JP5039195B2 - Forest CO2 flux measurement method - Google Patents

Description

The present invention relates to a method for measuring CO ₂ flux in forests. More specifically, the present invention relates to an improvement in the technique for evaluating the CO ₂ absorption amount of the entire forest, which has been evaluated by the eddy correlation method (EC).

In order to use the forest absorption source in the Kyoto mechanism as a greenhouse gas reduction measure, it is necessary to show and implement a method for evaluating the amount of CO ₂ absorbed in the forest absorption source. In order to quantify the impact of forests on global warming suppression and climate change, it is important to evaluate the CO ₂ balance of the entire forest (full carbon accounting). Such forest overall CO ₂ balance, which can be evaluated by integrating exchange amount CO ₂ (the flux), the current, the evaluation and prediction of the global CO ₂ balance has been actively conducted. In conducting such evaluations and predictions, the use value as verification data of energy and mass exchange (flux) measured on the ground surface is high. The development of an environment where data can be used as ground truth is being promoted worldwide. In addition, there is currently an evaluation method for CO ₂ absorption recommended by IPCC (Intergovernmental Panel on Climatic Change, a UN organization established in 1988, where scientific discussions on global warming issues are held). this evaluation method is the fact that can not be accurately grasp the CO ₂ emission from the soil of the CO ₂ balance of the forest, a new technique that the CO ₂ balance of the forest can be correctly evaluated there is also a surface that will be required in the future.

For example, if the size of a small park is large, there will be a difference in the amount of CO ₂ in and out, but in the case of a large forest such as tens or hundreds of ha, the amount of in and out is considered to be almost equal. it can. Therefore, when evaluating the amount of CO ₂ absorbed in the forest, assuming that the forest is uniform in the lateral direction, the flow rate of CO ₂ (exchange) (Quantity) can be treated as almost equal (see FIG. 1). Therefore, based on this idea, the amount of CO ₂ in the forest during the year can be measured by measuring, for example, the amount of air entering and exiting the vertical space in the upper space of the forest, for example, without considering the lateral flow rate. Is absorbed or released (see FIG. 1). More specifically, if the CO ₂ exchange rate in the vertical direction of the forest is kept tracked and integrated at the end, the absorption amount (release amount) for a certain period (for example, one year) can be found. The amount of absorption or release, that is, the amount of CO ₂ exchanged as an amount per unit time is referred to as “flux”, that is, so-called exchange rate (see FIG. 1).

Here, NEE representing the balance of the CO ₂ in the forest (Net Ecosystem Exchange: net ecosystem exchange amount), using a CO ₂ flux _{F C} is expressed by the following equation.
[Equation 7]
NEE _{= F} C + F _S
F _S is the amount CO ₂ temporarily present in the space in the forest (canopy in storage amount), repeated storage and release during the day, the balance is very small throughout the day. Therefore, NEE for one year can be obtained by integrating the time series of CO ₂ flux for one year. One year of NEE (mgCO ₂ m ⁻² yr ⁻¹ ) indicates the amount of CO ₂ absorbed by the forest ecosystem in one year due to the assimilation of photosynthesis (see FIG. 5).

Conventionally, an eddy-covariance method (EC) has been widely used as one of the methods for measuring the CO ₂ flux. The eddy correlation method, 'the temperature and humidity, CO ₂ measurements of density variations (c variation of vertical wind velocity (w)' The average covariance) (w'c '(statement not given the upper line Not, but represent the arithmetic mean))) to determine the turbulent flux, ie, as shown in FIG.
CO ₂ flux = turbulence of wind × CO ₂ density fluctuation = w′c ′
The CO ₂ flux is obtained according to the equation: This measurement method using the vortex correlation method is the measurement method with the fewest assumptions because it directly measures each physical quantity. Also, when using this eddy correlation method, facilities for directly measuring changes in wind flow and CO ₂ density over the forest, such as a tower built in the forest or the tower, are provided. The sensor etc. which were used are utilized (for example, refer nonpatent literature 1). As the sensor, an ultrasonic anemometer for measuring wind turbulence, a CO ₂ / H ₂ O analyzer for measuring CO ₂ density fluctuation, or the like is used.

Thiermann, V. "A displaced-beam scintillometer for line-averaged measurements of surface layer turbulence". Tenth symposium of turbulence and diffusion, 29 Sept -2 Oct, 1992, Portland, OR., Published by the American Meteorological Society, Boston, MA .: p244-p247 (1992)

  However, there is a view that the measurement accuracy is not sufficient in the case of the conventional measurement method. In other words, since the flux is driven by atmospheric turbulence near the ground surface in the lower atmosphere, the heat flux by the eddy correlation method is used in the stable atmosphere such as at night on a clear day, or in unstable conditions with large atmospheric convection. There are many phenomena (imbalance) in which the radiation flux from the radiometer is not balanced. Such a phenomenon (imbalance) is recognized as a problem mainly caused by the vortex correlation method.

Therefore, an object of the present invention is to provide a forest CO ₂ flux measurement method for accurately evaluating the amount of CO ₂ absorbed in the entire forest.

In order to achieve this object, the present inventor has conducted various studies. In order to evaluate the amount of CO ₂ absorbed in the entire forest with high accuracy, it was considered to measure the time average or spatial average CO ₂ flux in the entire forest. However, as described above, there is a view that the measurement accuracy is not sufficient with the eddy correlation method, and as a result of studying this problem, the present inventor performs measurement at one point over the forest, so-called “point measurement”. As a result of further study, we thought that there was a reason for the type, and as a result, we used a scintillometer that can take a long path (the distance between the transmitter and the receiver), and CO ₂ flux in a wide area I came to the idea of measuring. In the case of a scintillometer, a transmitter and a receiver are installed at a distance of several tens of meters or 100 m or more, and a change in the refractive index of the laser beam between them is measured, so that at least a CO ₂ flux over a wider area than in the vortex correlation method. Can be measured.

However, in the case of a scintillometer, only the temperature turbulence intensity can be measured, so that the data relating to “CO ₂ density fluctuation”, which is an element for obtaining the above-mentioned CO ₂ flux, cannot be directly obtained. There is also. That is, even be expected to increase the measurement accuracy by the flux measured in a broader area, it becomes impossible elements begin CO ₂ flux measurement to be able to determine directly for obtaining the CO ₂ flux End up.

Here, the present inventor further studied in order to obtain the CO ₂ flux while using a scintillometer that can take a long path as a measuring means. Therefore, in order to investigate the influence of the source area on the difference in measurement technique, we came up with a comparison of the dissipation rates of the source area, scintillometer, and eddy correlation sensor, and as a result of further investigations, we obtained some knowledge.

The present invention is based on such knowledge. First, as a method for measuring CO ₂ flux in a forest, when a laser beam emitted from a scintillometer transmitter over the forest is received by a scintillometer receiver, turbulence or temperature fluctuations in the forest are detected. The variation of the light intensity based on the scintillation due to the temperature is measured to obtain the extinction rate ε _T of the temperature variation, and further, the vortex correlation sensor over the forest is used to calculate the square root σ _T of the variation with respect to the temperature, Measure the square root σ _q , the square root σ _c of the variance of the variation with respect to CO ₂ , and then
Based on the above, the present invention has been invented in that it calculates the water vapor extinction rate ε _q and the CO ₂ extinction rate ε _c , and obtains the flux from the respective extinction rates ε _T , ε _q , ε _c which are scalar quantities. .

For example, when air turbulence or heat fluctuation occurs over the forest, light scattering occurs and the light intensity fluctuates. Therefore, the dissipation rate of turbulent kinetic energy or the dissipation rate of temperature fluctuation is obtained based on the light intensity that has changed in this way (see FIG. 2). Further, the heat flux can be obtained by multiplying the air turbulence and the CO ₂ density, or by multiplying the air turbulence and the temperature fluctuation. Here, there is a pair of relationship between flux generation and dissipation. Therefore, it is possible to calculate the flux indirectly by obtaining the dissipation rate.

