本研究采用淮河流域典型农田生态系统（寿县）冬小麦和水稻生长季内的3年通量观测数据（2007、2008和2009年），旨在评估三种模型模拟农田日蒸散量的性能。第一种是两步模型（PM−K_c模型），另外两种是一步模型（Rana−Katerji（R-K）模型和advection-aridity（AA）模型）。结果表明，淮河流域冬小麦和水稻生长季内的涡度相关数据的能量闭合度是合理的，介于0.84~0.91，R2约为0.80。冬小麦日蒸散量呈现先缓慢下降后迅速上升的变化趋势，而水稻日蒸散量呈现先上升后下降的变化趋势。在PM-K_c模型中，采用校正作物系数（K_c）的模拟性能优于采用联合国粮农组织（FAO）推荐的K_c的模拟性能。R-K模型和AA模型中校准的关键参数具有较好的通用性。经过校准后，PM-K_c模型的模拟性能最佳，R-K和AA模型均低估了冬小麦和水稻的日蒸散量。与R-K模型相比，AA模型的模拟性能较好，尤其在水稻日蒸散量模拟中。总体而言，本研究揭示了PM-K_c模型在淮河流域以及全球其他类似气候地区估算水稻和小麦作物需水量时的稳定性。

Potato grows in most part of China, it achieves higher yield and better quality in Gansu Province than in others. With global warming, its growth duration has been prolonged and sowing date become earlier than before. Therefore, to regulate its sowing date and growing period is of great significance for better harvest. In this study, experiments were conducted with six sowing-date treatments of potato in Dingxi, which is in the Loess Plateau of central Gansu Province in Northwest China in 2010. The growth period, morphological index and change in yield and their relationships with temperature, precipitation, and other climatic factors were investigated for each treatment. Results show that the crop with different sowing dates experienced different climate conditions, leading to distinct growth duration, plant height, and leaf area index. The growth duration was shortened due to a delay in sowing date. For each 15-day delay in sowing, the growth duration was reduced by 12 days on average. A significant linear relationship was found between numbers of days either from seeding to emergence or from flowering to harvest and mean temperature over the corresponding period. Dry matter accumulation, tuber fresh weight, and final yield were all decreased because of insufficient cumulative temperature over the shorter growing periods. Marked differences in tuber yield were discovered among the six treatments of sowing date, the potato planted on May 27 giving the highest yield. The potato planted either earlier or later would produce invariably lower yield than the treatment of May 27. Late May therefore can be taken as the optimum sowing time of potato in this region because the crop can fully utilize thermal resource. We conclude that to postpone sowing time is a good practice for potato production to adapt to climate warming in the Loess Plateau of central Gansu, China.