April 2016

Yale’s Climate Community Considers Africa

By Eric Ellman, Yale Climate & Energy Institute

Eric Ellman is an environmental scientist and features writer for AR3.

Tropical meteorology is one of the great remaining mysteries of Earth science, says Yale scientist Bill Boos. He doesn’t call it “THE” great mystery, but you suspect that’s how he feels.

Billions of lives depend on rain-fed agriculture. For Boos, a Yale associate professor who uses math and fluid mechanics to numerically model continental-scale tropical weather, global warming prompts humanitarian concerns. His colleague Matt Huber, director of the Purdue University and University of New Hampshire Climate Dynamics Predictions Laboratory, feels them too. Using similar rigorous computational tools, Huber develops high-resolution maps that predict where temperature and humidity become lethal.

Catastrophes in East Africa and the Middle East justify their efforts. Prolonged drought preceded Europe’s current refugee crisis. A severe shortage of water also played a role in the death of 400,000 Ethiopians during famine from 1983-1985.  In 2015 the UN warned of jeopardized food supplies for 15,000,000 in the Horn of Africa. Across huge swaths of the planet, where multitudes lack air conditioning and rely on capricious seasonal rains, what additional risk does climate change pose?

The Intergovernmental Panel on Climate Change (IPCC) – the world’s largest consortium of climate scientists – offers little help to those planning regional interventions in East Africa and the Middle East. While science has helped develop a consensus on the climate future of North America and Europe, Africa is not so lucky.

Boos cites multiple reasons for the ‘hole’ in climate-science that exists in Africa. For one, there is less meteorological data to begin with, and the complexity of the continent’s geography is confounding.  Sandwiched between well-formed monsoon systems in West Africa and Asia, drought-prone East Africa is impacted by not one system, but both. Describing all of this in mathematical models is part of the challenge Boos is taking on.

But understanding complicated meteorology is only part of the problem. Making use of that information is another. Existing IPCC climate models present the Earth’s surface as a nested set of boxes measuring more than 100 kilometers to a side. At such coarse resolution, a small nation like Djibouti, for example, is reduced to just a couple of pixels; a scale that precludes informed decision-making related to irrigation, agriculture, hydroelectric and other infrastructure. As the region anticipates billions of dollars in funding for climate-related adaptation over the coming decades, climate models with greater accuracy and precision are desperately needed.

Step one: Understand tropical weather

Boos, who made his reputation using fluid mechanics to challenge prevailing notions about the Asian monsoon, now uses the same tools to run the multitudinous calculations required to understand a different weather feature: The Intertropical Convergence Zone.

The  ITCZ is a rising conveyor belt of warm air that develops near the Equator where two enormous rotating atmospheric cells meet. The air sheds its excess moisture as it cools, returning to earth thousands of miles away near the Tropics of Cancer and Capricorn, flowing back towards the Equator. The two systems (called “Hadley Cells” after George Hadley who first described them in 1735) circulate like opposing pinwheels. If you were driving south in Kenya, predominant tailwinds would become headwinds as you crossed the ITCZ. Beneath the ITCZ you would experience rain; a few thousand miles in either direction, where desiccated air descends back to Earth, it would be dry.

Current theory uses an energy budget to describe seasonal migration and the wet and dry zones of the ITCZ. As ice retracts in the Northern Hemisphere each summer, less sun reflects back to space, and the reduced temperature gradient draws the convergence zone north. Over the skies of Africa, the changed gradient brings rain to the Sahel.

In reality, many factors impact tropical precipitation. Prediction is complicated by a lack of meteorological stations and historic records, which is not at all surprising in a region with many competing priorities. In the meantime, certain generalizations have emerged to describe how global warming impacts hydrology. “The rich get richer, and the poor get poorer,” is an ironic refrain scientists borrow to describe how regions that currently get rain will get more, and those that don’t will get less.

But even that aphorism is more consistent over the world’s oceans than over its landmasses.  Above Africa, the future is even less clear: the most recent IPCC Assessment Report shows stark disagreement regarding which regions will grow wetter and which will grow drier in decades ahead. Imagine the consternation of regional planners: what guides long-range infrastructure decisions when it is unclear whether the future will be wet or dry?

Step two: Develop “actionable” science

Beyond forecasting how the ITCZ will wander, and how changes in the West African and Asian monsoons will impact African precipitation, climate model output should support societal needs.  Scientific tools ought to address pressing policy questions, like “what to grow in the Sahel (and when)”, or “where to build a hydroelectric dam in Ethiopia (and how high).”

Huber believes we are on the cusp of building tools that do.  Moore’s Law, which predicts that computer processing speed will double every 18 months, is catching up with the expansive computational demand of tropical weather prediction.

The first IPCC Assessment Report, published in 1990, depicted future climate trends on a 500km grid scale. Continental features that it outlined were as indistinct as constellations. Subsequent reports reflected increasingly faster and cheaper computer processing, enabling more complex analysis at ever-finer resolution. By the 2007 Assessment Report IV, the grid scale had shrunk to 110km. Mountain chains were now discernible.

Last May, Huber and Boos demonstrated opportunities to further increase the accuracy and precision of climate modeling as part of a Yale-led delegation to the East Africa Risk and Opportunities Summit in Djibouti. Before hundreds of delegates and cabinet-level officials from East Africa and the Arabian Peninsula, they shared how banks of super-computers, could simulate convective processes on a grid scale measuring just 3-4km to a side, in a process known as ‘dynamical downscaling’. Huber’s time slice of a typical end-of-century day in the Arabian Peninsula demonstrates how future heat and humidity will describe the limits of outside human activity.

In anticipation of developing such high-resolution climate assessments, Yale Climate and Energy Institute (YCEI) postdoctoral researcher Srinath Krishnan has already developed the first “mechanistic” model that utilizes Huber’s high-resolution climate output for another purpose. Krishnan’s Lyme disease model incorporates life cycle knowledge of the Blacklegged tick and other vectors of the disease to generate future range maps for New England, where Lyme disease is a prime health and safety concern. The analog for Africa would be maps that project where future temperature and humidity foster conditions for the vectors of malaria, dengue fever, sleeping sickness and other maladies.

Africa may be one of Earth’s obvious Achilles’ heels with respect to global warming, but the tools described above are applicable globally.