However, in this method, the CO ₂ flux itself cannot be obtained directly, and only two of sensible heat flux (heat transfer) and wind speed shear (turbulent strength) are obtained directly. Therefore, a new method for obtaining the CO ₂ flux or the water vapor flux by applying this method is introduced. That is, as described above, since there is a pair of relationship between flux generation and dissipation, in the present invention, the flux is indirectly calculated by obtaining the dissipation rate. In this case, it is characteristic that the flux corresponding to the spatial scale of the scintillometer is calculated from the measured value by the vortex correlation sensor.

More specifically, it is possible to indirectly measure the flux by using a scintillometer, but in this scintillometer, only the strength of temperature and turbulence is known. Therefore, the ratio of the dissipation rate of the temperature and water vapor, the temperature and introduced a relational expression equal to the square of the ratio of the dispersion of water vapor, from the data measured by the eddy correlation sensor, the variance of the change relating to the temperature square root sigma _T and steam The ratio of the square root σ _q of the variance of the fluctuations with respect to temperature, and the ratio of the square root σ _T of the variance of the fluctuations with respect to temperature and the square root σ _c of the variance of the fluctuations with respect to CO ₂ , from which, depending on the scintillometer, The extinction rate ε _{q of} water vapor that cannot be measured and the extinction rate ε _c of CO ₂ are calculated (see FIG. 3). If the extinction rates ε _T , ε _q , ε _c that are scalar quantities can be obtained in this way, a flux (for example, CO ₂ flux) can be obtained from these results.

In some cases, underestimation of the heat flux was observed in the eddy correlation method. In contrast to such a measuring method, in the case of the present invention, the heat flux by the scintillometer relaxes the imbalance of the heat balance, and a more appropriate evaluation result can be obtained. In addition, since the CO ₂ flux is based on the same measurement principle as the sensible heat and latent heat flux, a more reliable value can be obtained by applying a scintillometer.

Here, as a result of repeated studies, the present inventors have found that the turbulent kinetic energy and scalar dispersion dissipation rate (dominant parameters for flux calculation) obtained from the scintillometer applied on the forest are orthogonal to the measurement path. We found a phenomenon that characteristically exceeds the dissipation rate of the vortex correlation sensor in the wind direction, and thought that one of the causes of this phenomenon is that the source area differs greatly between the scintillometer and the vortex correlation sensor. Based on this knowledge, when carrying out the above-mentioned CO ₂ flux measurement, the transmitter and the receiver are arranged or arranged so that the direction of the path as the irradiation direction of the laser light in the scintillometer is orthogonal to the wind direction in the forest. It is preferable to convert. From the viewpoint of expanding the measurement area by the scintillometer, if the path direction is orthogonal to the wind direction as described above, a greater effect can be obtained when measuring the CO ₂ flux using the scintillometer (FIG. 4). reference). For example, if the paths are orthogonal when the wind direction is constant or stable, the measurement accuracy is increased. Further, particularly when the measurement path is made longer, the scintillometer has an advantage that a wider area measurement is possible because the flux can be derived in a wide area. In the case of the present invention, if at least one of the transmitter and the receiver can be moved, the path direction can be appropriately changed according to the wind direction.

The above is the forest CO ₂ flux measurement method using a scintillometer, but the present inventor has further studied. If the scintillometer is used, the above-described effects can be obtained, but if the optical path of the laser beam is interrupted, measurement becomes impossible, so that data during bad weather such as rain or fog cannot be obtained. When the wind direction is orthogonal to the measurement path of the scintillometer, a value representing the region is obtained, but when the wind direction is parallel to the path, an evaluation value equivalent to the vortex correlation method is obtained. In order to obtain the amount of CO ₂ absorption by integrating the flux, it was considered necessary to stably calculate a flux having a scintillometer spatial scale over a long period of time. As a result of further investigation, the ratio ε _TSAS / ε _TEC of the temperature fluctuation dissipation rate by the scintillometer and the vortex correlation method is (i) the ratio of the measurement range of the sensor (in other words, the ratio of the source area, RA = A _SAS / A _EC ) and (ii) Atmospheric turbulence intensity (σ _w / u) is highly correlated, so these two elements are used as parameters to convert the vortex correlation measurement value into a flux equivalent to a scintillometer. I came up with an idea.

The invention according to claim 1 has been conceived based on such an idea. In the CO ₂ flux measurement method for measuring the CO ₂ absorption amount of the entire forest, the case of using a scintillometer and the case of using the eddy correlation method The ratio of extinction ratio of temperature fluctuation of ε _TSAS / ε _TEC to the ratio RA (= A _SAS / A _EC ) of the measurement range A _SAS when using a scintillometer and the measurement range A _EC when using a vortex correlation sensor , And the high correlation with the atmospheric turbulence intensity σ _w / u, the relational expression with these two elements RA (= A _SAS / A _EC ) and σ _w / u as parameters
Is prepared in advance based on the measurement results in a certain period actually obtained by using a scintillometer and a vortex correlation sensor, and a constant value is given to the ratio RA of the measurement range in Equation 10 for all wind directions. The wide measuring range of the scintillometer is applied, and then the turbulence intensity σ _w / u of the atmosphere is measured by the eddy correlation sensor actually installed above the forest.
The sensible heat flux H _EC obtained by the vortex correlation method is corrected to the sensible heat flux H _SAS corresponding to that obtained by the scintillometer, and the latent heat flux λE _EC and CO ₂ flux F obtained by the vortex correlation method are corrected. Regarding _cEC , the following relational expression
Based on the above, the latent heat flux λE _SAS and the CO ₂ flux F _cSAS corresponding to those obtained by a _{scintillometer} are corrected, and the same measurement result as that obtained by the scintillometer is obtained. In this case, RA is the spatial scale of the scintillometer that varies depending on the wind direction, and σ _w / u represents the turbulent state of the atmosphere.

  In the case of the present invention, the function (Formula 10) having the variables (i) and (ii) described above as variables is prepared in advance as an experience value based on past measurement results actually using a scintillometer and a vortex correlation sensor. . Subsequent measurements are performed only using the vortex correlation method, and actual measurements using the scintillometer are no longer performed, but the same measurement results as when using the scintillometer are obtained by correcting the flux using the relational expression prepared in advance. Obtainable. In other words, it is not necessary to actually install a scintillometer in the measurement target forest, and measurement data can be obtained by performing actual measurement using only the vortex correlation sensor according to the same procedure as before. For example, the measurement result using only the vortex correlation sensor tends to be underestimated than the actual value, but in the present invention, this result is corrected or interpolated based on a specific relational expression, and it is obtained as if using a scintillometer. Similarly, a measurement result closer to the true value can be calculated.

According to the above-mentioned CO ₂ flux measurement method, it is possible to measure the flux in a wide area by using a scintillometer that can extend the measurement path. For example, in the case of the conventional eddy correlation method, the measurement area is very narrow because it is measured at a point, whereas in the case of the flux measurement method according to the present invention, a large area can be measured, so it does not depend on fluctuations in the narrow area. Average and highly accurate measurement data can be obtained. Thereby, it becomes possible to evaluate the CO ₂ absorption amount of the entire forest with high accuracy.

In the above-mentioned CO2 flux measurement method, a wider measurement area can be secured by making the path orthogonal to the wind direction or by approximating it (see FIG. 4). According to this, it becomes possible to evaluate the CO ₂ absorption amount of the entire forest with higher accuracy.

According to the first aspect of the present invention, it is possible to obtain a highly accurate measurement result similar to that using the scintillometer without using the scintillometer. In such a case, since it is not necessary to provide a tower for installing each of the transmitter and the receiver, it is possible to evaluate the CO ₂ absorption amount with high accuracy while suppressing the cost. In short, by applying corrections derived from different measurement values in the measurement area, it is possible to obtain the same results as if measurement was performed with a scintillometer, and to eliminate the heat flux imbalance by the vortex correlation method. It is possible to make it.

  Moreover, because the scintillometer is an optical measuring device, it is easy to be affected by the weather, such as the operating rate drops in the case of rain or fog, but it has been made possible to obtain highly accurate results without actually using the scintillometer. The measurement method according to the present invention also has an advantage that measurement can be performed at any time without being affected by the weather.

  Hereinafter, the configuration of the present invention will be described in detail based on embodiments shown in the drawings.

