 show consistent seasonal variation trends in wind vector (WV), air temperature (Tair), specific humidity (q), air pressure (P), precipitation (Prec.), downward longwave radiation (DLR), and upward longwave radiation (ULR).

3.2.2. Energy Components and CO2 Flux

[26] Figure 5 shows the seasonal variation of weekly means of (1) net radiation (Rn), (2) sensible heat flux (H), (3) latent heat flux (LE), (4) soil heat flux (G1) measured at a depth of 0.01 m in soil, and (5) CO2 flux image Rn was calculated by using the four radiation components (i.e., DSR, USR, DLR, and ULR). H, LE, and image were measured by fast response instruments and calculated using equations (2, 3, and 4). The gaps in Figure 5a were caused by the data gaps in DSR and USR shown in Figure 4a. Gaps occurring in H, LE, and image were caused by instrument problems.

Figure 5.

Seasonal variations of weekly mean net radiation (Rn), sensible heat flux (H), latent heat flux (LE), soil heat flux (G0), and CO2 flux image during the period from May 2004 to June 2005 at the grassland site.

[27] All of Rn, LE, and image varied seasonally and dramatically, but no obvious change occurred in H and G1. The relationship between sensible heat flux and the East Asian monsoon onset, given in Figure 5b, agrees with the model output by Xue et al. [2004]. The negative sign in image means that vegetation absorbed CO2. The variations of LE and image are generally consistent. However, the weekly oscillation in LE was weaker than that in image The latent heat flux was the main consumer of net radiation (Rn). Because Rn mainly depends on DSR, the obvious seasonal variation in Rn was mainly caused by the significant seasonal variation in DSR. During the wet summer season, (1) the grass grew well, and the strong photosynthesis led to the larger LE release and the larger image absorption; (2) G1 was positive and was about several watts per square meters. In winter, (1) the grass became yellow and dead, and grass CO2 absorption therefore weakened; (2) G1 was negative and was about several watts per square meters.

3.3. Diurnal Variations on Monthly Average Basis

3.3.1. Radiation Components

[28] In order to investigate the diurnal variation of the radiation components, the monthly means of the diurnal variation in the radiation components (DSR, USR, DLR, and ULR) are given in Figure 6 where composite analysis method is used and the short lines are error bars. We find that diurnal variations in DSR, USR, and ULR occurred in all months, whereas significant diurnal variations in DLR only occurred in the period from May to November 2004. Diurnal variations were dramatic in summer, but weaker in winter. On average, the maximum peak values of DSR and USR occurred in August 2004 and reached 726.1 and 87.2 W m−2, respectively. The minimum peak values of DSR and USR occurred in February 2005, and reached 245 and 25.2 W m−2, respectively. The maximum peak value of DLR occurred in August 2004, and reached 451.5 W m−2 and the minimum peak value of DLR occurred in December 2004 and reached 340.3 W m−2. The maximum peak value of ULR occurred in June 2004 and reached 519.4 W m−2 and the minimum peak value of ULR occurred in February 2005 and reached 417.4 W m−2. Figure 6e shows that the diurnal variation in surface albedo during the winter is much lower than other months, which is a clear indication that in winter there is less vegetation and more bare ground.

Figure 6.

Diurnal variations of weekly mean downward shortwave radiation (DSR), upward shortwave radiation (USR), downward longwave radiation (DLR), upward longwave radiation (ULR), and surface albedo during the period from May 2004 to June 2005 at the grassland site.

3.3.2. Energy Components

[29] The monthly diurnal variation courses of net radiation (Rn), sensible heat flux (H), latent heat flux (LE), soil heat flux (G1), and CO2 flux image are given in Figure 7 where composite analysis method is used and the short lines are error bars. Figure 7a shows that the diurnal variation pattern of Rn is similar to that of DSR; that is, the diurnal variation was significant in summer and weak in wintertime. The maximum diurnal variation occurred in August 2004 and the peak value of Rn reached 558.4 W m−2. The minimum diurnal variation occurred in February 2005 and the peak value of Rn was 180 W m−2.

Figure 7.

Diurnal variation of weekly mean net radiation (Rn), sensible heat flux (H), latent heat flux (LE), soil heat flux (G0), and CO2 flux image during the period from May 2004 to June 2005 at the grassland site.

