(Abstract)

The Intergovernmental Panel on Climate Change (IPCC) published their fourth report, which described how and why climate systems are changing, the impact assessment on mankind and ecosystems, and many of its possible future impacts. However, the simulation of future climate change by global climate models has included many of the generic uncertainties. The difficulty of estimating the impact of the change on the hydrological cycle should be one of problems because the transport of moisture is associated with the energy transport. The water cycles have significant impacts on the regional and global climate systems, respectively. The hydrological cycle is a highly non-linear portion in Earth's climate system. It is the critical result of climate change and plays a key role in regulating the climate, such as positive and negative feedbacks.

The vegetation influence climate through physical, chemical, and biological processes that affect planetary energetics, the hydrological cycle, and atmospheric composition, and this action has the different effect on many time scales, for example on a sub-daily scale (e.g. stomatal closure, solar elevation angle), on a seasonal time scale (e.g. vegetation phenology) and on multi-year scales (e.g. changes in land cover, extreme climate change). These complex and nonlinear atmosphere-biosphere interactions can dampen or amplify anthropogenic climate change. Therefore understanding how energy and water are exchanged between the atmosphere and terrestrial biosphere becomes an important component in understanding and predicting climate change which includes global water cycle. However, the influence of vegetated land surface on large-scale climate is still poorly understood in the real world. This theme is able to approach the system of land-atmosphere interactions on a range of space and time scales using observation experiment products to develop hypotheses, prepare parameters and validate models.

In nature, soil-vegetation-soil transfer system explicitly considers the role of vegetation in affecting water and energy balance by taking into account its 1) physiological (stomatal conductance) and 2) structural (canopy aerodynamics) properties. These two vegetation functions are also the basis of evapotranspiration parameterizations in physically-based hydrological and climate models. However, most current SVAT schemes and hydrological models do not parameterize the characteristics of vegetation as a dynamic component. The aim of this study is to understand the role of vegetation for evapotranspiration which is critical factor on the partitioning of land surface energy and water cycling, regionally and is linking factor for both microclimate and hydrometeorology, globally.

In chapter 2, Land surface energy partitioning in various vegetated surfaces is controlled by climatological and biophysical factors. The "big leaf" model represents these conditions by three resistance parameters: bulk surface resistance, aerodynamic resistance, and climatological resistance. Author tries to understand land surface energy partitioning in various vegetated surfaces with the flux and meteorological data from 27 FLUXNET sites, and using the big leaf model of bulk surface, aerodynamic, and climatological resistance in Penman-Monteith approach. Here, author introduces a new method to estimate the total resistance and normalized surface resistance and shows that the Priestley-Taylor coefficient can be related to the total resistance, and the Bowen ratio can be described by air temperature and normalized surface resistance. The proposed empirical method in this chapter is useful for estimating evapotranspiration using remotely-sensed data and for understanding the energy balance partitioning in land surface models.

In chapter 3, the author introduces MATSIRO (Minimal Advanced Treatments of Surface Interaction and RunOff), the well-known one of 3rd generation land surface model, as main three sub-schemes related with canopy structural and plant physiological effects on land surface energy partitioning: radiation transfer, aerodynamic surface roughness, and plant physiological process schemes in MATSIRO. Atmospheric general circulation models used for climate simulation and weather forecasting require the fluxes of radiation, heat, water vapor, and momentum across the land-atmosphere interface to be specified. These fluxes are calculated by sub-models called land surface parameterization scheme (LSPs). Recent LSPs represent globally the system of soil-vegetation-atmosphere transfer as advances in plant physiological and hydrological research. It incorporates biogeochemical and ecological knowledge. The author applied four modifications based on sensitivity tests in order to improve the simulation accuracy:

1) The role of tree stems on radiation energy balance and aerodynamic transfer is considered. In early LSPs, leaf area index and canopy height are used as manner of describing the above-canopy structure. However, only both canopy structure components are not sufficient parameters to represent land properties in the surface boundary layer because of passing over the impact of stem morphologies as the permeable obstacles for the rain falling, the light penetrating, and the wind going through. Therefore, in recent LSPs, stem area index (SAI) is added with leaf area index (LAI) which is one of main key phonological parameters in LSPs. In order to consider SAI, LAI is replaced by plant area index to estimate interception loss and roughness length. The simple weighted factor with the ratio of SAI and LAI on PAI is used. Nevertheless, the simple application of PAI in previous studies is not available to represent the effects of stem because the whole skins of leaves and stem are not exposed to outside and the reflection and drag coefficient for leaf and stem object are different. Therefore, the author applies the exposed weighted factor using Monsi-Saeki's light extinction in MATSIRO.