1 to 8 show an embodiment of the present invention. The CO ₂ measurement method according to the present invention is to measure and evaluate the CO ₂ flux in the entire forest. When the laser light emitted from the transmitter 2 of the scintillometer 1 over the forest is received by the receiver 3 of the scintillometer 1, the forest Measurement of light intensity fluctuation based on scintillation due to turbulent flow or temperature fluctuation on the surface to obtain temperature fluctuation extinction rate ε _T , or prepared as an empirical value in advance, and further eddy correlation sensor over forest Measure the variance of the variation with respect to temperature (or its square root σ _T ), the variance of the variation with respect to water vapor (or its square root σ _q ), the variance of variation with respect to CO ₂ (or its square root σ _c ), and then formula
Based on the above, the water vapor extinction rate ε _q and the CO ₂ extinction rate ε _c are respectively calculated, and the flux is obtained from the respective extinction rates ε _T , ε _q , ε _c which are scalar quantities.

First, various terms in this specification will be described. First, in general, in the case of “flux”, the energy of the turbulence (momentum flux) carried in the vertical direction by the turbulence of the air current and the movement speed per unit area of heat / material (sensible heat flux, latent heat flux, CO ₂ flux) Means that. In the present specification, the exchange amount of CO ₂ or the like per unit time is called a flux. The “vortex correlation method” is a technique for calculating flux from fluctuation components of horizontal and vertical wind speeds and masses measured using an ultrasonic anemometer and a CO ₂ / H ₂ O variometer. “Dissipation rate” is a value representing the dissipation of atmospheric turbulence energy and heat / material density fluctuations. The turbulent flux is calculated from the balance equation for the generation and dissipation of turbulent energy in the surface boundary layer.

  In this embodiment, first, an outline of a flux calculation method using the scintillometer 1 will be described. Next, the flux calculation method of extending the method and correcting the measured value of the vortex correlation method using a function determined from the spatial scale of the scintillometer 1 will be described.

A scintillometer 1 (SAS: Small Aperture Scintillometer is used in this embodiment) is composed of a pair of a transmitter 2 that emits a visible laser and a receiver 3 that detects a laser intensity. The change in refractive index is regarded as the change in light intensity, and the extinction rate (ε _T ) of temperature fluctuation, which is a parameter for calculating the sensible heat flux, is calculated (see FIG. 2). The dissipation rate of temperature fluctuation can also be calculated from the time series of wind speed and temperature measured by the vortex correlation sensor. By applying Moninofkov's similarity law to the flux balance equation, the dissipation rate can be determined by the following equation as sensible heat H (Wm ⁻² ), latent heat λE (Wm ⁻² ) and CO ₂ flux F _C (mgCO ₂₎ m ⁻² s ⁻¹ ).
Here, _{u *} is the friction velocity ^{(ms -1),} ρ is air density ^(gm _{-3), k kar} Kalman constant (= 0.4), z is the measurement height (m), _{c p} is the specific heat at constant pressure of the air (Jg ⁻¹ K ⁻¹ ) and λ represent the latent heat of vaporization (Jg ⁻¹ ) of water. ε _q and ε _c are the extinction rates of the specific humidity (q kgkg ⁻¹ ) variation and the CO ₂ density (mgCO ₂ m ⁻³ ) variation, respectively. φε ₍ ζ) is a parameter that expresses atmospheric stability with a non-dimensionalized dissipation factor [Equation 19]
ζ = z / L _mo (L _mo : Monin Obkov length)
Is given as a function of

Subsequently, the contents of a method for calculating latent heat flux and CO ₂ flux by obtaining ε _q and ε _c from u _* and ε _T determined by the scintillometer 1 and dispersion of q and CO ₂ density fluctuations measured simultaneously are described. To do.

  The flux calculation by the scintillometer 1 is performed based on the dissipation method. Turbulence consists of generation and dissipation. The vortex correlation method corresponds to the direct measurement of scalar fluctuation and turbulent kinetic energy (Turbulent Kinetic Energy = TKE) generation, whereas the dissipation method uses scalar fluctuation and It is a technique for calculating their generation from the dissipation rate of TKE (a positive value representing energy dissipation). Dissipation method has been applied to measuring flux at sea because it is not affected by low frequency noise such as ship motion. In the flux calculation by the dissipation method, first, the dissipation rate of fluctuation of TKE and scalar is obtained, and then the flux is calculated from the balance equation of flux and dissipation rate. The scintillometer 1 is a device for measuring the dissipation rate of TKE and scalar fluctuation.

The principle of the dissipation method will be described below. The turbulent kinetic energy TKE (e _t ) and scalar variation balance equations in the horizontal uniform and steady ground boundary layer are written as follows:

Here, s represents a scalar such as temperature T, specific humidity q, or CO ₂ , and p is atmospheric pressure [Equation 22].
θ ′ _v = θ (1 + 0.61q)
The provisional potential temperature (where θ is Yutakai), _{e t} represents the ^{_{TKE (e t = (u 2}} + v 2 + w 2) / 2). Here, u, v, and w are wind velocity components in the average wind direction, the cross wind direction, and the vertical direction. Although no upper line is added in this sentence, the bar in the formula represents the average value, and the prime represents the fluctuation from the average value. Each term of Formula 20 is a generation term, a buoyancy term, a transport term, a pressure term, and a dissipation factor. Each term of Equation 21 is a generation term, a transport term, and a dissipation rate. The first terms u'w 'and w's' in both formulas (however, the upper line is omitted in the text) correspond to the momentum and scalar flux to be obtained. Assuming that both terms can be ignored in the contact boundary layer, the friction velocity u _* , defined by Equation 23,
Using Monin-Obukov length (L _mo ) and Karman constant (k _kar 0.4), u _* ³ / k _kar (z−d ₀ ) and k _kar (z−d ₀ ) u _* / (w ′s ') When dimensionless by ² , the terms of Equation 20 and Equation 21 are calculated at the measurement height z (m).
It is expressed. phi _epsilon and phi _.epsilon.s is a dimensionless dissipation rate of TKE and scalar, the phi _m and phi _s represent the vertical gradients dimensionless TKE and scalar. d ₀ is a zero plane displacement, and is a parameter for correcting upward the reference plane for measuring the height when the logarithmic distribution of the wind speed profile on the vegetation is assumed. The value of community height x 0.7 is often used. Further, here the dissipation rate of TKE and scalar variation is defined as follows (Brutsaert, 1982).

In a horizontal uniform contact boundary layer, these dimensionless dissipation rates are a function of the parameter ζ = (z−d ₀ ) / L _mo (stable ζ> 0−, unstable ζ <0) indicating atmospheric stability. Expressions 24 and 25 indicate that they are expressed. The Monin-Obukov length L _mo used here is defined by the following equation.

ρ is air density, T _v is the virtual temperature, H, λE is sensible and latent heat flux in the earth's surface. Solving the equations for s = θ and s = q in Equation 26 and Equation 27 for u _* , H, and λE, respectively,

Thus, if the function form of the dimensionless dissipation factor is known, Equations 30 to 32 can be solved as simultaneous equations by giving the dissipation factor, and fluxes u _* , H, and λE are obtained. The dimensionless function expression by the atmospheric stability parameter ζ is based on the establishment of the Monin-Obukov similarity (MOS) in the ground boundary layer, and indicates that the vertical profile of the physical quantity can be uniquely expressed by ζ. For example, φ _m ≡k _kar z / u _* ∂u / ∂z in Equation 24 and Equation 25,
The Buser-Deyer equation is often used to express φ _s ≡k _kar zu _* / w's' ∂s / ∂z as a function of ζ. Several function forms of dimensionless dissipation factor have been proposed based on the observation results. For example, the empirical formula by Kader (1992) is:

If the response of atmospheric temperature and specific humidity to atmospheric stability is similar _φ εc = φ ε0 = φ εq is satisfied, Equation 30, Equation 31, Equation 32 can be solved by iterative calculations. Furthermore, when it is assumed that the scalar dimensionless dissipation rate is similar, the flux can be calculated from the CO ₂ dissipation rate in the same manner as Equations 31 and 32.