[30] Figure 7b shows that seasonal variations in sensible heat flux were weaker than those in Rn and in LE. H showed a dramatic diurnal variation, whereas diurnal variations in Rn, LE, G1, and image gradually decreased for the period from October to December 2004. The maximum peak value of H occurred in October and reached 206.3 W m−2 and the minimum peak value of H occurred in February 2005 and was 88.2 W m−2.

[31] Figure 7c shows that the seasonal variation in latent heat flux was dramatic. Obvious diurnal variation of LE occurred in summertime, with a maximum peak value of 213.6 W m−2 occurring in July 2004. Diurnal variation of LE was not significant in wintertime and the maximum peak value was 131.7 W m−2 which occurred in January 2005. Obvious diurnal variation of LE in summer might be attributed to the following two facts: (1) precipitation frequently happened in summertime (as shown in Figure 2) and (2) the diurnal variation in net radiation was dramatic in summertime (as shown in Figure 7a).

[32] Figure 7d shows the seasonal variation of soil heat flux (G1) measured at a depth of 0.01 m. It is interesting to note that the diurnal variation of G1 in May and June 2004 was much more significant than in May and June 2005 when many meteorological factors, such as wind speed, air temperature, humidity, air pressure, and precipitation were almost identical for these 2 months. The reason turns out to be because our instrument installation damaged the natural status of grass and soil around the soil flux plates in May 2004 and that the soil flux measurement is very sensitive to the environment just above the plate. The area recovered after this, so we believe that values of soil heat flux (G1) measured in 2005 were good. Because the grass was very dense, the soil heat flux measured at the depth of 0.01m in soil was very small relative to sensible heat (H) and latent heat (LE) fluxes. Diurnal variation was still measured, and the peak value was about 20 W m−2. However, the seasonal variation in G1 was not significant.

[33] Figure 7e shows that the seasonal variation of image was similar to that of LE, but in the reverse phase. The most significant diurnal variation occurred in September 2004, and the peak value reached −0.77 mg m−2 s−1. Weak diurnal variation occurred in February 2005 when the grass was dead and in yellow and the peak value was only −0.3 mg m−2 s−1.

[34] In summary, Figure 7 shows that the seasonal variations in both latent heat and CO2 fluxes were stronger than those in sensible heat and soil heat fluxes, which implies that both latent heat and CO2 fluxes may be more significant climate signals than sensible heat and soil fluxes.

3.4. Energy Partitioning

[35] The monthly means of the Bowen ratio (βH/LE) were 0.52, 0.49, 0.44, 0.51, 0.56, 0.77, 1.11, 1.16, 1.35, 0.83, 0.66, and 0.44, for the period from May 2004 to May 2005 (except March 2005). It is obvious that the latent heat flux was the main consumer of available energy (RnG1) in summertime, and sensible heat flux was the main consumer of available energy in wintertime. Taking a yearly average, the Bowen ratio was 0.62, LE/Rn = 55%, H/Rn = 34%, and Re/Rn = 11%. Because the site was covered by grass, the soil heat flux was about zero. The proportion of latent heat flux “in net radiation (?)”, LE/Rn, reached a maximum value (0.75) in July 2004, and reached a minimum value (0.39) in January 2005. LE/Rn was less than 0.50 during the period from October 2004 through January 2005, and was larger than 0.5 for the rest of the time. H/Rn reached a maximum value (0.55) in February 2005 and reached a minimum value of 0.27 in August 2004. H/Rn was larger than 0.4 during the period from October 2004 to February, and was lower than 0.3 during summer 2004.

[36] The monthly mean of diurnal variation of residual energy (ReRnHLEG0) is plotted in Figure 8. It is obvious that the diurnal variation pattern of Re is similar to that of Rn; that is, diurnal variation was significant in summer and weak in wintertime. The maximum diurnal variation occurred in August 2004 and the peak value of Re reached 180.0 W m−2 and the minimum diurnal variation occurred in February 2005, when the peak value of Re was 45 W m−2.

Figure 8.

Diurnal variation of weekly mean residual energy (ReRnHLEG0) during the period from May 2004 to June 2005 at the grassland site.

