CC BY
18
ОСНОВАННЫЙ НА ГИМС-ТЕХНОЛОГИИ МЕТОД МИКРОВОЛНОВОГО МОНИТОРИНГА РАСТИТЕЛЬНОСТИ
СП. ГОЛОВАЧЕВ, ИРЭ, В.Ф. КРАПИВИН, ИРЭ, А. А. ЧУХЛАНЦЕВ, ИРЭ, А.М. ШУТКО, ИРЭ
В работе проведен анализ состояния исследований в области дистанционного СВЧ радиометрического зондирования почвы и растительности и намечены некоторые перспективы дальнейших исследований в данном направлении. Обобщены результаты теоретических и экспериментальных работ по разработке и верификации моделей СВЧ излучения земной поверхности при наличии растительного покрова. Рассмотрены связь параметров моделей с характеристиками почвы и растительности и определение этих характеристик по данным СВЧ радиометрического зондирования.
Обсуждаются современные достижения в области изучения глобального круговорота углерода с применением методов математического моделирования. Анализируются подходы к синтезу моделей глобального круговорота углерода. Рассматриваются проблемы изменения глобального климата в связи с ростом кон-
центрации углекислого газа в атмосфере и оцениваются вклады наземной биоты и Мирового океана в эти изменения. Предлагается новая концепция глобальных изменений окружающей среды, основанная на интеграции знаний в рамках системы адаптивного геоинформационного мониторинга, в которой микроволновая радиометрия обеспечивает оценку входных характеристик моделей почвенно-растительных формаций с высокой точностью.
Основой синтеза эффективной технологии мониторинга растительности является формула ГИМС = ГИС+модель, где в качестве составляющих выступают компьютерная технология картирования, методика имитационного моделирования, алгоритмы реконструкции двухмерных изображений по отрывочным в пространстве и фрагментарным во времени наблюдениям, эволюционное моделирование и управление базами данных.
MICROWAVE RADIOMETRY RELATED TO THE SOIL-VEGETATION
SYSTEM CONTROL
S.P. GOLOVACHEV, IRE, V.F. KRAPIVIN, IRE, A.A. CHUKHLANTSEV, IRE, A.M. SHUTKO, IRE
The solution of the majority of applied problems within global carbon cycle is difficult for the reason that effective methods of control of the soil-vegetation system (SVS) are insufficiently developed. During the last few years, the global carbon cycle problem has acquired a special significance because of the greenhouse effect. Knowledge of the state of the SVS al-
lows one to have a real picture of the spatial distribution of the carbon sinks and sources on the Earth's surface.
As is well known, among the types of remote sensing techniques, microwave radiome-try proves effective for observations of SVS environmental parameters. However, these observations are a function of different environ-
mental conditions mainly depending on the SVS type. That is why it is necessary to develop data processing methods for microwave monitoring that allow the reconstruction of the SVS characteristics with consideration of the vegetation types and that provide the possibility of synthesizing their spatial distribution.
As is noted in [1], the problem of microwave remote sensing of the vegetation cover requires the study of the attenuation of electromagnetic waves (EMW) within the vegetation layer. The solution of the problems arising here is made possible by the combination of experimental and theoretical studies. The vegetation cover is characterized by varied geometry and additional parameters. Therefore, a knowledge of the radiative characteristics of the SVS as functions of time and spatial coordinates can be acquired by means of a combination of on-site measurements and models. General aspects of such an approach have been considered by many authors [1-6].
One prospective approach to the solution of the problems arising here is GIMS-technology (GIMS = GIS+Model) [4]. A combination of an environmental acquisition system, a model of the functioning of the typical geoecosystem, a computer cartography system, and a means of artificial intelligence will result in the creation of a geoinformation monitoring system for the typical natural element that is capable of solving the many tasks arising in the microwave radiometry of the global vegetation cover. The GIMS-based approach, in the framework of the EMW attenuation by the vegetation canopies, allows the synthesis of a knowledge base that establishes the relationships between the experiments, algorithms and models. The links between these areas have an adaptive character giving an optimal strategy for experimental design and model structure.
The reliability of the assessment of the role of CO2 in the greenhouse effect formation depends on a detailed consideration of the global biogeochemical carbon cycle dynamics in the models and on the accuracy of the assessment of its characteristics. The accuracy of estimates of carbon fluxes in the terrestrial part
of the biosphere is the function of a detailed quantization of the SVS types and accuracy of the parameterization of the biocenotic processes. In this connection the world map of SVS is given in [4]. An exemplary scheme of carbon flux in this model is characterized in Figure 1 and Table.
The vegetation cover parameters change during the year depending on the weather situation . The specific biomass Qi of the ith type of vegetation at time t can be parametrized by means of the following equation:
dQ/dt = R - M - E
/ i i it
where Ri is the biomass productivity and Mi and Ei are the biomass losses at the expense of withdrawal and transpiration, respectively.
The function Mi(^,i^,t) reflects the set of natural MNi and anthropogenic MAi processes leading to vegetation biomass losses (Mi = MNi + MM ):
Mt (<y, t) = (t )Qi (<,y, t), where < y are the latitude and longitude, respectively.