Next, calculation of the dissipation rate by the eddy correlation sensor will be described. For use in comparison with the scintillometer 1, a method for calculating the dissipation factor from the measurement value obtained by the vortex correlation sensor will be described. For the calculation of the dissipation factor, an inertial dissipation method is used which is indirectly determined by using the turbulence characteristics in the small inertia region of the spectrum (frequency representation of energy and mass fluctuation). However, since the result of Fourier transform is generally largely disturbed and may cause an error in estimating the dissipation rate, here, a method for estimating the dissipation rate based on a structure function that is a real space representation of the spectrum is used. A quadratic structure function D _ab (r) corresponding to the spectrum of TKE and scalar variation is defined by the following equation.
Here, a and b represent wind speeds and scalars such as u, θ, and q, respectively, and r is a distance in real space. When the time difference between the time series data is τ, the relationship is τu = r using the average wind speed. According to Kolmogorov's theory, when r is in the small inertia region, the quadratic structure function and the dissipation factor have the following relationship.
Here, ε and ε _s represent the dissipation rate of TKE and scalar variation, respectively. α _uu is the Kolmogorov constant (approximately 0.55), and α _ss is a universal constant called the Obukhov-Corrsin constant (approximately 0.8). The proportionality constant C _s ² is a structure function constant, and the following combinations of a and b [T: T] (temperature), [q: q] (specific humidity) and [T: q] are the structure function constants C _T ^2, respectively. , C _q ² and C _Tq . Since the second-order structure function is proportional to r ^2/3 as shown in Equation 36 and Equation 37, fitting the curve of r ^{2/3 to} the structure function in the range of r corresponding to the small inertia region. Then, the extinction rate is obtained from Equation 36 and Equation 37, and the flow proceeds to flux calculation by the dissipation method (see FIG. 11). The turbulent fluctuation smaller than the path length d _s of the ultrasonic anemometer is not measured, and only a part of the vortex larger than the measurement height is measured. Therefore, 2d _s as a fitting range in which turbulent fluctuation measurement is sufficiently guaranteed. <R <(z−d ₀ ) / 2 is selected.

Next, a method for calculating the dissipation rate using the scintillometer 1 will be described. SLS40A (Scintec AG, Germany) is a commercially available displaced-beam small aperture scintillometer that separates a 0.67 μm wavelength HeNe laser into two beams with orthogonal polarization planes and dissipates TKE and temperature fluctuations from two-path scintillation. It is an apparatus that calculates the flux and calculates the flux. The SLS series is a product that achieves both operational and measurement stability and economy by using a single light source. The refractive index fluctuation due to turbulent flow can be measured as log dispersion of the light intensity of monochromatic light propagating in the atmosphere. The log covariance B ₁₂ of the light intensity when there are two independent light sources and light receivers is expressed by the following equation (Thiemann, 1992).

Here, x is the position on the optical axis of the path length L; K = 2π / λ is the optical wave number; wave number k turbulence; C _n ² is the structure function constants of a refractive index variation, l ₀ Internal scale ( When the turbulent vortex gradually collapses into small vortices, if it becomes smaller than this, the vortex cannot be maintained due to molecular viscosity and is decomposed as heat. J ₀ and J ₁ are zero-order and first-order Bessel functions of the first order, and represent sensitivity distributions determined by the distance d between the beams and the diameter D of the receiver. B ₁ and B ₂ representing dispersion due to a single detector can be obtained by setting d = 0 in Equation 38. Φ _{n in} Equation 38 is a dimensional spectrum of refractive index fluctuation,
F (kl ₀ ) is a three-dimensional spectral model of refractive index variation according to Hill and Clifford (1978). Formula 38 assumes a low scattering condition (Rytov approximation) of σ _x ² <0.3. The integrand in the equation is a load function that represents the sensitivity distribution of the scintillometer 1, and takes a bell-shaped function that has a maximum value at the center of the path. Substituting Equation 39 into Equation 38 and integrating over x and k yields:

f _B is a function representing the attenuation of B ₁₂ with respect to increases in l ₀ , d and D. Correlation coefficient r ₁₂ = B ₁₂ / (B ₁ B ₂ ) ^1/2 = B ₁₂ / B ₁ = B ₁₂ / B ₂ is a function of l ₀ , d and D, and d and D are considered as constants. An internal scale l ₀ is obtained from the correlation coefficient r ₁₂ . If l ₀ is known, C _n ² can be obtained from B ₁ (or B ₂ ). In Scintec software, r and f _B for l ₀ are prepared as a numerical table for path length L, and l ₀ and C _n ² are calculated in real time from B ₁₂ , B ₁ and B ₂ . From the obtained l _0, the dissipation rate of turbulent kinetic energy (TKE) is obtained by the following relationship.

Here, ν (m ² s ⁻¹ ) is a kinematic viscosity coefficient of air. The temperature T (K) contributes most to the change in the refractive index, and then the change in the specific humidity q (kgkg ⁻¹ ). Since a gas such as CO ₂ exists only in a minute amount in the atmosphere, it does not affect the refractive index fluctuation. The structure function constant C _n ² of the refractive index variation is the structure function constant C _T ² (K ² m ^−2/3 ), the specific humidity variation C _q ² (m ^−2/3 ), the temperature and the specific humidity T−. It is expressed as follows by the bond structure function constant C _Tq (Km ^−2/3 ) of q (Hill et al., 1980).

Constant A _T and A _q receives the wavelength of light, air pressure, the effect of temperature and specific humidity. In the wavelength λ = 0.67 μm used in the scintillometer 1 used in the present embodiment, the following equation by Andreas (1989) is used.
Here, P (hPa) is atmospheric pressure. Since the scintillometer 1 that uses a single wavelength as the light source can obtain only one value of C _n ² , C _T ² and C _q ² cannot be uniquely determined from Equation 42. In response to this problem, Thiermann assumed that the influence of specific humidity on C _n ² is one order of magnitude smaller than air temperature, so that this can be ignored, and C _T ² was obtained from only the first term on the right side of Equation 42. The flux calculation software attached to the scintillometer 1 used in this embodiment employs this calculation method, and only the momentum flux and the sensible heat flux are calculated.

In this embodiment, a technique for obtaining a value corresponding to the scintillometer 1 is also applied to scalars such as water vapor and CO ₂ density.

First, an existing method for correcting water vapor is apparent. As is clear from Equation 42, the scintillometer 1 is sensitive to fluctuations in water vapor, and C _{T under} conditions where the water vapor flux is much larger than the sensible heat flux. The need for correction for water vapor has been pointed out because ² is overestimated (Green and Hayashi, 1998). The classical correction method for the water vapor effect is the Bowen ratio (sensible heat to latent heat ratio: H / λE) correction proposed by Wesely (1976). Since Bowen ratio correction requires w′T ′ and w′q ′, the measured value by a sensor such as a vortex correlation system is used. Moene (2003) scrutinizes the correction of Bowen ratio, and the method using the standard deviations σ _T and σ _q of temperature and specific humidity has less measurement items than the method using the flux ratio, so the error is small. It is shown that. Ie
Thus, Equation 42 can be modified as follows with respect to C _n ² and C _T ² .
By measuring the average value and variance of T and q, and the correlation coefficient R _Tq with another sensor, C _T ^{2 in} which the influence of water vapor is canceled can be calculated.

Next, as an extension to the scalar flux, if the similarity between the structure function constant ratio and the dispersion ratio of water vapor and air temperature introduced at the time of water vapor correction is established, C _T ² by the scintillometer 1 and the vortex It is considered that C _q ² can be calculated from σ _T and σ _q by the correlation sensor. Similarly, when this similarity is extended to a gas flux such as CO ₂ , the following holds.
Thus, the structure function constants C _Tq ² and C _c ^{2 of the} water vapor and CO ₂ fluctuations are obtained from the dispersion of the air temperature, water vapor density and CO ₂ density obtained by the vortex correlation sensor and the structure function constants of the air temperature fluctuations obtained by the scintillometer 1. Is required. It should be noted that this calculation method is not derived by a pure scintillometer 1 but by a combination with a vortex correlation sensor. However, there is no method for measuring fluctuations at the same scale as the scintillometer 1 at present. Dissipation rates ε _q and ε _c are obtained by substituting the structural function constants for q and c into Equation 37, and the flow proceeds to flux calculation by the dissipation method by adding ε and ε _T.

