Forest mist

Prof Chris Huntingford explains more about new research showing the importance of plant transpiration in earth system models…

A paper in Nature Climate Change, led by Peking University and also involving the Centre for Ecology & Hydrology, looks at the performance of climate models in estimating the rate at which trees and plants can transpire.

Rainfall levels control how much water reaches the land. In any natural system, this water then takes one of the following routes: It either enters the soil and, then, via roots, trunks, branches or stems, is drawn up to leaves and transpired. Alternatively it is released by direct evaporation from either the soils, or off leaves following rainfall (the latter called "interception loss"). Any remaining water enters the river systems.

Understanding the fraction of water re-entering the atmosphere by plant transpiration, rather than by soil evaporation or interception loss, is important. Although all three evaporative mechanisms do return water to the atmosphere, the timescales are different – and this could impact on meteorological conditions, such as subsequent rainfall features.

"Understanding the fraction of water re-entering the atmosphere by plant transpiration, rather than by soil evaporation or interception loss, is important."

Hence, with so much interest in how rainfall patterns might change under global warming, computer models that predict future climate must depict accurately the ratio of transpiration (T) to the sum of all three evaporation quantities (evapotranspiration, ET).

Using measurements from multiple field sites, and comparing to a large set of available climate models, we find that these numerical computer simulations persistently underestimate transpiration.

"...we find that these numerical computer simulations persistently underestimate transpiration."

In our paper, we suggest a range of possible reasons for this. These include the need for better descriptions of energy exchanges within tree canopies, and how improvements might be needed to better describe interception losses. Achieving this will contribute towards making ever more accurate simulations of the current climate.

As confidence is built, this then allows better assessment of water exchange between the land and atmosphere, and how this might evolve in the decades ahead.

The analysis was led by Chinese authors from the College of Urban and Environmental Sciences at Peking University in Beijing, but working with others from the UK, France and Australia. Both the lead author, Xu Lian, and another co-author from China, Hui Yang, have recently spent time at the UK Centre for Ecology & Hydrology.

Chris Huntingford

Additional information

Full paper reference: Lian, X et al. Partitioning global land evapotranspiration using CMIP5 models constrained by observations. Nature Climate Change volume 8, pages 640–646 (2018). DOI: 10.1038/s41558-018-0207-9

See also a Nature Climate Change News & Views article about the research: Plants turn on the tap

Science areas: 

Issues: