Land surface energy exchange over the rice-winter rape rotation system and its response to winter rape abandonment in the humid lowland region, southern China

Exploring the characteristics and main influential factors of energy exchange and evapotranspiration (ET) in the rice-winter rape system and their response to winter rape abandonment is significant for advancing the knowledge on land surface processes. The daily and seasonal variations in the energy exchange from 2017 to 2020 and their governing environmental factors were examined for rice, winter rape, fallow seasons, based on a Bowen ratio energy balance measurement and path analysis together with principal component analysis (PCA). The results showed that the total latent heat flux (LE) of the rice-winter rape (spanning 317 days) and rice-winter fallow (spanning 332 days) rotation systems were 1458.4, 1652.1 MJ/m 2 , respectively. LE consumed the largest net radiation ( R n ) proportion, with the mean seasonal values of LE/ R n of 0.87, 0.55, and 0.61 for rice, rape, and fallow seasons. Sensible heat flux ( H ) stayed low for rice but tracked with the LE for rape and fallow. The mean Bowen ratio was 0.13, 0.87, and 0.73 for rice, rape, and fallow seasons. The Priestley–Taylor coefficient for rice mostly stayed below 1.26, while the Priestley–Taylor coefficient for rape and fallow fluctuated around 1.26 except for lower than 1.26 in the late stage. The variables describing heat and water conditions [ R n , air temperature ( T a ), soil temperature ( T S ), relative humidity (RH), vapor pressure deficit (VPD)] explained most of the total variance in LE, while precipitation, wind speed, and leaf area index had no significant effect on LE. R n was the dominant environmental factor controlling LE mainly through the direct effect, followed by RH mainly through an indirect effect. VPD was another important factor for rice LE, and T S and T a also considerably affect rape and fallow LE. The winter rape abandonment caused an increase in the average daily LE, ET, and Priestley–Taylor coefficient, and a decrease in the Bowen ratio relative to the rape.