  Next, the induction of the correction method of the vortex correlation method will be described.

In the following, the flux and the dissipation rate by the scintillometer 1 are distinguished by adding a SAS suffix, and the value corresponding to EC is appended by an EC suffix. Here, from the flux by taking the ratio of the flux in Equation 16 to Equation 18 Equation 47 to next equation 49, the scintillometer 1 and the ratio of the dissipation rate eddy correlation sensor (ε _iSAS / ε _iEC) eddy correlation method, scintillometer A flux corresponding to a spatial scale of 1 is obtained. The subscript _i in ε _i means that any subscript of T, q, or c corresponds to what extinction rate is expressed.

Here, it is known that ε _TSAS / ε _TEC changes in response to turbulence intensity (vertical wind deviation standardized by horizontal wind speed: σ _w / u) and wind direction. Further, the change in the wind direction with respect to the path of the scintillometer 1 affects the spatial scale of the scintillometer 1. That is, the change in the wind direction, that is, the angle of attack of the wind with respect to the path, changes the spatial scale (measurement range). However, the wind direction is considered to contribute the most to changes in the spatial scale, although wind speed and atmospheric stability are also affected. Therefore, epsilon _iSAS / epsilon _iEC that expressed by a function that the spatial scale of the turbulence intensity and scintillometer 1 and variable, and can be calculated flux corresponding from the measured value by the eddy correlation sensor spatial scale of scintillometer 1 Become.

The flux calculation corresponding to the spatial scale of the scintillometer 1 is as follows. First, the turbulent flow signal detected by the flux measurement sensor originates from the windward area, and its size (source area) varies with the wind speed, the atmospheric stability, and the measurement path length of the sensor. In order to quantitatively analyze the difference in spatial scale between the scintillometer 1 and the vortex correlation, the source area of the scintillometer 1 is calculated as follows by applying the footprint model of Kormann and Meixner (2001). That is, the footprint F (x) is a function representing the contribution to the flux at a position x (m) away in the windward direction, and the measured wind velocity u (ms ⁻¹ ) and friction velocity u _* (ms ⁻¹ ) are measured. Z (m) and the atmospheric stability parameter ζ are variables. Further, according to the calculation example of De Bruin et al. (2002), the sensitivity area D of the source area (A _SAS ) of the scintillometer 1 in the crosswind direction y (m) represented by F (x) and the path length of the sensor. It is obtained from the area surrounded by the isoline drawn by the product of x, y) (see FIG. 6). Here, D (x, y) is a function indicating that the sensitivity of the scintillometer 1 to detect a signal is the highest in the center portion of the path. In this embodiment, Wang et al. (1978) Sensitivity function D (x, y)
[Equation 50]
D (x, y) = (y / L _p (1−y / L _p )) ^5/6
Is standardized and used. Where L _p is the path length of the scintillometer 1. Further, the source area of the scintillometer 1 that changes depending on the wind direction is expressed as follows using the wind direction angle (θ) from the direction orthogonal to the path.
The ratio RA (= A _SAS / A _EC ) of the source area of the scintillometer 1 and the vortex correlation sensor is 1 when the wind direction is parallel to the path of the scintillometer 1 (θ = 90 °, 270 °) and increases when the wind direction is orthogonal. (See FIG. 7).

Next, the area average flux is calculated. Here, first, the shape of the function f (σ _w / u, RA) having the source area ratio RA and the turbulent flow intensity σ _w / u obtained as above as variables is defined, and ε _TSAS / ε _TEC obtained for a certain period of time. The coefficient of the function is determined by regression with respect to the actual measurement value.
Next, determine the RA directing flux evaluation value considered appropriate, assuming that the relationship is established for all the wind direction, obtaining the sensible heat flux H _SAS from Equation 47. The flux thus obtained is defined as the area average flux. Since the area average flux has the spatial scale of the scintillometer 1 with respect to all wind directions, and the flux evaluation can be performed only by the measurement value by the vortex correlation sensor, the flux under the weather conditions that the scintillometer 1 cannot measure can also be evaluated. Furthermore, since water vapor and CO ₂ are thought to be transported on the same principle as heat,
Assuming that, the area average flux can be similarly calculated for latent heat and CO ₂ (see FIG. 8).

  The above-described embodiment is an example of a preferred embodiment of the present invention, but is not limited thereto, and various modifications can be made without departing from the gist of the present invention.

Through 2002-2004, flux measurements including CO ₂ were conducted in a 54-year-old deciduous broad-leaved forest in the eastern foot of Asama, Chubu district. Here, “regional average flux calculation” of calculating the region average flux is performed by correcting the measured value of the vortex correlation method using the function determined from the spatial scale of the scintillometer 1, and the calculated region average flux Confirmed that the heat balance was correctly evaluated. It was also compared with the canopy photosynthesis amount determined from photosynthesis of the measurement result and CO ₂ flux. While the vortex correlation method underestimates about CO _{2 with} respect to the photosynthetic potential of forests calculated based on the photosynthetic characteristics and leaf distribution of the first-class Dake birch, the calculated value by the area average flux is good. From the agreement, it was found that the reliability of the evaluation value was improved by the area average flux even in the CO ₂ exchange amount. Furthermore, as a result of evaluating the CO ₂ absorption amount of the forest, it was almost the same as the evaluation result by the past eddy correlation method obtained in a place where the weather conditions and the tree species were similar. The above will be described as Example 1 below.

Here, 28m high steel scaffolding towers are placed 86m away from east and west, the vortex correlation method is performed at the top of the east side, and the area average flux is measured between the two towers. The measurement path of the scintillometer 1 was used. In the eddy correlation method, sensible heat, latent heat and CO ₂ flux were measured. A 30-minute block average of 10 Hz sampling time series was used for flux calculation. The NEE was calculated by using the CO ₂ density fluctuation measured by a closed-path CO ₂ analyzer (LI-7000, Licor USA) which is not easily affected by rainfall or the like and has little drift in measurement values. An air sample at the top of the tower was sucked into the observation building at the base of the tower via a decabon tube 40m having an inner diameter of 4 mm at 9.0 Lmin ⁻¹ , and a part thereof was introduced into the CO ₂ analyzer at 2.0 Lmin ⁻¹ . The time delay for the sample air to reach the analyzer from the tower top was about 9 seconds. It has been reported that the problem that the high-frequency component of the signal deteriorates due to the suction of the sample air hardly needs to be considered in forests where the scale of turbulence is large. In this test forest as well, it was confirmed that the closed-path system measured CO ₂ flux with high accuracy compared to the open-path (open9-path) CO ₂ analyzer. In order to determine the function f (σ _w / u, RA) for calculating the area average flux, regression was performed using the measured values on a clear day from June to November 2002 when the measured values of the scintillometer 1 were obtained. Calculation was performed.

Next, the missing data interpolation will be described. In order to calculate NEE (Net Ecosystem Exchange) from flux, it is necessary to compensate for missing measurements caused by various reasons such as instrument failure, maintenance, power outage, and to obtain a continuous time series. . Therefore, short-term missing within 1.5 hours performs linear interpolation, using a high meteorological parameters correlated with the CO ₂ flux otherwise, previously obtained temperature - perform interpolation from photosynthesis relationship - respiration and light (See FIG. 9). When measured values could not be obtained, the daily average solar radiation and temperature of this test forest were estimated from AMeDAS data.

