next up previous contents
Next: Seasonal and interannual variability Up: Results Previous: Comparison of the calculated   Contents

Environmental forces

Based on the long term measurement data, we can investigate the response of the measured NEE to the environmental factors. The main variables controlling the behaviour of the vegetation is the incoming photosynthetically active photon flux density (PPFD), temperature, soil mositure and vapor pressure deficit (VPD, defined as the difference between the saturated water vapor pressure and the actual water vapor pressure at a given temperature in the atmosphere).

PPFD is the most immediate environmental control on photosynthesis (Malhi et al., 1999). The major influence of temperature on net carbon balance is through its effects on rates of both autotrophic and heterotrophic respirations (Malhi et al., 1999). Since we can not distinguish between these two types, we may only describe the net effect of respiration. Vapor pressure deficit controls stomatal closure, thus it has a direct effect on the rate of photosynthesis. High VPD causes stomatal closure, decreasing photosynthesis (Anthoni et al., 1999). Lack of soil moisture also reduces carbon uptake by causing stomatal closure, but also affects soil carbon and nutrient release by restricting microbial decomposition (Malhi et al., 1999).

PPFD, VPD and air temperature are currently measured by our system. Soil moisture measurement will be installed in the near future at Hegyhátsál.

Figure: Daily aggregated environmental variables for the 1997-1999 period. Upper plot: daily mean air temperature at 10 m; middle plot: daily sums of PPFD; lower plot: mean daylight water vapor pressure deficit.

Figure [*] shows the temporal course of the daily aggregated environmental variables controlling the carbon budget of the vegetation for 1997, 1998 and 1999 (i.e. for the whole 82 m EC measurement period, so far3.1).

Based on the PPFD and NEE time series, it is possible to construct the light response function of NEE for the different months during the growing season. This can be used for modelling purposes.

Figure: Light response curve of NEE based on all available data for August, 1997-1999. NEE calculated between 8 h and 19 h UTC+1 are plotted against PPFD. Number of samples: 788. X sign: NEE with VPD1.5 kPa; square sign: NEE with VPD\( \geq \protect \)1.5 kPa. Negative NEE indicates CO\( _{2}\protect \) uptake by the vegetation. The functional form of the solid line is eq. [*].

Figure [*] shows the NEE-PPFD function for August, 1997-1999. Data was measured between 8 h and 19 h UTC+1. Number of samples = 788. NEE was classified by the measured VPD. The figure shows that on average, higher VPD causes slight decrease (in absolute meaning) of NEE, as it is expected from the theory. The fitted equation is a Michaelis-Menten type rectangular hyperbola (Valentini et al., 1996; Markkanen et al., 2001). The NEE-PPFD function has the following form (solid line in figure [*]):

\end{displaymath} (3.6)

where \( a=-1.154 \) mg m\( ^{-2} \) s\( ^{-1} \), \( b=980.961 \) \( \mu \)mol m\( ^{-2} \) s\( ^{-1} \), \( c=0.174 \) mg m\( ^{-2} \) s\( ^{-1} \) for August, PPFD is given in \( \mu \)mol m\( ^{-2} \) s\( ^{-1} \), and NEE is in mg CO\( _{2}\protect \) m\( ^{-2} \) s\( ^{-1} \). Light response curve of NEE is estimated for each month during the growing season (March-October) using all available data from 1997, 1998 and 1999. The coefficients of the NEE-PPFD function is shown in table [*].

Table: Coefficients of the NEE-PPFD response function during the growing season estimated for the 1997-1999 period. The general form of the function is eq. [*]. The number of hourly samples (\( n \)) used for the fitting is also shown.
  \( a \) [mg m\( ^{-2} \) s\( ^{-1} \)] \( b \) [\( \mu \)mol m\( ^{-2} \) s\( ^{-1} \)] \( c \) [mg m\( ^{-2} \) s\( ^{-1} \)] \( n \)
March -0.246 73.525 0.192 350
April -0.492 244.387 0.184 340
May -1.201 663.331 0.246 493
June -1.302 600.823 0.324 636
July -1.389 799.208 0.259 374
August -1.154 980.961 0.174 788
September -0.622 226.867 0.194 497
October -0.482 63.649 0.299 318

Figure: Relationship between NEE and air temperature at 10 m during nighttime. NEE measured between March-August is plotted from all available years (1997-1999). Plus sign: data measured during March; grey asterisk: data from April-June; X sign: data from July and August. (September and October is not plotted for clarity). All data defines a single curve, although very high scatter is observed.