[37] Figure 9 shows the intercomparison of H + LE and RnG1. The surface heating rate ɛ is 0.76 and the correlation coefficient between H + LE and RnG1 is 0.86. Our analysis on the surface-heating rate is focused on the data collected on rain-free days, because the sonic anemometer would be in error during and after rain events. We also neglected the data collected during calibration activities, equipment modification, and large transient fluctuations. Our analyses are also limited to the situations where H, LE, and Rn are all larger than zero because the unavailable measurement errors would be significant when the energy components were close to zero.

Figure 9.

Intercomparison of the measured (H + LE) against available energy (RnG1) during the period from May 2004 to June 2005 at the grassland site.

[38] Theoretically, ɛ should be very close to 1.0. The energy imbalance that occurred for these measurements is unexpected because the experiment was carried over a relatively flat, homogeneous site with sufficient fetches and flux calculations are sophisticatedly treated. Such energy imbalances have also been encountered in other major field campaigns and caused difficulty for their climate applications [e.g. Kahan et al., 2006]. Previous researchers [Foken and Oncley, 1995; Panin et al., 1996; Wicke and Bernhofer, 1996; Foken et al., 1999; Kahan et al., 2006] concluded that the causes of the imbalance of the energy budget were usually related to the errors/uncertainties in the individual energy component measurements and the influence of different footprints on the individual energy components. For our site, (1) we neglected the soil heat storage of the top surface soil layer of 0.01 m; (2) because we measured the soil heat flux at only one point, this point may not be fully representative of the whole site; and (3) the footprints of net radiation, sensible heat, latent heat, and soil heat fluxes are not consistent. These, together with the unavoidable uncertainties that occurred in the individual energy component measurements, are the main causes of the energy imbalance encountered.

3.5. Soil Temperature

[39] Surface radiation and energy budget balances are related to variations in soil temperature and soil water content. Figure 10 shows the seasonal variation of hourly mean soil temperatures at six depths (0.01, 005, 0.10, 0.20, 0.40, and 0.70 m). The seasonal variation trends of soil temperature are close to those of air temperature and specific humidity.

Figure 10.

Temporal variation of soil temperature (K) at depths of 0.01, 0.05, 0.1, 0.2, 0.4, and 0.7 m.

[40] As may be expected, soil temperature in shallow layers changed seasonally and showed dramatic changes. There is evidence of seasonal variation in soil temperature measured at 0.70-m depth. In general the range of seasonal variation measured in the deep layer was much less than that of soil temperature measured in the shallower layers. The highest soil temperatures occurred in July 2004, and were 312.6, 310.8, 307.2, 306.2, 303.6, and 302.2 K for depths of 0.01, 005, 0.10, 0.20, 0.40, and 0.70 m, respectively.

[41] We also examined the diurnal variation of soil temperatures. Results show that soil temperatures diurnally changed in shallow layers, diurnal variation trends weakened with increasing depth, and almost no diurnal variation occurred with soil temperature measured at a depth of 0.7 m.

3.6. Case Study of Diurnal Cycles

[42] In this section we investigate the diurnal cycle of the radiation components, energy fluxes, CO2 flux, and energy balance for clear days under specific radiation environments: (1) on 6 May 2005, the daily downward shortwave radiation reached the strongest value during our experimental period; and (2) on 10 January 2005, the daily downward shortwave radiation reached the weakest value during our experimental period. Figure 11 shows the diurnal cycle of radiation components for these 2 days, the mean diurnal cycle of radiation components for the whole experimental period, and the corresponding daytime surface albedo. The shortwave radiation difference between these 2 days was mainly caused by the difference in surface cover. The diurnal cycles of shortwave radiations are obvious.

Figure 11.

Diurnal variation of downward shortwave radiation (DSR), upward shortwave radiation (USR), downward longwave radiation (DLR), upward longwave radiation (ULR), and surface albedo on 6 May 2005, on 10 January 2005 and on average at the grassland site.

[43] The maximum values of downward shortwave radiation and upward shortwave radiation were 1074.5 and 154.4 W m−2, respectively, on 6 May 2005; 572.3 and 63.7 W m−2, respectively, on 10 January 2005; and 607.7 and 72.5 W m−2, respectively, on average. The corresponding surface albedo values were 0.143, 0.104, and 0.116, respectively. The summer surface albedo is higher than the winter surface albedo because the fraction of bare soil increased in winter and the albedo of soil is lower than that of grass in this area.