The flux Ei is calculated by means of the formula [8]:
T7 / ^^ p e V c / ^a \
Ei (< y,t) = —-r— ,
where e' (Tc) is the saturated vapor pressure inside the canopy foliage (in units of Pa), ea is the vapor pressure in the canopy air space (Pa), rc is canopy resistance (sm-1), rb is the bulk leaf boundary layer resistance of the canopy (sm-1), p is air density (kg-m-3), cp is the air specific heat (J-kg"1-K"1) and y p is the psychrometric constant (Pa-K-1).
The actual plant productivity is approximated as follows: Rt = S-c (1 + tiT • AT/100) exp (-£ / Q) min { A A &},
where dT and are indices of dependence of
production on the temperature variation AT and biomass Qi, respectively; A is the index of
limitation of production by the factor Z: e = illumination, Z = pollution, W = soil moisture, B = nutrient salts of the soil and c = atmosphere CO2 concentration. The S'z functions ac-
tually used in the framework of real situations are calculated based on existing or preliminary receiving data. Thus, the role played by the atmospheric CO2 concentration CA in photosynthesis is described by the relation 8] = bCA /(CA + C0.5),where C0.5 is the CO2 concentration for which 8'c = b /2 . The influence of
the solar radiation intensity e(^,y,t) on photosynthesis is parametrized by the relation 8 =8* exp(1 -88), where 88= e/e' and e8 is the optimal illuminance for ith type of plant. A more detailed description of the correlations between the biocenotic processes is given in [4, 6].
Figure 1 represents the World Ocean as a complex hierarchic system. Modelling the organic carbon cycle in this system and atmosphere-ocean exchange processes are described by the 3-D model of oceanic ecosystem. Different items of this model were described in the papers [4,6,9,10].
Following Nitu et al. [11], by the depth z the ocean was divided into four basic layers: photic to well-heated depths (Qu = U[z, ,zj+1], z0 = 0; i=0,1,...,m-1); intermediate (QP = U[zi ,zm], i = m,...,m+n-1); deep (Ql = U[z, ,zm], i = m+n,..., m+n+l-1), and bottom QF. By the hydro-physico-ecological characteristics the layer Qu as a function of latitude 9, longitude y , and season t can be attributed to warm or cold waters, the layer QP is photic, but always with low water temperatures, in the layer QL the phytoplankton is not produced. Finally, the layer Qf plays the role of a boundary layer.
The vertical transport in the ocean is determined by advective fluxes and gravitational sedimentation of dead organic matter (the flux H 2C0ij). An advective transport from the
i -th into the j -th reservoir of the ocean is considered proportional to the concentration of carbon in the respective reservoirs:
H^ = Cai (a = U,P,L),
where X2jj =Vj / Vj , Vj is the water volume transported per unit time from the i-th reservoir into the j-th reservoir; Vi is the volume of the ith reservoir.
The following algorithm is used for a parameterization of the process of carbon sedimentation. The flux under unit area of the ocean is supposed to decrease exponentially with depth. If we denote the inflow of organic matter in the i-th reservoir as gi and the net outflow of organic matter from the water surface as H20T, we obtain:
H20,! = H20,T ; H20,i =gM(o,/o,.1)exp[-(z, - z^D,], (i = 2,...,m + n + l);
where o, is the area of the i-th reservoir, Ds is the characteristic time of the organic matter particles sedimentation before their decomposition. The rate of decomposition in each reservoir is equal to: Rd,, = H^ - H^+1, (i = 1,...,m+n+l); Rd,f = HcW:m+n+l -Hcl6 .However, if the time of the transition of the organic matter particles from one layer to another is short compared to Ds , then it is better to take H2C0,i = ^1Ca,j , Hd- = X4CFjl . In addition to these fluxes one should take into account the fluxes of detritus decomposition, solution of bottom sediments, and carbon consumption in the process of photosynthesis: Hf7. = Const., Hfs. = X3DL,j ,
H22,, =^3 Du,, , H21,, = C31Ro,j .
One of the promising results of modelling the effect of «fertilization» due to changing concentrations of atmospheric CO2 has been discussed by Alexandrov and Oikawa [5] who, based on the TsuBiMo model, obtained a globally-distributed estimate of the contribution of various types of vegetation to the change of CO2 flux on the «atmosphere-land» interface. To compare this result with the global simulation model (GSM) [6] and mathematical model of biosphere (MMB) [4] data, Figure 2 presents curves of the dependence of net primary production of the surface vegetation on changes of the atmospheric CO2 partial pressure. The difference between these curves shows that a consideration in GSM and MMB of additional feedbacks compared with a simple TsuBiMo model specifies this dependence and recommends a more thorough structure of the MMB units as the GSM modernized version. Nevertheless, it should be borne in mind that despite an apparent simplicity of the
parametric descriptions assumed in TsuBiMo, the result obtained with its use is reliable due to a detailed description of the spatial structure of vegetation covers (50*50 km).
