The goal of this project was to determine if there is more or less evaporation1 from the land surface when dryness in the the air2 increases. This is a particularly relevant problem because atmospheric aridity is expected in increase in the future, and it is an interesting problem because there are two competing factors:
Plants can sense increasing dryness in the air and close up the pores (stomata) on their leaves to conserve water for later use. This closure reduces evaporation.
Drier air naturally demands more water from the land surface, encouraging evaporation.
So the answer to whether evaporation increases or decreases in response to drier air comes down to which effect dominates: plant response (reduces evaporation) or atmospheric demand (increases evaporation). To test which effect dominates we developed a very simplified analytical model based on recent results blending observations and plant theory. According to the model, ecosystems are capable of a broad range of behavior in response to increased atmospheric dryness, from strongly reducing evaporation to allowing large increases evaporation. Ecosystem behavior depends both on environmental conditions and plant type. For example, to access carbon in the air plants must keep stomata open. So, plants that were bred or evolved to prioritize carbon gain (e.g. growth) over water conservation will tend to keep stomata open even as aridity increases. For these plant types (e.g. many crops), evaporation will likely increase in response to increasing aridity. However, other plant types like many conifer trees prioritize resilience to aridity over growth, and are more likely to close their stomata in response to aridity. For these plant types, evaporation will likely decrease in response to atmospheric aridity. See the paper for more detail: for example the role of the environment, etc.
As an aside, the design of the project allowed us to test a range of proposed mathematical models for plants' stomatal behavior. While in our analyses we used a newer model by Medlyn and coauthors that blends theory and observations, we found that had we used different models our results would have been drastically different: the shape of the dryness-evaporation curve completely changes depending on model choice! We would hope that different models of the same physical processes would give qualitatively similar results (that is, they would only differ in the details), but this is not the case. This is all somewhat troubling, especially because different land surface and earth system models ("climate models") vary in their choice of stomatal model, and there is no consensus on which model is most representative of plant behavior (I would guess that it even depends on the plant type and environmental conditions). To me, the identification of this divergence (and contradiction) in stomatal model behavior was our most noteworthy result. It really highlights the need for future research understanding plant response at the ecosystem scale: plant response to dryness is very much an open area of research. I would be skeptical of anyone with strongly held convictions on the nature of the evaporation response to increased atmospheric dryness! Hopefully in a few years we as a community can reach consensus on how to model plant response, and we can redo our analysis; we might even get very different results :).
I am too technologically illiterate to set up a comment system on this page, but comments and questions are very welcome and encouraged through Github’s issue system: just click here! (I know it’s kind of a hack but it should work well enough.)
Evaporation as used here also includes water released by plants, which is often called "transpiration" in some circles. For more discussion on these terms, I recommend Miralles and coauthor’s really nice note on the topic.↩︎
"dryness" as defined by vapor pressure deficit (VPD). In more familiar terms, as relative humidity decreases, vapor pressure deficit increases. However, if temperature increases and relative humidity stays the same, vapor pressure deficit (dryness) also increases. More technically: VPD = (1−RH)es, where es is the saturation vapor pressure of water, which is an exponentially increasing function of temperature.↩︎