Monitoring evapotranspiration on a green roof : feedback from two summer periods.

The greening of cities has become a major component of current urban development policies. This greening is taking new and variable forms, and is increasingly associated with rainwater management techniques. This vegetated surfaces increase can potentially encourage evapotranspiration. Increasing this process has a twofold advantage. On the one hand, It provides stormwater runoff reduction benefits and, on the other, it promotes cooling in the urban environment. However, in order to better quantify and optimise evapotranspiration, it is necessary to be able to assess it for different kind of urban surfaces. In an urban environment, where surfaces are very heterogeneous, it is therefore necessary to have continuous measurements, over the long term (several seasons) and, if possible, on different types of surfaces with relatively small areas: just a few dozen m².

In order to document the capacity of urban vegetated surfaces to evapotranspire, a study carried out in 2022 and 2023 on a green roof, emblematic of urban greening solutions, tested the Eddy Covariance (EC), energy budget closure (EB) and a transpiration chamber (Ch) methods for measuring evapotranspiration on this type of surface and continuously estimated the evapotranspiration of this roof. Moreover, with the aim of eventually being able to compare the evapotranspiration of different urban vegetated  surfaces, we also looked at the evaporative fraction in relation to water availability and net radiation.

The results show that EB method tends to overestimate evapotranspiration in relation to the Eddy Covariance, whereas Ch tends to underestimate it. The evaporative fraction of this green roof is generally quite low, averaging 0.2, but can exceed 0.5 on some days. This evaporative fraction is also highly variable over the measurement period.

This shows that for this type of vegetated surface, their capacity to use the energy available for evapotranspiration is generally quite low and not constant. While a higher water content favours high evaporative fractions, this is not always sufficient. Average net radiation of at least 300W.m-2 also seems necessary. If these conditions are met, there must also be other conditions favourable to evapotranspiration, not observed here, but linked to the physiology of the vegetation.