2) The hydrological impact of the vertical distribution of plant root is critical on land water cycle. In nature, roots define the biologically and chemically most active zone of the soil profile. The depth of the rooting zone of the vegetation cover, or rooting depth, determines the extent to which soil moisture can be extracted by vegetation for transpiration. In LSPs, root vertical distribution is treated as a static component. The parameter of vertical root distribution in MATISRO is determined by look-up table of vegetation types. In other word, most of LSPs assume that root distribution profile in one vegetation type is not changed. However, the changes in leaf and stem biomass are reflected in not only leaf area index and vegetation height but also rooting depth, as following plant allometric relationships. In most grassland and cultivation area, transpiration is increased and runoff is decreased. In fact, the root fractions in first soil layer of grassland and cultivation type are relatively larger different than other vegetation types. In Africa region, decreased runoff might be caused increased transpiration. This area is dry, and thus first layer soil moisture is quite variable through the amount of rainfall.

3) Photosynthetically active radiation (PAR) energy reaching on the vegetated surface is a key determinant of physiological processes. In order to procure PAR variable in the model, the ratio of PAR to Rs (PAR/Rs) is commonly used to convert Rs into PAR. PAR estimated from incoming solar radiation (Rs) reduces the weather data requirements. It will reduce the weather date requirements for the model. Several models simply use 0.5 for PAR/Rs as the constant ratio. However, previous field experiments have been represented the variations of PAR/Rs, not constant ratio. The previous empirical equations for estimating PAR/Rs are derived from the data of relatively long time-scale (e.g., daily and monthly) and limited measurement sites. It is not suitable for current biosphere model having with several hourly and global scale simulation. Here, we represent the exponential correlation between hourly PAR/Rs and sky clearness index (0~1) using data set of 54 measurement sites in Ameriflux. PAR/Rs is increased until 0.6 by cloudy conditions of below about 0.2 clearness index. To the contrast, PAR/Rs of over 0.2 clearness index is retained 0.42 as the constant value. When our empirical equation is replaced with previous one in biosphere model, it will be possible to bring -4~2 percent different on stomatal conductance, which is most critical parameter for plant transpiration.

4) The scale-up method from leaf to canopy considering diffuse radiation is important on climate change. Scaling processes from a single leaf to an entire canopy is necessary when considering the role of stomatal response to environmental variables on interaction between biosphere and atmosphere because of complex environmental and physiological gradients within the canopy. Several studies have shown that diffuse light, which enters a complex canopy from all directions, reaches leaves more evenly than light in the direct solar beam. Consequently, canopy light use efficiency increase with fraction of diffuse radiation and canopy photosynthesis has less tendency to saturate. Beer's law, which is the assumption theory of Monsi and Saeki theory does not account for the loss of scattered radiation if the extinction coefficient is calculated as G/cos(θ), where G is the mean area projected onto a surface normal to the sun's ray per unit leaf area, and θ is the zenith angle of incident radiation. Because of its simplicity, Beer's law has been used in most models. One previous study found that the absorption of intercepted and scattered radiation could be accounted for by reducing the canopy extinction coefficient. The absorptions of direct beam and diffuse radiation are not calculated separately. It is necessary for the canopy scaling-up considering with diffuse radiation in global climate simulation.

In chapter 4 and 5, canopy structural and plant physiological effects are investigated, respectively. In chapter 4, the author hypothesizes that a functional relation between albedo and roughness length can be sought for a vegetated surface since both land surface parameters are conceptually related to vegetation structure parameters such as leaf area index, canopy height, canopy density and crown area. The author investigated this relationship between α and z0 by using vegetation structure parameters collected from observed data from 48 measurement sites worldwide covering various vegetated surfaces. Based on these data, an inverse nonlinear relationship between roughness length and albedo is found. It is shown that this observed relationship can be empirically related to canopy structure parameters such as leaf area index and canopy height. Our results demonstrate that plant canopy structure formed by the pattern of biomass partitioning with the accumulated carbon assimilation determines the key land properties for surface energy budget: albedo and roughness length.

In chapter 5, recent simulation results by land surface parameterization schemes (LSPs) used for climate modeling represent the change of global water balance through stomatal regulation. The author investigates the stomatal response to humidity and CO2 concentration. Real stomatal behavior which is summarized in previous literatures is compared with the calculated stomatal conductance, and the method of stomatal conductance is modified considering both influences of humidity and CO2 concentration. The author modified two parts of stomatal conductance model; 1) leaf temperature calculation and 2) performance of the Ball-Berry model under the elevated CO2. Those two model modifications are very important to represent plants' physiological effects realistically in any similar stomata models. Transpiration decreasing by "CO2-induced stomatal closure" is under-estimated in original MATSIRO (Ball-Berry model) because of relatively high VPD in tropical regions. However, it is equivalent to "runoff increase" because the increased surface soil moisture by reduced transpiration results in less water stress on soil evaporation. Therefore, "runoff increase" by the CO2-induced stomata closure will possibly occur, but the magnitude of increase may vary from region to region, because of the differences in humidity and soil moisture conditions. These results highlight certain important impacts of plant physiological changes on the global hydrological cycle. However, it is critically important to understand the real physical processes at work; 1) validate the modified Ball-Berry model using observed data and 2) conduct on-line simulations considering precipitation change.