Moreover, in order to compare with the flux measurement value, the leaf area and the amount of photosynthetics of individual leaves were measured. The range of 60 m (N−S) × 140 m (E−W) centered on the tower was divided into 10 m × 10 m sections, which were used as the criteria for acquiring all biological data. Here, in order to clarify the seasonal variation in leaf area, 19 litter collection devices (litter traps) with an opening of 1m x 1m were arranged in a staggered pattern at a height of 1.3m from the forest floor. Litter leaves and twigs were periodically collected during the fall season in early November. In addition, the amount of fallen leaves was investigated using three litter traps that showed the average amount of fallen leaves from June to the beginning of September, taking into account the fallen leaves and branches other than the fallen leaves due to strong winds. The leaf area of the collected deciduous sample was determined from the linear regression equation of dry weight and area of deciduous leaf obtained in advance, and the LAI (Leaf Area Index) of the survey area was determined from the total leaf area. The measurement of photosynthetic amount is a portable photosynthetic transpiration measuring device (LI-6400) for any individual leaf at an altitude of 9 m in the birch (Betula ermanii Cham.) Adjacent to the north side of the tower where the eddy correlation method was performed. , Licor USA). Bake birch is known to be a priority species, accounting for 44.3% of the above-ground (branch + trunk) carbon content of the trees that make up the test forest, based on prior tree surveys. This apparatus has a small leaf chamber, and can control the temperature, humidity and CO ₂ concentration in the chamber by following the external environment. The birch leaves are fully saturated with light intensity in 7 levels (2000, 1500, 1000, 500, 200, 100, 0 μmolm ^-2 s ^-1 ), and the photosynthetic capacity is considered to be the highest from July to September. The reaction of the amount of photosynthesis with respect to the light intensity over a period (photo-photosynthesis reaction) was examined.

As a result, the following results were obtained. First, the annual flux data missing rate from 2002 to 2004 was 18%, 5%, and 3%, respectively. The high missing rate in 2002 was due to the power outage caused by lightning strikes, and there were other causes such as maintenance, bad weather and equipment failure. Here, FLUXNET (Falge et al., 2002), Asia Flux (Asia Flux Steering Committee, 2003) and others have proposed quality control (QC) of turbulent data for reliable flux measurement. Yes. In QC, where turbulence is not feasible, the vortex correlation method is considered unreliable, and the standard is set for the friction speed u _* and the continuity of the time series data. And For this reason, the missing rate of standard forest flux sites is said to be around 30%. However, in the present embodiment, comparison is made with the data of the scintillometer 1 including the measured values in a turbulent flow poor state or a large convection state in which the reliability of the vortex correlation method is considered to be low, and based on the result, the area average flux Therefore, QC was not applied and all data was used.

Subsequently, the correction function was determined. Here, the dissipation rate ratio ε _TSAS / ε _TEC of temperature fluctuations by the scintillometer 1 and the vortex correlation sensor is higher than the SAS at the neutral time when the mechanical turbulence intensity (σ _w / u) is large. , (Σ _w / u) is small and unstable, the SAS dissipation rate is larger than EC (see FIGS. 10A and 10B). Since the dissipation factor and the absolute value of the flux are proportional to each other as shown in Equation 47, this result suggests that the vortex correlation sensor underestimates the flux under conditions where convection prevails or in stratified atmospheric conditions. Kanda et al. (2004 ). Further, as the source area of the scintillometer 1 becomes larger than that of the vortex correlation sensor (RA becomes larger), the upper end of dispersion of ε _TSAS / ε _TEC tends to increase (FIG. 10 (c), FIG. 10 ( d)). Since the average value of signals originating from a wider area can be detected by increasing the spatial scale of the sensor, it is possible to capture atmospheric turbulence with a large spatial scale, resulting in the above σ _w / It was thought that the relationship of ε _TSAS / ε _{TEC to} u was emphasized.

Therefore, the asymptotic value in the region where f (σ _w / u, RA) in FIGS. 10A and 10B is large is considered as an offset, and the above relationship is expressed by an exponential decay function (Equation 54, Equation 55). (Refer to the right column of FIG. 10). Here, the coefficients a ₁ , a ₂ , and a ₃ in the formula are determined by the least square method for daytime and nighttime. The reason why the coefficients differ between night and day is that the footprint model used to calculate the source area is expressed by different functions at night and day.
[Formula 54]
f (σ _w / u, RA) = y ₀ + a ₁ exp [− (σ _w / u−x ₀ ) / a ₂ ]
[Equation 55]
x ₀ = (RA-1) / a ₃

FIG. 12 shows a comparison between the function f (σ _w / u, RA) and the extinction rate ratio ε _TSAS / ε _TEC . The dispersion at night (Fig. 12 (b)) is larger than that during the day (Fig. 12 (a)) because the calculation accuracy of the temperature fluctuation extinction rate ε _TEC by the eddy correlation sensor is lowered in a poor turbulent state. It was thought to be the cause.

Next, in order to calculate the optimum RA that is considered to obtain the area average flux, the relationship between the annual heat flux balance and RA was examined (FIG. 13). Imbalance rate I as RA increases for both daytime instability and nighttime stability
[Formula 56]
I = (H + λE−Rn) / Rn
Was close to 0, and the balance of heat balance was seen. In particular, the effect was strong when the night was stable. In addition to latent heat and sensible heat, heat stored in the canopy atmosphere and plant bodies and underground heat transfer (stored heat in the canopy) are involved in the heat balance, but the value is at most 10 tens Wm- ² . In addition, from the viewpoint of preventing the increase of errors due to extrapolation with almost no measured value exceeding RA of 20, an RA indicating an imbalance rate = −0.3 was obtained (10.8 at night, 23.1 at daytime). As a result, f (σ _w / u, RA) was expressed as a variable of σ _w / u, and a region average flux was obtained based on the measured value of the vortex correlation sensor.

Here, the characteristics of the area average flux will be described. The region average flux shows a larger value than the flux by the vortex correlation method in a region where σ _w / u is small, and shows a smaller value than the vortex correlation method in a region where σ _w / u is large. As a result, nighttime release and daytime absorption peaks are greatly estimated, and the amplitude of daily fluctuations increases. FIG. 14 shows an example of measurement of heat and CO ₂ flux. In the heat flux, the heat balance was improved by the area average flux, but under conditions where the atmosphere was neutral, the area average flux was lower than the vortex correlation method (see FIG. 14A). In addition, there were some cases where the effect of improving the heat balance was small. It is considered that there may be a time zone where the contribution of heat storage within the canopy is large. F _C Similarly tendency emission-absorption peak is enlarged is observed also (see B in FIG. 14), and canopy photosynthesis amount calculated on the basis of F _C, will also vary community respiration.

Furthermore, a comparison was made between the area average flux and soil respiration. Soil respiration, in which CO ₂ is released by the respiration of animals and roots in the soil and decomposition of organic matter by microorganisms, and respiration by plant bodies on the ground, add up to CO ₂ release from the forest. Call quantity. Information corresponding to the community respiration rate can be obtained from CO ₂ observed on the forest. On the other hand, soil respiration can be measured by a chamber method or the like. Therefore, in order to verify the validity of the CO ₂ flux, the community respiration rate and the soil respiration rate were compared.

First, canopy photosynthetic amount _{A g,} between the canopy respiration rate _{R e} and NEE [number 57]
NEE = −A _g + R _e
There is a relationship. Since A _g = 0 at night, NEE is the community respiration rate. The canopy respiration rate is highly correlated with temperature, and Equation 58 is used to parameterize the canopy respiration rate with temperature (Lloyd and Taylor, 1969).

Here, T (K) is a temperature, T ₀ (K) is an arbitrarily determined reference temperature, and R ₀ represents a CO ₂ flux at T ₀ (K). _{Also, Q 10} represents a temperature coefficient, relative to CO ₂ flux at the reference temperature, the CO ₂ flux ratio when the reference temperature was 10K rises. According to the relationship between air temperature and community respiration rate (2004) (see FIG. 15), data of σ _w /u>0.3 or more are shown for ease of viewing. Therefore, in order to analyze the community respiration rate of the forest, parameters Q ₁₀ and R ₀ were determined from the relationship between the data of CO ₂ flux (F _C ) at night and the temperature obtained from SAS and EC (see Table 1). ). Here, in consideration of snow accumulation during the winter season, the temperature at a height of 3 m above ground was used for T instead of the ground temperature. The community respiration rate _Re includes the soil respiration rate (respiration by organisms in the soil and respiration of the roots) and the respiration rate of the above-ground part of the plant.

