A new study by Columbia Engineering – the first to investigate the long-term effects of dryland soil moisture-atmosphere feedback – shows that soil moisture provides negative feedback on the availability of dryland surface water, offsetting some of the anticipated decrease.
Researchers have predicted that global warming would increase the availability in wet regions of surface water-freshwater supplies resulting from precipitation minus evapotranspiration-and decrease the availability of water in dry areas.
This expectation is focused mainly on atmospheric thermodynamic processes.
When temperatures in the air increase, more water evaporates into the air from the ocean and soil.
Since warmer weather can contain more water vapor than dry air, the current trend of water availability is expected to be enhanced by a wetter environment, triggering the “dry-get-dry and wet-get-wet” responses of the atmosphere to global warming.
A team of engineers from Columbia, led by Pierre Gentine, Maurice Ewing and J. Lamar Worzel, Professor of Earth and Environmental Engineering and member of the Earth Institute, asked why coupled climate model forecasts do not project substantial “dry-get-dry” responses over arid, tropical and temperate regions with an aridity index of less than 0.65, even though researchers are using the high-emission global warming scenario. Sha Zhou, a Lamont-Doherty Earth Observatory and Earth Institute postdoctoral researcher who studies land-atmosphere interactions and the global water cycle, indicated that input from soil moisture-atmosphere may play an important role in future dryland water availability predictions.
The new research, published today in Nature Climate Change, is the first to demonstrate the significance for these projections of long-term shifts in soil moisture and related soil-moisture-atmosphere feedback.
In addition to what would be expected in the absence of soil moisture feedback, the researchers established long-term control of soil moisture through air circulation and moisture transport that largely mitigates possible declines in future water availability in drylands.
“These feedbacks play a more significant role in long-term surface water changes than thought,” Zhou says. “Since changes in soil moisture negatively affect water supply, this negative feedback may also partially reduce the warming-induced increase in the magnitude and frequency of extreme high and low hydroclimatic events, such as droughts and floods.
We may experience more frequent and severe droughts and floods without negative feedback.
With a novel statistical method they developed for the research, the team integrated a special, idealized multi-model land-atmosphere coupling experiment.
The algorithm was then applied to observations to explore the crucial role of soil-moisture-atmosphere feedback in future changes in dryland water availability and to explore the thermodynamic and dynamic processes underlying future changes in the availability of water due to these feedbacks.
Important decreases in the availability of surface water (precipitation minus evaporation, P-E) were observed in response to global warming over oceans in dry regions, but only minor decreases in P-E over dry regions. Zhou proposed that land-atmosphere processes are connected to this phenomenon. “Over drylands, soil moisture is expected to decrease significantly under climate change,” she explained. “Changes in soil moisture would further affect atmospheric processes and the hydrologic cycle.”
It is expected that global warming will decrease the supply of water and therefore soil moisture in drylands.
However, this new study found that drying soil moisture actually adversely affects the availability of water – reducing soil moisture decreases evapotranspiration and evaporative cooling, and increases dryland surface warming relative to wetlands and the ocean.
The contrast of land-ocean warming amplifies the variations in air pressure between the ocean and the land, contributing to greater wind drift and transfer of water vapor from ocean to land.
Our Arb, Our Arb