黄土高原农田水碳通量观测与陆面过程模式参数优化

The study will focus on the hot scientific issue on the water and carbon cycles of terrestrial ecosystems. Based on the combination of field comprehensive observations and model simulation, we will investigate the characteristics and its environmental effects on water and carbon exchanges over Loess Plateau farmland ecosystems, and attempt to make parameters estimation in the land surface processes model. Along the averaged annual precipitation gradient, the three representative farmland ecosystems in the Loess Plateau will be selected for observation and modeling. The eddy covariance technique was used to observe water and carbon fluxes of the farmland ecosystems, and the meteorological gradient systems were used to measure the corresponding environmental factors. Then, the relationship between water and carbon fluxes and environmental factors will be analyzed. Partitioning of the net ecosystem carbon exchange into gross primary productivity and ecosystem respiration by means of ‘non-linear regression method’, and partitioning of evapotranspiration into soil water evaporation and vegetation transpiration, combination of stable isotope techniques, were conducted. The temporal and spatial characteristics of water and carbon fluxes and their components in farmland ecosystems in in the Loess Plateau will be analyzed. Afterwards, we will use the observe data of water and carbon fluxes to examine the CABLE land surface model, and trying to make parameters optimization about plant physiological parameters (Vcmax, Jmax) of the CABLE model at the ecosystem-level, based on "Bayesian" modeling approach. The relationships between plant physiological parameters and precipitation, soil moisture, evapotranspiration and carbon exchange were analyzed under the background of climate variances in the Loess Plateau, trying to establish the land surface process parameters for farmland in the Loess plateau.

本项目将围绕陆地生态系统水碳循环问题，采用野外综合观测与模型模拟相结合的思路，研究黄土高原农田生态系统水碳交换特征及其环境影响，并尝试对陆面过程模式进行参数估计。沿年均降水量梯度，在黄土高原地区选择三个典型农田生态系统开展水碳通量观测和模拟研究。采用涡度相关技术对农田水碳通量特征进行观测；采用气象梯度系统对相应的环境要素进行观测。采用“非线性回归法”将净生态系统碳交换分解为总初级生产力与生态系统呼吸；采用双源模型结合稳定同位素技术，将生态系统蒸散分解为土壤蒸发与植物蒸腾。分析水碳通量各组分的时空变异性特征。利用水碳通量观测资料检验CABLE陆面过程模式；结合“贝叶斯”方法，在生态系统水平上对CABLE模式中植物生理参数（最大羧化速率、最大电子传递速率）进行优化。分析不同地域气候差异背景下植物生理参数与降水、土壤湿度、蒸散和碳交换的关系；尝试建立生态系统水平上黄土高原农田陆面过程参数方案。

准确观测与模拟黄土高原陆地生态系统与大气间水碳交换过程，对理解生物圈与大气圈相互作用以及评估气候变化对生态系统影响有重要意义。研究结果表明，夏季PPT对年ET的贡献大于秋季，而生长前期和早春期的极端干旱气候制约了ET的增长。退耦系数与冠层电导(gc)之间存在强的相关性（R2=0.85）。这表明ET对(gc)敏感，随着日间时间的发展，gc对ET控制作用增强。ET的生理调控与土壤水分密切相关。干旱严重影响了ET对环境因子（如短波辐射和饱和水汽压差）的响应方式。黄土高原草地生态系统是一个弱碳源，生态系统净碳吸收的年际变化与年降水量显著相关。东亚夏季风通过主导气候因子，如年降水量和季节性干旱，对草地二氧化碳交换起着至关重要的作用。干旱胁迫对总初级生产力（GPP）的抑制作用大于对生态系统呼吸（Reco）。土壤水分降低会导致Reco对土壤温度（Ts）的敏感性减弱，而在严重干旱条件下，且Ts较高时，由Reco-Ts指数增长函数估计的GPP值可能过高。此外，还发现春季增温能够增加草地的碳吸收，并补偿夏季干旱导致的降低。通过NW-MNPS模式优化物理过程参数化方案、陆面过程参数化方案、边界层参数化方案、辐射方案等，对西北区域数值预报业务试验系统的预报结果进行检验分析，结果表明：NW-MNPS模式对西北区域气象要素预报整体效果较好。