Long-term variations in the ratio of transpiration to evapotranspiration and their drivers in a humid subalpine forest

Amsterdam

Elsevier

2025

13 S.

Sonderdruck

40005850

Agricultural and forest meteorology: An International Journal ; 311 (2021) , 13 S.

Identifying the temporal variations in transpiration (T) and its contribution to evapotranspiration (ET) (T/ET) is of great significance for understanding the mechanisms of ecosystem water distribution and energy partitioning. However, there is a lack of knowledge on the long-term variations in T and ET in high-altitude subalpine regions with low temperature and high humidity. Therefore, the T and ET of a subalpine coniferous forest in Mount Gongga was simulated during the growing season from 2005 to 2021 using three machine learning models and a generalized nonlinear complementary principle model. Results showed that the machine learning models performed better in simulating T than the often-applied Penman-Monteith model. The mean daily and growing season T were 1.18 ± 0.14 mm d-1 and 217.60 ± 17.76 mm yr-1, respectively. There was a decreasing trend of T during 2005–2021, with a rate of -2.46 mm yr-1 (P < 0.05). Variation in T was mainly influenced by net radiation, wind speed, vapor pressure deficit, and relative humidity, and the magnitude of these effects varied at different temporal scales (daily, monthly, and annual). Mean growing season T/ET was 0.46 ± 0.03. There was no significant trend in T/ET before 2016 (P > 0.05), but the T/ET significantly decreased thereafter with a rate of 0.01 yr-1 (P < 0.05). There was no significant difference in T/ET among years with different precipitation at our study site which had always abundant precipitation of >1400 mm. Changes in T/ET were more sensitive to air temperature, and the effect of meteorological factors on T/ET varied at daily, monthly and annual time scales. The decrease in T/ET was primarily due to the continuous increase in temperature in recent years. Our findings indicate that future climate warming will lead to an increase in water resources in subalpine humid regions.
Keywords: Subalpine coniferous forest; Machine learning; Penman-Monteith equation; Water relation