中文摘要：

作物系数是计算作物需水量必不可少的参数。利用2008年野外水分试验和4个气象站近26年的土壤水分和气象等常规观测资料,以相关分析和回归分析等统计学方法为基础,根据水量平衡原理计算了内蒙古典型草原区的作物系数,分析了其在生长期和不同站点间的变化规律;建立了典型草原标准作物系数与返青后年日数和大于0℃积温的模拟方程,相关指数在0.94以上。在分析湿润指数、叶面积指数和盖度与作物系数关系的基础上,提出标准作物系数的气候修正方法和胁迫条件下作物系数的修正方法。同时,与修正后的联合国粮农组织（FAO）推荐值比较后得出,生长季标准作物系数的平均值为0.60,最大值为1.02;不同生长阶段作物系数的典型值分别为：初始生长期0.40,生长中期0.93,生长后期0.80,相应的阈值范围为0.35–0.45、0.85–1.00和0.70–0.90。通过旬蒸散量的模拟计算值与蒸渗仪实测结果的比较,平均相对误差在20%–24%之间,生长旺盛期大多低于10%,从而初步证明该文提出的方法在内蒙古典型草原区有较好的适用性。

英文摘要：

Aims Crop coefficients are the key parameters in calculating crop evapotranspiration. There are few reference values for crop coefficients of natural grass in FAO irrigation and drainage papers. Our aim was to define crop coefficients under different water conditions for typical steppe in Inner Mongolia based on actual observation data so as to accurately calculate the water demand and actual evapotranspiration. Methods Data were collected from field experiments at Xilinhot （44°08′03″N, 116°19′43″ E, elevation 990 m）, Inner Mongolia in 2008, and meteorological data from 26 years were obtained from four stations. Using the water balance equation, we calculated crop coefficients of typical steppe using statistical analyses and developed the standard crop coefficients of typical steppe and simulated equations through comparison and validation with rec-ommend values by FAO. We established a climatic correction method for the standard crop coefficients and a correction method of crop coefficients under water stress. Important findings Standard crop coefficient values of typical steppe were fitted with trinomial using days after return of the green period of grass and accumulated temperature （≥0 °C）, with correlation coefficients 0.94. Crop coefficients of typical steppe were described as an exponential equation with relative leaf area index （LAI） and relative plant cover. Average value of standard crop coefficient of typical steppe for the growing season was 0.60, and the maximum value was 1.02. Typical values of crop coefficients in different growing stages were 0.40 for initial stage, 0.93 for medium stage and 0.80 for later stage, and threshold ranges of three growing stages were 0.35–0.45 for initial stage, 0.85–1.00 for medium stage and 0.70–0.90 for later stage. Compared with measured data from a lysimeter, the simulated values for 10 days of evapotranspiration indicated an average relative error of 20%–24% for whole growing season, and 10% for the active growing stage. The metho