Although Q ₁₀ by the parameter Q ₁₀ and the region average flux canopy respiration based on the eddy correlation method showed a different tendency in differences in annual both clear trend between the average temperature was observed. Next, the community respiration rate was estimated from the daytime air temperature and Formula 58, assuming that the relationship between the temperature and the respiration rate at night also holds in the daytime. Missing interpolation was performed from Equation 57 using the relationship between the temperature and the group respiration rate thus obtained and the group photosynthesis amount described later. Previously, the yearly soil respiration rate of this test forest evaluated using the soil respiration chamber in 2002 was 6.6 tCha ⁻¹ yr ⁻¹ , while the community respiration rate based on the eddy correlation method and the regional average flux was In each of 7.4 tCha ⁻¹ yr ⁻¹ and 7.3 tCha ⁻¹ yr ⁻¹ (see Table 2), there was no clear difference between the flux measurement methods. On the other hand, in 2003 and 2004, with respect to 3.9tCha ^-1 yr ^-1 and 7.2tCha ^-1 yr ^-1 eddy correlation method, area average flux 6.7tCha ^-1 yr ^-1 and 8.9tCha ^-1 yr ^{- A} large value of ¹ was shown. As for the community respiration rate, since the respiration rate above the plant is added to the soil respiration rate, it can be said that the evaluation value by the flux exceeds the chamber method. However, the dispersion of the group respiration rate with respect to the temperature was large (see FIG. 15), and the error at the time of curve fitting was large. The community respiration rate calculated from the area average flux was consistent with the soil respiration chamber measurement, and it was confirmed that the community respiration rate was correctly evaluated by the area average flux.

Subsequently, a comparison was made between the area average flux and the amount of community photosynthesis. In this example, in order to verify the validity of the CO ₂ flux, a comparison with the amount of community photosynthesis was performed. The amount of community photosynthesis evaluated by the CO ₂ flux represents the amount of photosynthesis in the entire forest. On the other hand, the amount of leaf photosynthesis can be measured by a leaf chamber. However, forests have a complicated three-dimensional structure, and there are several times as many leaves per forest floor area. For this reason, in order to compare the photosynthesis amount measured in a single leaf with the community photosynthesis amount, it is necessary to extend the measurement result of the single leaf to the community. The amount of canopy photosynthesis can be evaluated from the maximum amount of photosynthesis and the leaf area of the leaf located at the top end of the canopy.

First of all, we performed lametallization of the amount of canopy photosynthesis. The difference between NEE and the group respiration rate indicates the group photosynthesis amount (A _{g in} Formula 59) by the anabolic action. Canopy photosynthesis (μmolm ⁻² s ⁻¹ ) P _m represents the maximum photosynthesis (μmolm ⁻² s ⁻¹ ), and α represents the initial photosynthesis efficiency. A density expression (mgCO ₂ m ⁻² s ⁻¹ ) multiplied by the molecular weight of CO ₂ is also used as a unit of photosynthesis. It was determined parameter fit the formula 59 in the plot of F _C minus the respiration found by replacing the daytime temperature to the amount of incident light and formulas 58 according to Equation 57. Thus to calculate the A _g and R _e from the parameter and the incident light quantity and temperature obtained were missing interpolation.

Next, the calculation result of the leaf area for converting the photosynthesis amount of individual leaves into the canopy photosynthesis amount P _can is shown. The leaf area index (LAI (Leaf Area Index)) accumulated from the amount of fallen leaves was 5.6, 6.3, and 6.8 from 2002 to 2004, respectively, showing an increasing trend.

  Moreover, when the change of the leaf area index | exponent shown in FIG. 16 is seen, defoliation is continuing also in the period from June when leaf expansion was completed to September when defoliation starts. Bake birch, the preferred species of the test forest, was found to extend for the second time after the spring leaves. If leaf expansion by secondary extension continues during the period from June to September, if the amount of fallen leaves during that period is included in LAI, LAI will be overestimated. As values close to the actual leaf area per forest floor area, the values in early September immediately before the start of defoliation were 5.4, 5.6, and 5.8, respectively. Did not change. The leaf litter weight of Dake birch in total leaf litter weight was about 41%, 38.2%, 42.1%, 42.7%, showing a value close to 44.3% of the above-ground carbon content. In terms of leaf area, it was about 31% at 29.1%, 31.9%, and 32.5%.

Subsequently, the amount of community photosynthesis was calculated. _{P m} by the area average flux, the maximum value in July and August, the value was ^{_{1.7,1.9,1.8 (mgCO 2 m -2 s -1}} ). On the other hand, the average values of the maximum photosynthetic amount of the birch leaves measured by the leaf chamber were 0.8 ± 0.2, 0.9 ± 0.1, and 0.9 ± 0.1 (mgCO ₂ m ⁻² s ⁻¹ ), respectively. Assuming that the maximum photosynthesis amount of the birch leaves obtained in the leaf chamber is the maximum photosynthesis amount at the top of the _canopy, the amount of light _Iink absorbed by the leaf group at the position of the integrated leaf area F is expressed as follows according to Beer's law. Ru
Here, I ₀ is the incident light at the top of the canopy, K is an extinction constant within the canopy determined by the leaf angle, etc., and is 0.5. a is the leaf transmittance '(= 0.14). Next, the photosynthesis amount P with respect to the incident light _Iink can be calculated by setting PAR in Equation 59 to _Iink . FIG. 17 shows the amount of light absorption I _ink of the leaf group with respect to the integrated leaf area and the amount of photosynthesis P obtained by substituting I _ink into Equation 59. Then, by integrating P by F, the photosynthesis amount P _{can of the} entire community _can be calculated.
Plants in the swarm layer generally have lower photosynthetic ability (smaller maximum photosynthesis amount) than the upper end of the canopy, and it was thought that the amount of swarm photosynthesis was actually smaller. In addition, this model does not take into account the nap phenomenon caused by pore blockage due to lack of moisture. Therefore, it is appropriate to consider this calculation result as the potential of community photosynthesis. It is considered that the instantaneous value of the flux is distributed below the potential of canopy photosynthesis, for example, the moisture state in the plant body affects the opening and closing of the pores, and the exchange amount of CO ₂ changes depending on the turbulent state. As shown in FIG. 18, top and _{P CAN} distribution of _{A g} by the region mean flux substantially overlap. This can be said that the area average flux well represents the photosynthetic characteristics of the forest.

In addition, FIG. 19 shows the community photosynthesis potential derived from the potential and flux of community photosynthesis at the maximum photosynthesis effective radiation amount 2000 μmolm ⁻² s ⁻¹ observed in this test forest. The underestimation of the eddy correlation method is also apparent in CO ₂ flux, and the maximum photosynthesis amount by the area average flux is about 1.5 times the maximum photosynthesis amount by EC, approaching the theoretical potential of community photosynthesis. As a result, it was shown that forest photosynthesis was correctly evaluated by the area average flux.

Subsequently, the validity of the area average flux was evaluated. The present inventor previously determined that the flux is properly measured for CO ₂ measured by the same principle based on the matching of the heat balance equation by the application of the scintillometer 1 as the basis of the validity of the flux measurement. In this example, the results of comparing the community respiration rate and community photosynthesis amount derived from the area average flux with the community photosynthesis amount based on the community respiration rate and the photosynthetic characteristics of individual leaves, It was in good agreement with the measured values of the amount and the amount of photosynthesis. Thus, the validity of CO ₂ Evaluation by region average flux was shown. In addition, using the heat flux balance as a criterion, the methodology for evaluating the validity of flux measurement was supported. The results in the present example as described above capture the phenomenon peculiar to this test site. However, even at existing flux measurement sites and new measurement sites, site-specific observations are made by combining the vortex correlation method and the scintillometer 1. It was considered possible to calculate the area average flux.