[44] The longwave radiation components on 6 May 2005 were greater than those on 10 January 2005. The downward longwave radiation showed almost no daily change, in contrast to the upward longwave radiation.

[45] Similar to Figure 11, the daily cycles of the energy flux components and CO2 flux for the 2 days mentioned above were plotted in Figure 12. The maximum values of net radiation (Rn), sensible heat (H), latent heat (LE), soil heat (G1), and CO2 image fluxes were 810.1, 176.2, 262.3, 37.0 W m−2, and 0.31 mg m−2 s−1 on 6 May 2005. The maximum values of Rn, H, LE, G1, and image were 386.9, 238.1, 56.0, 22.0 W m−2, and 0.11 mg m−2 s−1 on January 10 2005. The maximum values of Rn, H, LE, G1, and image were 443.8, 178.6, 137.0, 31.5 W m−2, and 0.11 mg m−2 s−1 on average. On 6 May 2005, the latent heat flux was larger than the sensible heat flux; on 10 January, the sensible heat flux was larger than the latent heat flux; and on average, sensible heat flux was somewhat larger than latent heat flux. On 6 May 2005, the daytime CO2 absorption was significant because of the strong photosynthesis associated with the grass.

Figure 12.

Diurnal variations of net radiation (Rn), sensible heat flux (H), latent heat flux (LE), soil heat flux (G0), and CO2 flux image on 6 May 2005, on 10 January 2005, and on average at the grassland site.

[46] Figure 13 shows the energy partitioning for 6 May 2005, 10 January 2005, and the average. Latent heat fluxes were 125.6, 20.0, and 55.5 W m−2; sensible heat fluxes were 44.5, 49.8, and 42.7 W m−2; soil heat fluxes were −1.2, −3.2, and 1.3 W m−2; and residual heat fluxes were 25.0, 13.1, and 14.1 W m−2 for 6 May 2005, 10 January 2005, and the average, respectively.

Figure 13.

Surface energy partitioning on 6 May 2005, on 10 January 2005, and on average at the grassland site.

4. Summary and Conclusions

[47] In order to investigate energy partitioning and CO2 exchange over the land surface in the tropical monsoon climate environment and to investigate which surface energy components are strong climate signals eddy covariance measurements were conducted on fluxes of moisture, heat, and CO2 in a near-surface layer over a grassland surface in southern China during the period from May 2004 to July 2005.

[48] The measured values of wind vectors, air temperature, specific humidity, air pressure, and precipitation are compared to historical data collected during the period from 1960 to 2003. Results show that the experimental period is representative of the local climate for this area.

[49] All four radiation components seasonally changed, resulting a seasonal variation in net radiation. The components also changed diurnally, although downward longwave radiation showed almost no diurnal variation during the period from December to April. Summer surface albedo was higher than winter surface albedo because in winter the fraction of bare soil increased.

[50] Appropriate correction was made for turbulent fluxes. The seasonal variations in both latent heat and CO2 fluxes were stronger than those in sensible heat and soil heat fluxes, which implies that both latent heat and CO2 fluxes may be more significant climate signals than sensible heat and soil fluxes. Latent (sensible) heat flux was the main consumer of available energy in summer (winter).

[51] Surface energy partitioning was examined and the surface heating rate (ɛ) was found to be 0.76 during the experiment. The energy imbalance problem was encountered. The reasons for this were thought to be the following: (1) we neglected the soil heat storage of the top surface soil layer of 0.01 m and so the soil heat flux may be underestimated; (2) we measured the soil heat flux at only one point and this point may not fully be representational of the whole site; (3) the footprints of net radiation, sensible heat, latent heat, and soil heat fluxes are not consistent; and (4) the unavoidable uncertainties in the individual energy component measurements.

Acknowledgments

[52] This study was mainly supported by the National Key Basic Research and Development Projects (2006CB400500 and 2006CB403600), by the Centurial Program sponsored by the Chinese Academy of Sciences, and by the National Natural Science Foundation of China (NSFC) project (40575007). The field experiments were supported mainly by the Tropical and Marine Meteorological Foundation (200401), the National Natural Science Foundation of China (40375002 and 40418008), and Guangdong Natural Science Foundation (04003920, 033029, 2004A30401002, 2005B32601011, and 2004J1-0021). Authors are very grateful to four anonymous reviewers for their valuable comments, which led to substantial improvements of this manuscript.

Ancillary