ABSTRACT

The land surface interacts strongly with the atmosphere above it, as the atmosphere supplies water to the surface in the form of rain and energy, including sunlight and downwelling infrared radiation. The land in turn affects the atmosphere by providing water vapor through evaporation and transpiration, giving off sensible heat and upwelling infrared radiation, and blocking the wind with trees and other obstacles, among other effects. Land-atmosphere interactions are thus an important topic in climate science, and a key goal research in this area is to understand the feedback mechanisms through which land-surface processes influence the atmosphere in ways that produce further effects on the land and vice versa. Much of the work in this area is focused on precipitation and soil moisture, particularly the extent to which evaporation serves as a source for later precipitation which further controls the amount and distribution of soil moisture.

Here the PIs go beyond soil moisture-precipitation feedback to consider mechanisms that link land surface characteristics to cloudiness and the subsequent shading effect of cloud cover on the surface. One of these is a feedback in which sunlight falling on moist soil produces evaporation, which leads to the formation of clouds or fog, which shades the soil and limits further evaporation. Previous work by the PIs suggests that this negative feedback mechanism plays an important role in limiting evaporation in the Amazon during the rainy season. An additional question pursued in this research is the extent to which small-scale differences in surface cover, such as exist between adjacent forested and deforested patches of the Amazon, produce differences in cloudiness as near-surface air converges into and rises above drier and hence warmer patches.

A key concern in studying such effects is that climate models have limited ability to represent them. Climate models rely on parameterizations to represent clouds and precipitation, and parameterizations have difficulty capturing the diurnal cycle of cloudiness. This is a severe limitation for studying the effect of cloud shading on evaporation, as the effect depends on whether clouds develop when the sun is high in the sky or near or below the horizon. Clouds simulated in climate models are also unlikely to respond to small-scales patchiness in surface cover, as models only represent aggregate cloud cover and surface conditions over grid boxes which extend at least tens of kilometers in each direction.

The PIs use two separate modeling strategies to circumvent these difficulties, the first of which is a limited domain cloud resolving model (the Weather Research and Forecasting model, or WRF) constrained to relax back to a specified background temperature profile. This configuration is based on the weak temperature gradient (WTG) approximation, which assumes that temperatures well above the surface are horizontally uniform due to the weakness of the Coriolis force over tropical regions such as the Amazon. The WRF-WTG framework allows for very high resolution simulations (grid spacing of one or two kilometers) over limited domains on which the processes of interest can be represented with some realism. The second approach uses a technique known as superparameterization, in which a somewhat simplified cloud resolving model is placed in each grid column of a climate model, creating a hybrid model which represents both the cloud scale and the large scale (see AGS-0425247).

Using these two modeling strategies the PIs perform a number of model experiments to determine the effects of the proposed mechanisms, including experiments in which the land surface turbulent heat flux is prescribed and simulations in which the diurnal cycle of land surface fluxes is reduced by imposing a very large soil heat capacity. The model experiments are complemented with analysis of relevant observations from a number of observing stations in the Amazon, some in deforested regions and some representing the transition from wetter to drier conditions.

The research has societal value as well as scientific interest, as it seeks to improve understanding of climate variability and change in the Amazon, a region of high biodiversity which plays a substantial role in the global water and carbon cycles. In addition, a variety of education and outreach activities are organized around the work, including work with high school students in Harlem, work with a STEM center housed at Cal State Los Angeles, and an undergraduate recruitment effort through the Research in Science and Engineering (RiSE) program at Rutgers. The project also provides support and training for a graduate student and a postdoc.

Please report errors in award information by writing to: awardsearch@nsf.gov.