【作者】 米娜；

【导师】 王盘兴； 于贵瑞；

【作者基本信息】 南京信息工程大学 ， 应用气象学， 2007， 博士

【摘要】 森林生态系统水热和CO₂传输过程是国际研究的前沿问题。理解森林生态系统水、能量和CO₂传输机制有助于了解森林生态系统在陆地生态系统碳水循环中的作用，并认识森林生态系统过程与功能对干旱与气候变化的响应规律。本研究将基于生理生态学过程，模拟生态系统下垫面与大气之间碳水通量交换的综合模型—EALCO（Ecological Assimilation of Land and Climate Observation）模型引入国内，将其应用在中亚热带人工针叶林生态系统，对模型进行参数化与初始化使其能够较好地连续模拟该生态系统三年连续的碳、水和能量通量交换状况，同时用通量观测数据对模拟结果进行检验，利用EALCO模型揭示该生态系统碳水通量的驱动机制及季节性干旱对人工针叶林生态系统碳吸收及其碳水通量耦合关系的影响，模拟和预测未来气候变化情景下人工针叶林生态系统对气候变化的响应和碳收支状况，以期为我国估计大区域尺度生态系统碳平衡的时间和空间格局特征提供模型储备。本文的主要研究成果如下：（1）将EALCO模型进行参数化和初始化，并改变模型的光合作用时段和落叶机制等，使之能够较好地模拟亚热带人工针叶林生态系统碳水通量状况，利用通量站的观测数据与模型模拟结果进行比对，检验了模型对水和能量通量特征的模拟效果：结果表明模型对两个重要的中间变量，即冠层温度和水势的模拟结果较为理想。通过将冠层气孔导度的模拟值与利用Penman-Monteith公式计算的表面导度进行对比发现，除2005年模型模拟的冠层导度偏高外，对其它两年的模拟效果较为理想。净辐射三年的模拟值与实测值之间的相关系数均在0.98以上，模拟效果理想。模拟结果表明，被森林地面吸收的净辐射(R_n，s)和被冠层吸收的净辐射(R_n，c之间的比率也表现出比较明显的季节变化形式，比率的月平均值在4或5月份达到最大值0.35左右，之后逐渐减小，至12月或1月降至最低值0.05左右。2003年和2004年潜热通量与显热通量的模拟结果好于2005年。总体来看，三年潜热通量和显热通量模拟值与实测值的相关系数平均分别为0.82和0.62。模拟结果显示，森林地面日平均潜热通量的模拟值介于5-40W m^-2之间，2003-2005年森林地面年总潜热通量分别占年总冠层潜热通量（LE_s／LE_e）的8％、11％和12％。观测与模拟的生态系统潜热通量（LE_e）占净辐射（R_n）的比例在年初至春季（DOY120）之前较小，在30％-50％之间，进入生长旺季，LE_e／R_n逐渐增加，在50％-80％之间。显热通量的日平均值无明显的季节变化规律，较高的显热通量通常在春季（DOY100前后）出现。（2）检验了模型对碳通量特征的模拟效果：从碳通量的日变化来看，模型能够较好的模拟碳通量的日变化形式，对2003-2005年的模拟结果进行统计分析表明，模型对NEP的模拟相关系数为0.67，斜率及标准差分别为0.90和3.96μmol C m^-2s^-1，可见模型可以在很大程度上反应出该生态系统NEP的日变化趋势。模型对土壤呼吸在半小时尺度上的模拟值接近或略高于静态箱-气象色谱法的土壤呼吸观测值。从碳通量的季节变化来看，模型对2003-2005年GPP、R_e和NEP的模拟结果也较为理想，2003年7月和10月生态系统受到干旱胁迫期间，模型对GPP和R_e均存在高估现象。利用箱式法的观测数据与模型模拟的土壤呼吸日总量进行了对比，除2005年生长季期间的模拟值与观测值差异较大，其它年份的模拟值与观测值基本一致。对生态系统各种呼吸组分的模拟结果显示，该生态系统自养呼吸占生态系统总呼吸的比例较为固定为88％。就生态系统自养呼吸来讲，在全年中，植被的维持呼吸均大于生长呼吸。2003-2005年维持呼吸占自养呼吸的比例分别为77％、72％和76％。对土壤呼吸各组分的模拟结果表明，三年中，模拟的根系自养呼吸占土壤总呼吸的70％。（3）分析了季节性干旱对人工针叶林生态系统碳水通量的影响：研究结果表明深层土壤水分含量是决定冠层导度的主要因素，进而影响GPP的大小；干旱对GPP的影响比对R_e的影响更为强烈，因此碳吸收能力的下降主要由GPP下降剧烈引起，对GPP下降的环境因素进一步分析表明，在晴天正午之前对光合作用能力产生抑制的因素主要是深层土壤水分，正午之后深层土壤水分匮缺与高温共同影响生态系统光合作用能力，两者对光合作用能力的削弱作用各占一半。2003年高温干旱期间，生态系统呼吸值下降33％左右，生态系统呼吸的降低主要是由植物自养呼吸和土壤异养呼吸的减小共同作用引起。其中自养呼吸的减小主要是由生长呼吸的减小所导致。2003年持续的夏季高温致使生态系统的碳吸收能力下降，2004年10月的水分匮缺对生态系统呼吸的影响更为强烈，因此碳吸收能力并未受到影响，可见生态系统碳平衡的两个组分R_e和GPP对干旱响应的方式与程度有所不同，是造成森林生态系统源／汇强度变化的根本原因。此外，干旱还会使植物碳水之间的耦合关系发生改变。（4）分析了人工针叶林生态系统对气候变化的响应；研究结果表明，气候变暖会增加ET，从而降低土壤的水通量。但在气候变暖与CO₂浓度升高的共同作用下，ET可能会减少，主要原因在于CO₂浓度的变化导致了冠层气孔导度的变化。气候变暖会提高生态系统的生产力。与降水增加相比，CO₂浓度的增加更能有效地提高生态系统的生产力。CO₂浓度升高所引起的总初级生产力的提高，不会被由温度增加所导致的呼吸速率增加而抵消，NEE仍然有增加趋势。更多还原

【Abstract】 Water, heat and carbon dioxide transfer processes in the forest ecosystem arefocused intensively by the international researches. Understanding the mechanism ofwater, heat and carbon dioxide transfer in the forest ecosystem is important forassessing the role of the forest ecosystem in the water and carbon cycle of theterrestrial ecosystem and learning the ecosystem process and function of forest inresponse to the drought and climate change. EALCO model is a process-based modelused to simulate carbon, water and energy exchange between the atmosphere and landsurface. In this study, EALCO model is parameterized to