They let the iodine go on purpose, and then they went looking for it.

December 2, 1949.
46.55° N, 119.64° W

The Columbia River comes down out of Canada, runs south through Washington, and makes a wide bend east of the Cascades. The land inside that bend is desert. Sagebrush, cheatgrass, less than a foot of rain in a year, wind most afternoons. What made it farmland was the river and the pumps.

Two towns sat on the north bank. White Bluffs and Hanford. A few hundred people in each, a ferry, a bank, a high school, a Grange hall. They grew soft fruit under irrigation, cherries and apricots and peaches, the kind of crop that has to be picked in a narrow window and moved fast. Some of the families had been on that ground since before the First World War.

In the spring of 1943, the federal government took all of it. Condemnation notices went out. The land was required for a war project, and the notices did not say what the project was. Families were given weeks. The fruit was still on the trees when they left.

They scattered into Yakima and Prosser and the Willamette Valley and wherever else there was work. The orchards were pushed down. Some of the buildings were used and the rest were knocked flat, and the whole stretch of country became the Hanford Engineer Works, a federal reservation the size of a small county with guards on the roads into it.

Its product was plutonium.

The reactors went along the river, because a reactor of that design needed enormous volumes of cold water and the Columbia had it. The chemical plants went inland, miles back into the desert, in a grid the maps called the 200 Areas. The building that did most of the early work there was T Plant. It was a windowless concrete box several hundred feet long, with walls thick enough to hold in what happened inside, and the men who ran it worked everything by remote handling and watched through periscopes.

The work was called separation. Uranium slugs came out of the reactors after irradiation, fiercely radioactive, and went into nitric acid. The plutonium was pulled out of the resulting solution in stages, and what was left was waste.

Dissolving irradiated uranium releases gas. One of the gases is iodine 131. Its half-life is eight days, which makes it intensely radioactive and short-lived at the same time. It goes up the stack, drifts with the wind, and settles on whatever is underneath. Grass, mostly. Cows eat grass, and the iodine passes into the milk. In a body it goes to the thyroid, and it delivers the highest dose to the smallest thyroids, which belong to children.

During the war, none of that slowed anything down. The stacks ran. The plutonium in the Trinity test and in the bomb dropped on Nagasaki came out of those buildings.

Afterward, the operating practice tightened. Fuel was left to sit for months after it came out of a reactor, so the short-lived iodine could decay before the metal was dissolved. Filtration went onto the stacks. Health physics crews sampled air, vegetation, and milk on the site and in the towns around it, and the readings went into files nobody outside the fence would see.

On September 3, 1949, an Air Force weather reconnaissance aircraft flying a route over the North Pacific came back with radioactive material on its filter paper. More flights confirmed it. The debris was analyzed, and the analysis said the Soviet Union had detonated a plutonium device in late August. President Truman announced it on September 23. The American monopoly had lasted a little over four years.

The announcement told the public what had happened. It did not tell anyone in Washington what they most wanted to know, which was how much more was coming.

A plutonium program has to reprocess fuel, and reprocessing puts fission products into the air above the plant that does it. If an aircraft could fly a sampling run downwind of a Soviet separations plant, and if the isotopes caught on its filters could be worked backward, the United States could estimate how much material the Soviets were separating and how fresh their fuel was. The technique existed on paper. It had never been checked against a release of known size.

Only one place in the world could produce a release of known size with every variable written down in advance. In the fall of 1949, the plan was drawn up at Hanford and given a name in the files. Green Run. Green meant the fuel: short-cooled, roughly sixteen days out of the reactor instead of the usual ninety or more, with the iodine still near full strength. The dissolving would go ahead without the stack filtration in line. Aircraft and ground teams would follow the plume and measure it.

The operation needed weather. The planners wanted stable air and a wind that would carry the plume in a predictable direction, far enough and steady enough for the instruments to stay with it. The forecast for the night of December 2 was judged good enough.

The dissolving started that night and ran into the next day.

The men on the ground crews were spread out across the basin in trucks with survey meters, and most of them had been told only where to stand and what to write down. The aircraft flew their patterns. The plant did what the plan said.

The forecast was wrong. The air over the basin did not lift and disperse. An inversion settled and held the plume low, and weather moving through brought rain and snow that pulled the material out of the air and put it on the ground. The plume went south and east, over Richland, over Kennewick and Pasco, over Walla Walla, over the pasture and the wheat and the dairy farms in between.

Hanford's own monitoring picked it up. Vegetation samples from the towns downwind came back above the levels the site's own staff used as limits. The numbers went into reports, and the reports were classified.

Nobody under the plume was told. Not that week, not that year. Farms ran their winter routine. Herds were milked. Milk from cows that had grazed downwind went into cans and onto trucks and into refrigerators in Pasco and Kennewick and the small farm towns north of them, and children drank it, because milk is what children drank.

Hanford kept producing for decades. Reactors were added along the river and eventually shut down one at a time. The downwind towns grew. Over the years, families in the farm country north and east of the site accumulated a set of medical problems they had no framework for. Thyroid disease turned up again and again. So did miscarriages, birth defects, and cancers in people who had never worked around anything worse than a tractor. In Franklin County there was a stretch of country road that the people living on it started calling by a grim nickname, because of how many of the houses along it had somebody sick in them.

The people making those observations had nothing to point at. The site's operating records were classified. Its release figures were classified. Green Run was classified.

That held for thirty-six years.

In February 1986, after sustained requests under the Freedom of Information Act from a downwinders' group and from reporters covering the site, the Department of Energy released thousands of pages of Hanford's historical operating records. The documents laid out what the plant had put into the air and the water, year by year, going back to the war. A reporter for the Spokane Spokesman-Review worked through them. Green Run was in the stack.

The reaction ran through the Tri-Cities and the farm counties for months. People who had spent thirty years being told there was nothing over their heads now had page numbers. Public meetings filled. State health officials who had been assuring residents for years found themselves reading the same documents for the first time.

What the papers also showed was that the deliberate release had not been the largest one. The wartime and early postwar years had put far more iodine over the same farmland, month after month, as an ordinary byproduct of making plutonium on a deadline. Green Run stood out for a different reason. It was done on purpose, on a schedule, with people positioned in advance to measure what happened.

The state and federal response was to commission studies. A dose reconstruction project spent years modeling how much radiation individual people downwind had received, working from where they had lived, what they had eaten, and how much milk they had drunk. The Hanford Thyroid Disease Study, run by the Fred Hutchinson Cancer Research Center, examined people born in the downwind counties during the peak release years. Its findings, reported at the end of the 1990s, did not establish a link between the estimated doses and thyroid disease. Many of the people who had been studied rejected the conclusion, and the argument over the modeling has never closed.

Downwinders filed suit in 1990. The consolidated litigation ran for a quarter century. A small number of plaintiffs won at trial, verdicts were appealed, most claims were dismissed or dropped as the plaintiffs aged out of the case, and the remainder were settled in 2015 with no admission of liability. No one was charged. No agency was penalized.

Hanford is now the largest environmental cleanup in the country. Millions of gallons of waste sit in underground tanks in the same desert grid, and some of the tanks have leaked. The schedule is written in decades.

T Plant is still standing in the 200 West Area, the same concrete box, kept in service long after almost everything around it was retired.

The iodine put into the air on the night of December 2, 1949, was gone within a couple of months. The record of it stayed shut for thirty-six years.