By CHARLES J. HANLEY, AP Special Correspondent

To the quiet green solitude of an English country estate they retreated, to think the unthinkable.

Scientists of earth, sea and sky, scholars of law, politics and philosophy: In three intense days cloistered behind Chicheley Hall’s old brick walls, where British saboteurs once secretly trained, four dozen international thinkers pondered the planet’s fate as it grows warmer, weighed the idea of reflecting the sun to cool the atmosphere, debated the question of who would make the decision.

The unknown risks of “geoengineering” — in this case, tweaking Earth’s climate by dimming the skies — left many uneasy.

“If we could experiment with the atmosphere and literally play God, it’s very tempting to a scientist,” said senior Kenyan earth scientist Richard Odingo. “But I worry.”

Arrayed against that worry is the worry that global warming at some point — in 20 years? 50 years? — may abruptly upend the world we know, by melting much of Greenland into the sea, by shifting India’s life-giving monsoon, by killing off marine life.

If climate engineering research isn’t done now, climatologists say, the world will face grim choices in an emergency. “If we don’t understand the implications and we reach a crisis point and deploy geoengineering with only a modicum of information, we really will be playing Russian roulette,” said Steven Hamburg, a U.S. Environmental Defense Fund scientist.

The question’s urgency has grown as nations have failed, in years of talks, to agree on a binding long-term deal to rein in their carbon dioxide and other greenhouse-gas emissions blamed for global warming. The Intergovernmental Panel on Climate Change (IPCC), the U.N.-sponsored science network, foresees temperatures rising as much as 6.4 degrees Celsius (11.5 degrees Fahrenheit) by 2100, swelling the seas and disrupting the climate patterns that nurtured human civilization.

Science committees of the British Parliament and the U.S. Congress urged their governments last year to look at immediately undertaking climate engineering research — to have a “Plan B” ready, as the British panel put it, in case the diplomatic logjam persists.

Britain’s national science academy, the Royal Society, subsequently organized the Chicheley Hall conference with Hamburg’s EDF and the association of developing-world science academies. From six continents, they invited a blue-ribbon cross-section of atmospheric physicists, oceanographers, geochemists, environmentalists, international lawyers, psychologists, policy experts and others, to discuss how the world should oversee such unprecedented — and unsettling — research.

An Associated Press reporter was invited to sit in on their discussions, generally off the record, as they met in large and small groups in plush wood-paneled rooms, in conference halls, or outdoors among the manicured trees and formal gardens of this 300-year-old Royal Society property 40 miles northwest of London, a secluded spot where Britain’s Special Operations Executive trained for secret missions in World War II.

Provoking and parrying each other over questions never before raised in human history, the conferees were sensitive to how the outside world might react.

“There’s the `slippery slope’ view that as soon as you start to do this research, you say it’s OK to think about things you shouldn’t be thinking about,” said Steve Rayner, co-director of Oxford University’s geoengineering program. Many geoengineering techniques they have thought about look either impractical or ineffective.

Painting rooftops white to reflect the sun’s heat is a feeble gesture. Blanketing deserts with a reflective material is logistically challenging and a likely environmental threat. Launching giant mirrors into space orbit is exorbitantly expensive.

On the other hand, fertilizing the ocean with iron to grow CO2-eating plankton has shown some workability. Massachusetts’ prestigious Woods Hole research center is planning the biggest such experiment. Marine clouds are another route: Scientists at the U.S. National Center for Atmospheric Research in Colorado are designing a test of brightening ocean clouds with sea-salt particles to reflect the sun.