Furthermore, forest CO ₂ absorption based on the area average flux was evaluated. As shown in Table 2, NEE showed large annual fluctuations. In addition to the meteorological conditions such as precipitation, solar radiation, temperature and water vapor saturation, nitrogen load and pests were considered to be combined. For example, in 2004, the defoliation period was earlier than in other years, and the production volume was high despite the decrease in leaf volume during September. NEE by the area average flux exceeded the eddy correlation method (see A in FIG. 20), and as a result, the community respiration rate and the maximum community photosynthesis amount were also greatly estimated (see B in FIG. 20). The average value of the NEE of 3 years 180gCm ^-2 yr ^-1 in eddy correlation method, in the region average flux was about 1.5 times the 266gCm ^-2 yr ^-1. Previously, the flux measurement by eddy correlation method carried out from 1998 to 1999 in a 40-year-old dake birch forest with similar conditions to this test forest was 163 gCm ^-2 yr ^-1, which is close to the result of this example. there were. Moreover, when the flux was corrected so as to eliminate the heat balance imbalance, it was 214 gCm ^-2 yr ^-1 , which was about 1.3 times. Ecosystem respiration rate (R _e) and Comparing the GPP represented by the sum of net ecosystem production (NEE) (Gross primary production) , 875gCm -2 yr -1 in the conventional method, after the correction 1146gCm ^-2 yr ^-1 was 1156gCm ^-2 yr ^-1 equivalent result by 3 years average 811gCm ^-2 yr ^-1 and the area average flux of this test forest. Although this uses a correction method different from that of the present example, underestimation by the vortex correlation method is considered to have occurred as in the result of this example.

As described above, in this embodiment, for the purpose of evaluating the CO ₂ absorption amount in the forest with high accuracy, by correcting the measurement value of the vortex correlation method using the function determined from the spatial scale of the scintillometer 1. The area average flux was calculated. Here, we verified the certainty of the area average flux by comparing the area average flux adjusted to match the heat balance with the community respiration and forest photosynthesis in the forest. As a result, the community respiration rate calculated based on the area average flux was consistent with the evaluation result of soil respiration rate by the chamber method developed previously. In addition, the amount of canopy photosynthesis calculated from the eddy correlation method, which is the conventional method, is significantly lower than the amount of canopy photosynthesis derived from the measurement results of the leaves of the birch, while the amount of canopy photosynthesis by the region average flux is well It became clear to correspond. From the above, it was found that the method developed by the inventor as an area average flux evaluation method improves the reliability of the conventional eddy correlation method in the heat balance in forests and the CO ₂ balance in communities. For example, the three-year average value of CO ₂ absorption evaluated by the conventional method is 180 gCm ^-2 yr ^-1 , which is equivalent to the value of deciduous broad-leaved forest under similar location conditions. On the other hand, the average value of the amount of CO ₂ absorbed by the forest estimated using the measurement method according to the present invention over three years was 266 gCm ⁻² yr ⁻¹ , which was 1.5 times the evaluation value by the vortex correlation method.

Dissipation rate ratio and wind direction between scintillometer 1 (also referred to as SC) and vortex correlation sensor (SC) with respect to wind speed, turbulence intensity, TKE (turbulent kinetic energy), temperature, specific humidity, and CO ₂ An experiment was conducted on the relationship (see FIG. 21). The angle in the figure represents the azimuth and the north is 0 °. The larger the value in the graph, the larger the ratio of the extinction rate by the scintillometer 1. For example, in terms of wind speed, the north wind tends to increase the ratio of the extinction rate by the scintillometer 1.

For each of the friction velocity u _* , the sensible heat flux H, the latent heat flux λE, and the CO ₂ flux fc, the flux calculation values of the scintillometer 1 and the vortex correlation sensor were compared (see FIG. 22). In the figure, data of a west wind parallel to the path of the scintillometer 1 indicated by a square symbol and data of a north wind perpendicular to the path of the scintillometer 1 are indicated by a round symbol. In each period, as shown in the figure, the ratio between the measurement result by the scintillometer 1 and the measurement result by the vortex correlation sensor tended to increase the measurement result by the scintillometer 1.

It was confirmed that the heat balance imbalance was improved by displaying the imbalance rate in the non-leafing period in each case of the scintillometer 1 and the vortex correlation sensor (see FIG. 23). The imbalance ratio RI and the heat balance equation can be obtained by the equations shown in the figure. As a result, the CO ₂ flux measurement method according to the present invention was found to improve the imbalance rate over the conventional method (see FIG. 23).

It is the schematic for demonstrating the definition and measuring method of the flux in the flux measuring method of this invention. It is a figure which shows the measurement principle of a scintillometer. It is a figure which shows the flow of the data processing of the flux calculation in this invention. It is a figure for demonstrating the difference in the measurement area | region by the path length of a scintillometer also in the case of a vortex correlation sensor. It is a diagram illustrating the concept of CO ₂ exchange forest ecosystems. It is a figure which shows a footprint and source area when a wind direction is orthogonal to a path | pass. It is a figure which shows the relationship between source area ratio RA and wind direction (theta). It is a figure which shows the procedure of area | region average flux calculation. It is a figure which shows the flow of interpolation of missing measurement data. The relationship _{between the} turbulent intensity σ _w / u and the source area ratio RA with respect to ε _TSAS / ε _TEC and the relationship between the correction relationship f (σ _w / u, RA), the turbulent intensity σ _w / u and the source area ratio RA are shown. FIG. It is a figure which shows the content (June 10, 2002) of the example of a structure function and the dissipation rate determination by curve fitting. The figure which shows the comparison of (epsilon) _TSAS / (epsilon) _TEC and f ((sigma) _w / u, RA), a) represents daytime, b) represents nighttime. It is a figure which shows the relationship between source area ratio RA and imbalance rate I (however, I is an average value for 3 years). Is a diagram showing an eddy correlation method EC and area average flux SAS relationship in diurnal heat fluxes and CO ₂ flux. It is a figure which shows the relationship (2004) of temperature and a community respiration rate, Comprising: Only the condition part of (sigma) _{w /} u> 0.3 is represented for easiness to see. It is a figure which shows the change of LAI of the leaf fall period based on litter integration. It is a figure which shows the profile with respect to the integrated leaf area (F) of the light-absorption amount ₁₀ and photosynthesis amount (P) of a leaf group. It is a figure which shows the relationship (July, 2004) with the potential of canopy photosynthesis and the amount of incident photosynthesis (PAR). It is a figure which shows the comparison of a canopy photosynthetic potential and a canopy photosynthesis amount (PAR = 2000micromol-2s ^{- 1} ). It is a diagram showing a comparison of net ecosystem exchange amount by the area average flux and eddy correlation flux (NEE) and canopy maximum photosynthesis (A _g). It is a figure which shows the relationship between the dissipation factor ratio between a scintillometer (SC) and a vortex correlation sensor (EC), and a wind direction about each element. It is a figure which shows the comparison result of the flux calculation value by a scintillometer and a vortex correlation sensor regarding each of friction speed u _* , sensible heat flux H, flux λE of latent heat, and CO ₂ flux fc. It is a figure which shows that the imbalance rate in the non-leafing stage in each case of the scintillometer and the vortex correlation sensor was improved.

1 scintillometer 2 transmitter 3 receiver

Claims (1)

In the CO ₂ flux measurement method for measuring the CO ₂ absorption of the whole forest, the ratio ε _TSAS / ε _{TEC of the} rate of extinction of temperature fluctuation when using a scintillometer and using the vortex correlation method, when using a scintillometer Based on the fact that the correlation between the measurement range A _SAS and the ratio RA (= A _SAS / A _EC ) of the measurement range A _EC with the eddy correlation sensor and the atmospheric turbulence intensity σ _w / u is high, Relational expression with these two elements RA (= A _SAS / A _EC ) and σ _w / u as parameters
Is prepared in advance based on the measurement results in a certain period actually obtained using the scintillometer and the vortex correlation sensor, and all the values are obtained by giving a constant value to the ratio RA of the measurement range in Equation 1. The scintillometer's wide measurement range with respect to the wind direction is applied, and then the atmospheric turbulence intensity σ _w / u is measured by the eddy correlation sensor actually installed above the forest, and the following relational expression
The sensible heat flux H _EC obtained by the vortex correlation method is corrected to the sensible heat flux H _SAS corresponding to that obtained by the scintillometer, and the latent heat flux λE _EC and CO ₂ flux F obtained by the vortex correlation method are corrected. Regarding _cEC , the following relational expression
The basis, corrected to latent heat flux RamudaE _SAS and CO ₂ flux F _CSAS corresponds to a case where determined by scintillometer, CO ₂ flux measurement of forest, characterized in that to obtain the same measurement result as that measured by the scintillometer Law.