simulate the ecosystemcarbon, water and energy exchange process in the man-planted evergreen forest.Modeling result was tested by use of the flux tower measurements. The mainobjective of the study was to （1） investigate the driving factors of the carbon andwater fluxes of the ecosystem （2） study the effect of seasonal drought on carbonassimilation and the coupling of carbon with water （3） simulate and evaluate thecarbon and water exchange conditions under the scenarioes of climate change.Main results of the study were as follows:（1） EALCO model is parameterized and initialized to simulate the ecosystem carbon,water and energy exchange of the man-planted forest from 2003 to 2005. Simulatedecosystem water and energy exchanges are analyzed on half-hourly, daily, and annualtime scales and compared with tower eddy correlation flux measurements.The results show that the simulated two critical variables, T_c andψ_c, werereasonable compare to the observations. Comparing the canopy conductance with thecalculation from the Penman-Monteith equation we found that the modeling results of2003 and 2004 are just satisfied while the results of 2005 are overestimated. Thecorrelation between the simulated and observed results of net radiation for the 3 yearsis more than 0.98. Modeling result shows that the ratio between net radiationabsorption by the forest floor and by the canopies showed an obvious seasonal pattern,with monthly mean values ranging from 0.35 in May to 0.05 in December. The modelreproduced well the annual course of daily latent heat （LE） and sensible heat flux （H）above the forest canopy as measured by EC method at 2003 and 2004. Thecorrelations between the simulated and observed results of H and LE for the 3 yearsare 0.82 and 0.62, respectively. The computed daily-averaged forest floor latent heatdensity （LE_s） ranged between 5 and 40 W m-2 throughout the three years and theseasonal sum of LEs accounted for 8, 11 and 12% of that of LE_e in 2003, 2004 and 2005, respectively. The latent heat fluxes of the whole ecosystem accounted low valueof Rn at the beginning of the year （from 30 to 50%） and increase with the growing ofthe forest （from 50 to 80%）. The daily-averaged sensible heat flux did not show adiscernible seasonality. Peak-averaged values often accrued in the early spring.（2） Simulated plant, soil and ecosystem CO₂ exchanges are analyzed on half-hourly,daily, and annual time scales and compared with tower eddy correlation fluxmeasurements and estimates from various authors.At half-hourly timescale, the simulated plant CO₂ exchange explained 67% of thevariance of the measured CO₂ fluxes derived from eddy correlation measurements.The simulated soil respiration was close to or higher than the soil chambermeasurements in the three years. At a coarser timescale, the model was verysuccessful in simulating variations of GPP, Re and NEP of the three years except forthe overestimation of GPP and Re in July and October of 2003 when during thedrought