Accurate determination of crop transpiration in greenhouses is very important in making exact irrigation scheduling and climate control.The most common method for calculating crop transpiration is the Penman-Monteith (PM) model.In this study,we analyzed the estimating methods of cucumber transpiration by PM method in greenhouse of South China.We measured meteorological data,transpiration and plant growth indicators of cucumber during 2 planting seasons (autumn-winter season in 2017 and spring-summer in 2018) in a Venlo-type greenhouse in south China.The PM model was used to predict the transpiration based on the data averaged over 30-min intervals using different approaches in the calculation of aerodynamic resistance (Perrier logarithm approach and heat transfer coefficient approach).The results showed that transpiration of the cucumber plants was mainly affected by meteorological factors when there was sufficient water supply.The fluctuations of transpiration were consistent with solar radiation,air temperature and canopy temperature,and reached the maximum at noon,while relative humidity of air had the opposite trends.Canopy temperature of the cucumber plants were lower than the air temperature during daytime.The maximum differences between air temperature and canopy temperature were 4.4 ℃(spring-summer season) and 2.0 ℃ (autumn-winter season) due to the evaporative cooling during high transpiration rates at noon time.By analyzing the response relationships between the stomatal resistance and the meteorological factors,a sub-model was constructed by using the effective leaf area index and stomatal resistance to simulate the canopy resistance of the PM model.A significant exponential relationship between stomatal resistance and solar radiation was observed in the study(R~2 = 0.89).The simulated canopy resistance determined by stomatal resistance and leaf area index remained constant during the night,with 533 and 574 s/m for spring-summer and autumn-winter season,respectively,and 96 and 112 s/m during the daytime when transpiration was high.The transpiration simulated by PM model were compared with the values measured by lysimeters.Results showed that the convection types of the greenhouse were conducted through mixed convection in about 80% and 94% of the study periods,respectively,for the spring-summer and autumn-winter seasons,while free and forced convection occurred at the rest periods at different times in the 2 planting seasons.In spring-summer season,free convection occurred in 5% of the study period during the day,forced convection occurred in 15% of the study period at night;In autumn-winter season,2% of the study period was free convection,which all occurred during the day,4% of the study period was forced convection,which occurred day and night.The variation ranges of aerodynamic resistance calculated by the Perrier logarithm approach and the heat transfer coefficient approach were small,and the average values was 388 and 383 s/m for the spring-summer and the autumn-winter season by the Perrier logarithm approach and 141 and 158 s/m for the spring-summer and the autumn-winter season by the heat transfer coefficient approach,respectively.The study showed that based on the PM model to simulate the cucumber transpiration had a good consistency with the measured values.However,the PM model underestimated transpiration when the Perrier logarithm approach was applied,the determination coefficients were 0.87 and 0.91 for spring-summer and autumn-winter seasons,respectively,and the PM model can accurately simulate the greenhouse cucumber transpiration using the heat transfer coefficient approach,the determination coefficients were 0.91 and 0.95,respectively,for spring-summer and autumn-winter seasons.