気候変動に関する政府間パネル(IPCC)は2007年に公表した第4次報告書で、地球表層の平均気温が上昇していることは疑う余地がないと断定し、それが人為起源である可能性が非常に高いと結論付けた。その根拠は気候モデルと呼ばれる数値モデルに、人為起源の排出に伴う大気中の温室効果ガス濃度の上昇やエアロゾルなどを入れた場合といれない場合とで、いれた場合でないと20世紀後半の気温の上昇がうまく説明できないから、という論理による。

このように、IPCCによる科学的報告書が国際政治や経済に及ぼす影響の大きさを考えると、気候モデルが地球の気候システムの応答を的確に表現しているかどうかは非常に重要な問題であり、その不確実性を減少させることは喫緊の研究課題である。

植生は物理的、化学的、そして生物学的作用を通じて地球のエネルギー・水循環や大気中の微量成分組成など気候システムそのものに大きな影響を及ぼしており、しかも、気孔の開閉といった日周期から、植生の発芽・葉の展開といった生物季節変化、そして植生被覆変化といった経年変化まで、さまざまな時間スケールで気孔システムと関わっている。そうした植生過程は様々なフィードバックを通じて人為的な気候改変を増進したり抑制したりする可能性があるが、いまだに未解明で知られていない点、あるいは気候モデルで適切に表現されていない側面がある。

そこで、本研究では、野外観測結果や室内実験結果に基づく知見を総合して、植生の生理現象と物理構造が蒸発散過程に及ぼす影響を適切に表現できるように気候モデルで用いられる陸面水文植生モデルの植生過程部分を改良することを目的とした。

第1章では、気候システムにおける植生の役割や水循環を通じた大気と陸面植生との相互作用などに関する既往の知見がまとめられている。

第2章では、地表面に到達する太陽エネルギーの顕熱と潜熱への分配に及ぼす植生構造の影響を明らかにするため、世界的に組織されたフラックス観測データの相互利用研究であるFLUXNETから地表面フラックス観測データを収集し、ペンマン・モンティース法を用いたビッグリーフ型の植生モデルに適用している。

第3章では、MATSIROと呼ばれる第3世代の陸面水文植生モデルを用いて研究が進められている。第3世代の陸面水文植生モデルでは、水収支、エネルギー収支に加えて、植物生理に基づき光合成に伴う気孔開閉がエネルギー・水収支や炭素固定量に及ぼす影響までが考慮されている。これに対し、既存の研究論文の文献調査に基づいて、外力として与える光合成有効放射量の推計部分、下向きの直達放射量と散乱放射量を算定する際に葉の分布のみならず枝や幹も考慮する点、土壌層における根の入り具合の分布の与え方、そして散乱放射量を考慮して樹冠層を積分して取り扱う手法、の4つの部分についてMATSIROの植生過程に改良を加えている。

第4章では、樹冠層の構造的効果について研究が進められ、植生に覆われた地表面では、どちらも葉面積指数や植生の高さ、密度、被覆率に関連しているので、反射率(albedo)と地表面粗度との間には関数関係があるのではないか、という仮説が提示された。この仮説を検証するため、さまざまな植生タイプを含む48のサイトにおける観測データが収集され、両者の間に非線形な逆相関が見出されている。グローバルに取得も可能なパラメータとして葉面積指数と植生の高さが選ばれ、それらによって地表面の反射率や粗度を推計する経験式が提案されている。

最後に、第5章では、大気中の湿度と二酸化炭素濃度が気孔開閉に及ぼす影響を通じて植物生理が水循環に及ぼす影響が詳細に研究された。その結果、現在の気候モデルで広く使われている計算法(Ball-Berryモデル)では、二酸化炭素濃度増大と大気中の湿度欠損に対する短期的な気孔抵抗の応答が逆変化になっていることを見出し、これに変わる定式化を提案している。また、葉面温度の算定手法についても、モデル中で一貫性を保てるような算定手法に修正している。これらの改良により、二酸化炭素濃度上昇時にはさらに蒸散量が減少する可能性があるが、その分裸地面における土壌水分欠損が減少し蒸発量が増大するため、結果としてグローバルな流量増加は、改良を施す前と大きくは変わらないことが示されている。

このように、本研究は、気候変動予測の不確実性を減らす対象として非常に重要かつ野心的な課題である陸面植生過程に関して、植物生理の基礎に立ち返り、気候モデルにおける数値モデルの改良に反映させて、その信頼性を高め、より現実的に二酸化炭素濃度の増大といった変化が植生過程を通じて気候変動に及ぼす影響を明らかにしたものであり、有用性に富む研究成果と評価できる。よって本論文は博士(工学)の学位請求論文として合格と認められる。