Therefore, the strategy of the modelling of the global carbon dioxide cycle should fol-
low a paradigm of the complete description of all direct interconnections and feedbacks with the use of a set of functional dependences, the choice among of which is made in the process of an adaptive-evolutionary numerical experiment.
Figure 1. The block diagram of the global biogeochemical cycle of carbon in the atmosphere-land-ocean system as the MMB item. The CO2 reservoirs and fluxes are described in Table
Table
Carbon flows in the Figure 1
Flow Origin of flow Flow Origin of flow
Hi Fuel burning H12 People vital functions
H2 Desorption H 13 Animal vital functions
H3 Sorption H14 Mortality of plants
H4 Rock erosion H15 Secretion by roots
H5 Volcanic emission H16 Deposition
He Photosynthesis in the ocean H17 Ocean depositions dissolving
H Respiration of plants H18 Detritus decomposition
H Burning of plants H 19 Water rising
H9 Decomposition of humus H20 Water descending and sedimentation
H10 People activities H21 Photosynthesis on land
H 11 Vital functions of biota in the ocean H22 River flow
300 400 500 600
Partial pressure of СО2 in the atmosphere (ppmv)
Figure 2. Comparison of three models of the
global carbon cycle to assess the response of net primary production of vegetation to changing concentrations of atmospheric CO2. Notation: TsuBiMo -Tsukaba Biosphere Model [5]
——
4
T 1 1 I 1 1 1 1 1 г
Figure 3. Model results for scenario IS92a with different spatial resolution in the SVS distribution (A9,AA) 1-(10°,10°), 2-(7.2°,9.0°), 3-(4°,5°), 4- (4°,5°) with the correction of soil moisture. The ordinate X=Ca(t)/Ca(1850)
Figure 3 shows dependence of atmospheric CO2 (Ca) dynamics on the different spatial resolution for SVS. As we see there exist some reserve in the precision of CO2 forecasting. Curve 4 was received with additional correction of soil moisture as input data for the MMB item that parametrizes the SVS production as function of CO2 concentration, soil moisture and temperature [4]. It demonstrates the microwave radiometry role in the increase of precision of vegetation model under the evaluation of CO2 sinks related to the concrete
spatial resolution. Really a consideration of satellite (for example, from EOS-Aqua) data about soil moisture with spatial resolution 1° x 1° or
0.5. x 0,5° will permit to have more precise estimations for spatial distribution of CO2 sinks and sources on the land. This study will be possible in the framework of the Global Carbon Project (GCP) [7].
References
1. Chukhlantsev A. A., Shutko A.M., and Golovachev S.P. Attenuation of electromagnetic waves by vegetation canopies in the 100-1000 MHz frequency band. ISTC/IRE Technical Report, #2059-1, 2003, 59 pp.
2. Ferrazzoli P. and Guerriero L. Passive microwave remote sensing of forests: a model investigation // IEEE Trans. on Geosci. And Remote Sensing, 1996, 34(2): 433-443.
3. Friedi M.A., Mclver D.K., Hodges J.C.F., Zhang X.Y., Muchoney D., Strahler A.H., Woodcock C.E., Gopal S., Schneider A., Cooper A., Baccini A., Gao F., and Schaaf C. Global land cover mapping from MODIS: algorithms and early results // Remote Sensing of Environment, 2002, 83(1-2): 287-302.
4. Kondratyev K.Ya., Krapivin V.F., Savinykh V.P., and Varotsos C.A. Global Ecodynamics: A Multidimensional Analysis. Springer/PRAXIS, Chichester U.K., 2004, 658 pp.
5. Alexandrov G., Oikawa T. TsuBiMo: a biosphere model of the CO2 - fertilization effect // Climate Res., 2002, 19: 265-270.
6. Kondratyev K.Ya., Krapivin V.F., and Phillips G.W. Global environnmental change: Modelling and Monitoring, Springer, Berlin, 2002, 319 pp.
7. Canadel J.G., Dickinson R., Hibbard K., Raupach M., and Young O. (eds.). Global Carbon Project: The Science Framework and Implementation. GCP Report No.1, Earth System Science Partnership, Canberra, 69 pp.
8. Sellers P.J., Bounoua L., Collatz G.L. , Randall D.A., Dazlich D A., Los S.O., Berry J.A., Fung I., Tucker C. J., Field C. B., and Jensen . T. G. Comparison of radiative and physiological effects of doubled atmospheric CO2 on climate // Science, 1996, 271: 14021406.
9. Nefedova E. I., Tarko A. M. Study of the global carbon cycle using the zonal model in the atmosphere-ocean system. Proc. of Russian Ac. Sci., 1993, 333(5): 645-647 (in Russian).
10. Alexeev V. V., Zaitsev S. I., Kokorin A. O. The growth of the rate of the atmosphere-ocean gas exchange in case of rough sea. Bull. of the USSR Acad. Sci., FAO, 1990, 26(3): 322-325 (in Russian).
11. Nitu C., Krapivin V.F., and Pruteanu E. Intelligent Systems in Ecology. Onesti, Bucharest, Rumania, 2004, 410 pp.