Investigation of nighttime NEE as a function of air temperature showed that it is not reasonable to calculate NEE-t response function for each month separately because of the large scatter of the observed data (Greco and Baldocchi, 1996; Valentini et al., 1996). Instead of this, one curve is fitted to all data measured during the growing season.

Figure [*] shows the NEE-air temperature function (air temperature measured at 10 m is used) during nighttime for March-August, 1997-1999. Number of samples used for fitting = 3356 (includes data from March-October, 1997-1999; September-October data is not plotted but used for fitting). Nighttime period before day of year (DOY) 100 and after DOY 260 is defined as the period between 18 h and 5 h UTC+1, and the period from 20 h till 3 h between DOY 100 and DOY 260. The fitted curve has the following form:

NEE=x\exp \left( y\cdot t_{10}\right)
\end{displaymath} (3.7)

where \( x=0.06 \) and \( y=0.063 \) (the linear correlation coefficient is only 0.21) and \( t_{10} \), the air temperature at 10 m, is given in Celsius. The Q\( _{10} \) coefficient (i.e. the ratio of the rate of respiration at one temperature to that at a temperature 10 degrees lower) is 1.88.

Outside the March-October period, another NEE-\( t_{10} \) function is constructed using both daytime and nighttime data using data from all available years. The coefficients of the fitted curve are \( x=0.041 \) and \( y=0.014 \) (data not shown). This curve shows good agreement with the previous one in the -5 - 0\( ^{o} \)C interval. This curve is supposed to be used for NEE modelling purposes both for daytime and nighttime in this period as a function of air temperature.

The same procedure is performed for the Japanese system. Air temperature measured by the Kaijo-Denki anemometer/thermometer at 3 m was used to determine the NEE-\( t_{3} \) relationship during nighttime. PPFD measured by the profile system was used to fit the NEE-PPFD curves in the same way as it is done in the case of the 82 m system.

Coefficients of the light response curve are presented in table [*] for the growing period of 1999 (except October, when no PPFD data was available).

Table: Coefficients of the NEE-PPFD function during the growing season of 1999-2000 for the 3 m system. The general form of the function is eq. [*]. The number of half hourly samples (\( n \)) used for fitting is also shown.
  \( a \) [mg m\( ^{-2} \) s\( ^{-1} \)] \( b \) [\( \mu \)mol m\( ^{-2} \) s\( ^{-1} \)] \( c \) [mg m\( ^{-2} \) s\( ^{-1} \)] \( n \)
March -0.386 63.833 0.244 823
April -1.403 603.01 0.209 1167
May -1.329 446.365 0.316 1338
June -1.507 636.586 0.34 621
July -1.303 451.294 0.389 650
August -0.815 612.294 0.234 1016
September -0.525 258.952 0.204 666
October -0.378 330.391 0.084 209

During the vegetative period, one NEE-\( t_{3} \) curve is constructed similarly to the 82 m system. Since air temperature at 3 m is measured by the system, the ensemble temperature response curve is estimated from all available data from 1999 and 2000. The coefficients of this ensemble curve are \( x=0.066 \) and \( y=0.08 \), to be used with eq. [*] for months between March and October. The Q\( _{10} \) coefficient is 2.23 here. This is higher than that estimated for 82 m for the same period mainly caused by the higher respiration rates of the vegetation (see fig. [*]) that is sensed by the 3 m system (source area, see section [*]) during nighttime.

For months between November and February, the exponential function defined by \( x=0.035 \) and \( y=0.04 \) should be used both for daytime and nighttime.

next up previous contents
Next: Seasonal and interannual variability Up: Results Previous: Comparison of the calculated   Contents
root 2001-06-16