Stratospheric Aerosol Climate Intervention Designed to Minimize Negative Impacts

Abstract:

Solar radiation management, by means of artificial stratospheric aerosols has been suggested as a response to global warming. We examine some of the potential impacts, particularly impacts on agriculture, through specific climate model and agricultural model simulations. Previous climate model experiments have used temperature as the metric for the desired climate, but here we use direct impacts as metrics, starting with agriculture. Later we will use Sustainable Development Goals as a guide to address other important impacts, including on human health, water resources, fisheries, and ecosystems.

We use the NCAR Community Earth System Model, version 2, using the Whole Atmosphere Community Climate Model version 6, for simulations using feedback control on sulfur emissions to the stratosphere, with impacts as metrics. Because this state-of-the-art model includes complete treatment of diffuse radiation, aerosols, and atmospheric chemistry, as well as built in simulations of crop production, it is ideal for this purpose. We know that if warming from greenhouse gases is countered with reduction of solar radiation to cancel out the globally-averaged radiative forcing, the planetary temperature can be kept from changing, but the hydrological cycle will be weaker, with a reduction of globally-averaged evapotranspiration and precipitation, particularly in summer monsoon regions. Temperature, precipitation, and insolation changes, as well as enhanced ultraviolet radiation, enhanced diffuse radiation, changes of surface O₃ concentration, and CO₂ will all impact crops. We start by examining previous simulations to develop a transfer function between stratospheric aerosol loading and various metrics of agricultural productivity. These include global caloric intake from major crops, as well as regional food availability. For model runs that did not include a crop model in the past, we use the Community Land Model offline, driven by climate model output.