中文摘要：

频繁出现的极端干旱事件会使水分循环的关键过程发生改变，进而显著影响生态系统的过程和功能. 为了定量分析极端干旱对半干旱草原水分平衡的影响，利用内蒙古羊草（Leymus chinensis）草原生态系统通量观测站1978-2005年的气象数据、野外实测数据，对基于过程的生物地球化学循环模型——DNDC（Denitrification-decomposition）进行参数化，模拟羊草草原生态系统2004-2005年各水分通量（蒸散、蒸腾、蒸发）和水分平衡的动态变化过程，分析极端干旱对蒸散、蒸腾、蒸发的影响. 通过与涡度相关实测数据对比发现：参数化后的DNDC模型能够很好地模拟羊草草原生态系统的水分通量，在峰值大小和时间动态上，模拟值与实测值有较好的一致性（R2 = 0.68，P 〈 0.0001），且实测和模拟结果都发现2005年的极端干旱使蒸散量显著降低（P 〈 0.0001）. 基于参数化后的DNDC模型对植物蒸腾和土壤蒸发也进行了定量分离，发现与2004年相比，极端干旱使年蒸腾量降低了57%，年蒸散量降低了30%，生态系统由水分盈余转为严重的水分亏缺，进而使总初级生产力降低了73%，影响生态系统的碳源/汇功能. 而且，羊草草原生态系统水分平衡发生改变的降水阈值为20.8 mm/月. 本研究表明，极端干旱显著改变了植被蒸腾和土壤蒸发对蒸散的相对贡献，其对植被蒸腾的影响要远大于对土壤蒸发的影响.

英文摘要：

Potential changes in both the intensity and frequency of extreme drought events can alter the distribution and dynamics of water availability and subsequently affect biological processes at the ecosystem level. This research aimed to study the effects of extreme drought on the water balance for a Leymus chinensis steppe in Inner Mongolia, China. In this study, a process-based biogeochemical model, Denitrification-decomposition （DNDC）, together with eddy covariance （EC） technique, was modified and employed for the first time to simulate and interpret dynamics of water fluxes including evapotranspiration （ET）, transpiration （T） and evaporation （E） and analyze the water balance for the targeted steppe during 2004-2005. Daily weather data of 1978-2005 in conjunction with soil properties and management practices for the very location were utilized to simulate the grass growth and water fluxes dynamics with DNDC. The modeled ET fluxes were compared with the eddy tower data. The results suggested that the seasonal patterns and magnitudes of simulated daily ET fluxes were basically in agreement with observations for the period of 2004-2005 （R2 = 0.68, P 〈 0.0001）, both showing that the extreme drought in 2005 significantly decreased ET fluxes for the targeted grassland （P 〈 0.0001）. Then, DNDC quantitatively separated T and E from ET, and found that extreme drought substantially decreased the annual transpiration and evapotranspiration by 57% and 30%, respectively. That shifted the ecosystem from water surplus to severe water deficit, which decreased gross primary productivity by about 73% and affected the ecosystem carbon sink or source function. The precipitation threshold for shifting water surplus and deficit function was found to be 20.8 mm per month in Leymus chinensis steppe ecosystem. The results suggested that extreme drought significantly changes the relative contribution of vegetation transpiration to ET fluxes, and that the effects of extreme drought are much stronger on vegeta