COS 98-5 - Effects of hot vs. dry vapor pressure deficit on ecosystem carbon and water fluxes

Abstract: Vapor pressure deficit (VPD), the difference between saturated and ambient water vapor pressure, defines atmospheric demand for moisture and exerts strong influence on plant fluxes of carbon and water. VPD has increased across the majority of the vegetated land surface and will likely continue to increase with further climatic change, largely driven by warming temperature. However, most leaf- and plant-level experiments examining the effects of VPD have manipulated it by changing humidity rather than temperature. Further, most large-scale empirical studies have considered VPD independent of its component temperature and humidity drivers. To understand the response of plant functioning to increasing VPD in a warming world, we must disentangle VPD’s temperature- vs. humidity-driven effects. In this research, we use carbon and water fluxes measured at diverse AmeriFlux sites across the U.S.A., along with site-measured meteorology and SMAP satellite rootzone soil moisture estimates, to address two main questions: (1) to what extent have similar historical VPD conditions been characterized by variable air temperature and relative humidity?, and (2) how are ecosystem carbon and water fluxes influenced by hotter vs. drier VPD of the same magnitudes? While VPD is mechanistically related to both temperature and humidity, sites experience considerable variability of temperature and relative humidity at a given magnitude of VPD. Specifically, over half of a site’s full temperature range and over a third of its humidity range were typically present within a particular observed magnitude of VPD, corresponding to absolute temperature ranges of 2.3 – 11.5 degrees C and relative humidity ranges of 8.7 – 41.1%. Sites with high mean annual temperature and low mean annual precipitation experienced the narrowest ranges of temperature and relative humidity at a particular magnitude of VPD. We also found strong and consistent statistical support for an interaction between VPD, temperature, and soil moisture influencing carbon (net ecosystem exchange, gross primary productivity) and water fluxes (latent heat) across diverse vegetation types. Carbon and water flux response variables typically varied by 12 – 25% across temperature for a given VPD, with the strongest interactions present at mid-ranges of VPD. Changing effects of VPD on ecosystem function depending on its “hotness” vs. “dryness” imply a nonstationary relationship between ecosystem function and VPD, with important implications for the role of atmospheric dryness in a warming world.