What We Publish

Crop yields in a geoengineered climate

Julia Pongratz, David B. Lobell, Long Cao, and Ken Caldeira · Nature Climate Change 2, 101-105 · 2012

Key finding. In climate-model simulations with sulphate aerosol added to the stratosphere to hold global mean temperature at control levels under doubled carbon dioxide, global mean yields rose relative to the unmanaged doubled-carbon-dioxide case by 14% for maize, 21% for wheat, and 8% for rice — 86, 99, and 25 million tonnes of production respectively — because temperature stress was removed while carbon dioxide fertilization was retained.

Two stacked line plots against latitude from 60 degrees south to 80 degrees north, with maize in red, wheat in blue, and rice in black, showing the difference between the solar geoengineering case and the doubled-carbon-dioxide case. The upper panel is yield change in per cent, positive almost everywhere and typically between 10 and 30 per cent, dipping below zero only for rice above about 35 degrees north. The lower panel is production change in millions of tonnes, near zero across the tropics and rising to a sharp maize peak of about 18 million tonnes near 40 degrees north, with wheat peaks either side.
Adding solar geoengineering to a doubled-carbon-dioxide world raises yields at nearly every latitude (upper). The production gain (lower) is concentrated in the northern mid-latitude maize belt, where the doubled-carbon-dioxide climate would otherwise have caused the largest losses; the one substantial yield loss, high-latitude rice, barely registers because little rice is grown there. Figure 3 from Pongratz et al. (2012), Nature Climate Change 2, 101-105. Reproduced under author reuse rights. Extracted from the published PDF and resized for web display.

What question did this research address?

Deflecting sunlight away from the Earth can hold global mean temperature steady in a high-carbon-dioxide world, but it changes regional climate and precipitation in ways that temperature stabilization alone does not describe. A prominent objection to solar geoengineering is that these changes would threaten the food and water supply of billions of people.

This paper asked what solar geoengineering would actually do to the yields of the three crops that supply most of the world's calories — wheat, maize, and rice — relative both to today's climate and to an unmanaged high-carbon-dioxide climate.

What did we find?

Three quasi-equilibrium simulations were run — a control at about 400 parts per million, a doubled-carbon-dioxide case, and a doubled-carbon-dioxide case with stratospheric sulphate raised enough to return global mean temperature to control levels. Yields were then estimated from published regressions of historical weather against country-level yields, with carbon dioxide fertilization added separately.

Relative to today's climate, doubling carbon dioxide changes global yields only modestly — maize down 3%, wheat up 6%, rice up 19% — as climate damage and carbon dioxide fertilization partly cancel. Warming, not precipitation change, drives most of the climate-induced losses.

Relative to today's climate, the geoengineered case raises yields for all three crops — maize 11%, wheat 26%, rice 28% — with climate-induced changes within plus or minus 1% at the global mean, so nearly all the gain comes from carbon dioxide fertilization.

The comparison that matters for policy is geoengineering against the unmanaged high-carbon world, and there the gains are 14% for maize, 21% for wheat, and 8% for rice. The large maize production losses in the northern mid-latitudes under doubled carbon dioxide are averted.

The one substantial loss is high-latitude rice, where cold-temperature limits are no longer relieved by warming. It does not show up in production totals because little rice is currently grown there.

The simulations do reproduce the expected weakening of the Asian summer monsoon — precipitation down about 10% in CAM3.5 and 14% in HadCM3L — but the resulting yield reductions are only a few per cent and are partly or fully offset by the averted warming.

Doubling and halving the assumed sensitivity of yields to temperature and precipitation changes the size of the response but not its sign.

Effects of the reduced and more diffuse sunlight itself were not simulated. A linear scaling of yield with insolation would imply a 2.2% loss, comparable to the tropical gains but small against the global mean gain, and increased diffuse light would work in the opposite direction.

Why does it matter?

It tests a specific and widely repeated objection to solar geoengineering and does not find support for it at the scale of latitude bands. The concern that stabilizing temperature would devastate global agriculture is not what these models produce.

The reason is worth stating plainly, because it also marks the limits of the result: the gains come mostly from carbon dioxide fertilization, which solar geoengineering does not remove. The technique is being credited here with letting a high-carbon world keep the fertilization benefit while shedding the heat.

A result at the global and latitudinal mean does not settle local food security. Small regions can move much more than the band containing them, which matters where subsistence farming leaves little room to adapt.

The paper's own conclusion is that solar geoengineering does nothing about ocean acidification and carries anticipated and unanticipated risks, so the most certain way to reduce climate risks to food security remains cutting greenhouse gas emissions.

Citation

Julia Pongratz, David B. Lobell, Long Cao, and Ken Caldeira (2012). Crop yields in a geoengineered climate. Nature Climate Change 2, 101-105.

Related