An accurate estimation of crop evapotranspiration and soil water balance under potato-legume intercropping is important for improving the crop water productivity of rainfed potato. This study quantified the evapotranspiration, yield, and soil water balance of potato (Solanum tuberosum L.) intercropped with lima bean (Phaseolus lunatus L.) or dolichos (Lablab purpureus L.) in comparison to the same crops under monocropping. The experiment was set up in three agroecological zones of Kenya: upper midland (with an altitude range of 1500–1653 m above sea level (masl)), lower highland (1892–1923 masl) and upper highland (2502–2594 masl). The dual crop coefficient approach was adopted with the SIMDualKc model to estimate crop evapotranspiration and compute the soil water balance. The dual crop coefficient partitions crop evapotranspiration into crop transpiration and soil evaporation by applying both soil evaporation and basal crop coefficient. The model was calibrated and validated using field data observed along four crop growth seasons. The basal crop coefficients, the ratios of soil evaporation to evapotranspiration, and that of transpiration to evapotranspiration were determined. The yields of different intercrops were converted into equivalent yield of potato based on price of the produce. Good agreement between the observed and simulated data for available soil water and crop evapotranspiration was found with modeling efficiency > 0.8. The residual mean square error was low and ranged from 0.01−0.08 m³ m^-3 for available soil water and 0.03−0.09 mm d^-1 for crop evapotranspiration. The transpiration to actual evapotranspiration ratio of potato-legume intercropping was 6−28% greater than that of the sole potato because legume intercrops fully covered the ground by mid-season, thus limiting the energy available for soil evaporation. Crop yields were significantly greater under intercropping as transpiration occurred near its potential rate, thus not limiting yields. These results support the dual crop coefficient method as an appropriate tool to estimate and accurately partition crop evapotranspiration under potato-legume intercropping.

马铃薯-豆科间作作物蒸散量和土壤水分平衡的准确估算对于提高雨养马铃薯的作物水分生产力具有重要意义。本研究定量蒸散，产量，和马铃薯的土壤水分平衡（马铃薯L.）与利马豆（间作棉豆大号。）或扁豆（扁豆曲霉L.) 与单一作物下的相同作物相比。该实验在肯尼亚的三个农业生态区进行：上中部地区（海拔高度范围为 1500-1653 m 海拔（masl））、下高地（1892-1923 masl）和上高地（2502-2594 masl）。SIMDualKc模型采用双作物系数法估算作物蒸散量并计算土壤水分平衡。双作物系数通过应用土壤蒸发和基础作物系数将作物蒸发蒸腾分为作物蒸腾和土壤蒸发。该模型使用沿四个作物生长季节观察到的田间数据进行校准和验证。确定了基础作物系数、土壤蒸发量与蒸散量的比率以及蒸腾量与蒸散量的比率。根据农产品价格，将不同间作的产量换算为马铃薯的等价产量。在模拟效率 > 0.8 的情况下，发现可用土壤水和作物蒸发量的观测数据和模拟数据之间具有良好的一致性。残差均方误差较低，范围为 0.01-0.08 m³ m ^-3用于可用土壤水，0.03−0.09 mm  d ^-1用于作物蒸散。马铃薯-豆类间作的蒸腾量与实际蒸散量之比比单种马铃薯高 6-28%，因为豆类间作在季节中期完全覆盖地面，从而限制了可用于土壤蒸发的能量。间作下作物产量显着增加，因为蒸腾作用发生在其潜在速率附近，因此不会限制产量。这些结果支持双作物系数法作为估计和准确划分马铃薯-豆科作物间作作物蒸散量的合适工具。