Conclusions

In the present dissertation an on-going program was described being carried out in Hungary to investigate the role of the temperate continental region in the global carbon cycle. Three closely linked measuring systems were described, which are operated near the village Hegyhátsál. The equipment is installed on a 117 m tall, free-standing TV and radio transmitter tower, and near the ground in close proximity of the tower.

One of the measuring systems is designed to measure the vertical profile of carbon dioxide mixing ratio and other meteorological elements. Design of the measuring system allows to calculate vertical fluxes of carbon dioxide by means of the classic, semi-empirical Monin-Obukhov similarity theory.

Another measuring system is designed to measure directly the vertical transport of carbon dioxide at 82 m height. The so-called eddy covariance technique is used which is the most common and reliable way currently available to determine the turbulent transport of carbon dioxide (and other scalars). The measurements made at 82 m are representative to an extedned, 8-10 km radius circle around the tower.

The installation of the second direct flux measuring system at 3 m provided the possibility to determine the carbon budget of the tower's surroundings. This measurement has a very limited flux footprint hence the results are representative only to the vegetation surrounded by the tower.

Despite the very different source area of the two measuring system, the local scale system may help to detect possible systhematic errors in the larger scale measurement (e.g. underestimation of the nighttime fluxes, intermittency, etc.) because of its small distance from the ground. During 1999 when both measuring system was in operation, yearly Net Ecosystem Exchange measured by the two systems was approximately equal. Even though the two spatial scales are very different, it may indicate that the measuring system at 82 m works well notwithstanding the different approach and possible further error sources (e.g. storage).

It has been found that the region acts as a net sink of carbon dioxide in an annual scale, but the amount of sequestered carbon dioxide is variable in time due to changes in the environmental conditions. The ecosystem sequestered 92-146 g C m $^{-2}$ year $^{-1}$ (0.92-1.46 t C ha $^{-1}$ year $^{-1}$ , or 3.37-5.35 t CO $_{2}\protect$ ha $^{-1}$ year $^{-1}$ , depending on the year). In 2000 the local system recorded extremely high carbon sequestration. Unfortunately, this high value cannot be confirmed due to the malfunction of the regional scale system.

Table

shows the result of year-long NEE studies conducted over very different ecosystems. Our result (i.e. NEE, not Gross Primary Production or total ecosystem respiration!) is comparable with results from boreal environment, temperate croplands (Falge et al., 2001) or in some cases with results from temperate deciduous forests or even with NEE measured at tropical rain forests (Grace et al., 1995).

As the carbon balance of the measured region is very sensitive to respiration, it was found that the region may switch from being a carbon source to a sink as climate change proceeds (Malhi et al., 1999; IPCC website, 2001).

Table: Comparison of year-long NEE measurements.

Author	vegetation type	NEE [gC m $^{-2}$ year $^{-1}$ ]
Malhi et al., 1999	boreal forest	-70
Lindroth et al., 1998	boreal forest	-70 - -220
Markkanen et al., 2001	boreal pine forest	-191 - -262
Black et al., 1996	boreal hazel nut/aspen	-130
Malhi et al., 1999	temperate forest	-585
Goulden et al., 1996	mixed temperate deciduous	-140 - -280
Valentini et al., 1996	temperate beech	-472
Yamamoto et al., 1999	deciduous forest (birch, oak)	-180
Greco and Baldocchi, 1996	deciduous forest	-525
Falge et al., 2001	corn (temperate region)	-467
Falge et al., 2001	wheat (temperate region)	-183
Falge et al., 2001	tallgrass prairie	-318
Malhi et al., 1999	tropical forest	-590
Grace et al., 1995	tropical forest	-100

It is extremely important to conduct long term, direct carbon dioxide exchange measurements to quantify the carbon budget of the vegetation and to estimate the effect of the global climate change (Baldocchi et al., 1996). Our current knowledge is insufficient to close the global carbon cycle, and to estimate the future of the biosphere under changing climatic conditions (Malhi et al., 1999).

No perfect method exists to quantify the behaviour of the biosphere because of methodological difficulties. There is also uncertainty in the estimation of the soil carbon stocks. One important future task is the investigation of the soil carbon fluxes (Malhi et al., 1999) to constrain the soil carbon content. The upscaling of the local, sometimes site-specific measurements also causes uncertainty. Our approach was to try to avoid the upscaling problem by developing a measuring system which is designated to be representative for the regional scale with mixed, patch-like vegetation. Our approach fits the trend of micrometeorological measurements, which is the relocation of the measuring sites from flat, homogeneous (ideal) terrains to heterogeneous, sometimes sloping terrains which are characteristic of the real world (Baldocchi et al., 2000). This new approach raises many new problems associated with the change in the dynamics of the atmospheric motions (e.g. changes in the roughness, humidity, albedo, vegetation type, possible onset of local circulations, advection, etc.), thus much more attention must be paid to ensure that the results of the measurements are reliable, and reflect the real dynamics of the region. Detailed overview was given in this work considering as many problems as possible to assure data quality.

The region was not expected to act as a strong sink compared to forests, but its current carbon sequestring role is not insignificant and should be considered in any integrated, continent-scale carbon balance study.

There is no current agreement in the localization of the ``missing sink'' (Malhi et al., 1999). Different research groups detected different places for the sink (e.g. North America or Siberia after Fan et al. (1998) or Bousquet et al (1999), respectively). There are very few studies that approaches consistency with the current, ground-based results.This is mainly caused by methodological problems (Malhi et al., 1999). Delevopment must be done in this field to be able to utilize the ground-based data to constrain the models that describe the global carbon cycle.

In summary, our current knowledge of the carbon cycle is deeper than it used to be 10-20 years ago, but still, a great deal of work is needed to gain better insight into the biochemical and meteorological processes that govern the carbon balance of the biosphere and the CO $_{2}\protect$ content of the atmosphere.