period. The simulated daily soil respiration was in consistent with thechamber measurements except for the year of 2005. Modeling results of thecomponent of ecosystem respiration show that the contribution of auto-respiration toecosystem respiration was 88% for the three years. As for autotrophic respiration,maintenance respiration was morn than growth respiration through the whole year andthe contributions of maintenance respiration to auto-respiration of 2003, 2004 and2005 were 77%, 72% and 76%, respectively. Modeling results of soil respirationcomponent show that root autotrophic respiration account for 70% of the soilrespiration.（3） The impact of seasonal drought on carbon and water exchanges of man-plantedforest was analyzed.Results show that deep soil water content decided the canopy conductancegenerally and GPP depended on canopy conductance to a large extent. Drought havemore intense impact on GPP than on Re, which lead to the decrease of net carbonexchange during the water stress period. Further analysis suggests that deep soil watercontent controls the canopy photosynthesis dramatically in sunny day before noontime during soil water stress. While after noon time both high temperature and deepsoil water content eliminate the GPP and their elimination percents are equal. Duringthe drought and high temperature period of 2003, ecosystem respiration decreased33%, which was caused by the decrease of plant autotrophic respiration and soilheterotrophic respiration. Among them the decrease of autotrophic respiration wasgenerally caused by the reducing of plant growth respiration. The net carbon exchange of 2003 was decreased due to the combination effect of summer drought and heatwave while in October of 2004 the water stress has more influence on ecosystemrespiration so the net carbon exchange was not influenced largely, from which we cansee the net effect of ecosystem carbon balance depends on how these two quantifiesare affected relatively to each other. In addition, drought also has impact on thecoupling of carbon and water fluxes.（4） Climate change effects on the carbon and water fluxes of the man-planted forest.The ecosystem model EALCO was used to evaluate possible changes inecosystem carbon and water exchange and accumulation under changes inatmospheric CO₂ concentration and accompanying changes in air temperature andprecipitation. Model results indicated that warming will cause increased ET andtherefore reduced soil water flux. Combination impact of warming and increased CO₂concentration will lead to reduced ET, which attributed to the reduced stomatalconductance under the high CO₂ concentration. Warming will cause increased growth.Compared to increased precipitation, the increase of CO₂ concentration improved theecosystem production effectively. Future CO₂ induced enhancements of grossphotosynthesis would not offset by temperature-induced increases in respiration, thecombined influences of the scenario resulted in a 23% increase in NEP.更多还原