The forest, thick with thin ponderosa pines, is stifling. A tangle of deadwood clutters the ground, making it difficult to hike on anything but an established trail or a dirt road. The canopy overhead has mercifully broken up the glare of the noonday sun into a mosaic of light and shadow. But the chiaroscuro effect only adds to the sense of disquiet-of confusion-that pervades these woods. Bill Armstrong of the Santa Fe National Forest looks up quickly as the warm wind that has been blowing steadily through the treetops for the past fifteen minutes suddenly gusts. He doesn't say anything. He doesn't have to. It's mid-May in the American Southwest. It's fire weather. What makes things particularly spooky this day is where we are-in northern New Mexico's Jemez Mountains, where two major wildfires have broken out over the past three years. In a forest so tinder-dry that lighting a match is unthinkable. And, above all else, in a spot just a mile or two from about the last place in the world anyone in their right mind would want to see a forest fire: Los Alamos National Laboratory, birthplace of the atomic bomb and home to a witch's brew of radioactive materials and toxic chemicals.

We find a campfire ring in a small clearing. The charcoal isn't warm, so it wasn't used the previous night. But it looks fresh and is probably the remains of a bonfire from the previous weekend. "Usually you find them right in the woods," Armstrong says. "The teenagers come out here to party." There are no empty beer bottles or cigarette butts to be seen, but it's easy enough to imagine a cluster of teenagers from the nearby town of Los Alamos, which borders the lab, standing around catching a buzz and smoking Camels.

We climb back into Armstrong's U.S. Forest Service pickup and drive to a relatively spacious patch of forest. Armstrong, who's been sounding the alarm bell about the high fire hazard around the lab longer than almost anyone, walks briskly here. This is one of the few places on the lab's heavily wooded western border where the danger has been reduced through a combination of cutting trees and burning vegetation. It's more open, more like the parklike stands of ponderosa pine that existed 100 years ago, before the era of fire suppression. Still, the area is at best a few acres in size and is dwarfed by the surrounding wall of trees. "Burning breaks up the canopy," Armstrong says, squinting in the bright sunlight. "I'm not sure it can stop a major fire. But it doesn't hurt. It gives us a chance."

The extreme fire hazard in the forest around Los Alamos lab exists in forests throughout the interior West and the story by now is familiar: The suppression of wildfires for much of the century has led to unnaturally dense forests that pose an unprecedented threat of catastrophic infernos. Where previously fire burned through western forests every few years-twenty to twenty-five years at higher, moister elevations-now decades go by between fires. Virtually every sapling becomes a tree because fire is not allowed to thin things out. The trees compete with one another for nutrients, water and light; and because sustenance is limited, the result is an abundance of smallish, stunted trees. In ponderosa forests, which pose the greatest wildfire threat because of their inherent dryness, particularly dense stands are known as "doghair thickets." In some areas of the Jemez Mountains, there are as many as 2,000 ponderosa pines per acre, compared to 25 to 80 trees per acre when fire was a frequent visitor to the landscape.

It's not just that there are too many standing trees. All the trees and branches that used to fall down and that fire used to convert to ash to enrich the soil now just lie there year after year, like kindling waiting for a match. They're known as "ladder fuels" because they enable fire burning on the ground to climb into the crowns of trees and go airborne, igniting treetop after treetop. Theenormous speed, heat and unpredictability of crown fires is what keep firefighters awake at night. Such unnaturally crowded forests are prone to disease, which makes them even more susceptible to the ravages of fire. Parasites can have a field day since it's so easy to travel between trees. Dwarf mistletoe, for example-a yellow-greenish growth- is epidemic in the Jemez Mountains. It's made the fire hazard worse because it sucks the moisture out of trees, making them drier than they otherwise would be. It also deforms trees by forcing branches to twist upon themselves, which in turn concentrates explosive resins. "Mistletoe burns like hell," Armstrong says. The extraordinary buildup of fuel has occurred across entire landscapes in the West, which leads foresters like Armstrong to predict that the fires of the not too distant future could be of enormous size and intensity. "There's a potential for 100,000-acre crown fires,'' says Craig Allen, a leading Southwestern fire expert who's based at the U.S. Geological Survey's Jemez Mountains Field Station in Bandelier National Monument, just south of Los Alamos lab.

It isn't just the density of the forest that makes Los Alamos lab vulnerable to a forest fire. There are other ominous factors: - Intense wildfires are increasing in frequency. Three major ones have broken out near the lab in the past twenty-one years, two of them in the last three years. - The Jemez range receives some of the most violent lightning storms in the West. From 1985 to 1994, a staggering 160,000 ground strikes were recorded-an average of 16,000 a year. In 1986 alone, 23,000 strikes were detected. - Outbreaks of fire are frequent. From 1975 to 1996, 372 fires erupted in the 40,000 acres of the Santa Fe National Forest and Bandelier National Monument adjacent to the laboratory. - Droughts are commonplace. From late April to the Fourth of July, northern New Mexico is generally dry, windy and-in June at least-hot. In addition, periods of prolonged drought, lasting several months to several years, have occurred in the past and will again. The 1950s were particularly dry-so dry that normally drought-tolerant ponderosa pines died during that era. - The last twenty years have been unusually wet. In the Southwest, wet doesn't really mean wet. It means less dry. Nonetheless, higher than normal precipitation over the past two decades has produced an explosion of trees in the Jemez range, greatly exacerbating the fire danger. Armstrong, who wrote an influential 1998 report analyzing the fire situation near the laboratory, took some of this data and fed it into a computer model. The model spit out a chilling forecast: There's a 30 percent chance that a fire 5,000 acres or larger will break out in the heavily wooded area west of the lab in the next five years.

The danger to Los Alamos goes beyond the simple threat of fire damage. Severe flooding is certain to follow in a forest fire's aftermath. The reason involves the lab's peculiarly complicated topography. Flanked on the west by thickly forested mountains, the lab sprawls across forty-three square miles of the Pajarito Plateau, a shelf above the Rio Grande Valley. The plateau is cut with numerous fingerlike canyons that run east and drain no fewer than seven watersheds. These canyons have served as dumping grounds for the lab over the years, and many are contaminated with radioactive and chemical materials. One contains a decommissioned nuclear reactor.

Were the upper portions of these watersheds to be denuded by fire, the flash floods that would roar down these canyons once the summer thunderstorm season began would be enormous. "Even if a fire doesn't burn into Los Alamos, even if it just burns in the watersheds above [the lab and town], you're going to have accelerated runoff and erosion," Allen says. "The canyons would get scoured out for one to three years."

The flooding that would occur the first summer after a fire would be the worst, Allen adds, because there would literally be no vegetation in the burned area to hold any of the water back. Making matters worse, the charred soil left behind by a fire would be water repellent.

Following the two largest fires that have struck the Jemez Mountains in the past quarter-century-the 1977 La Mesa fire and the 1996 Dome fire-peak flows 200 times greater than the normal peak flows were recorded in canyons in and downstream of the burned areas, according to Allen. After the flooding subsided, the mouth of one of the canyons was buried in charcoal-bearing sediment five to six feet deep, according to Michael Dale, a hydrologist with the New Mexico Environment Department, which monitors off-site contamination from the laboratory.

There's no question such flooding would wash significantly higher levels of lab contaminants than normal directly into the Rio Grande-which the canyons empty into-and from there downstream into Cochiti Lake, a popular fishing and swimming area. How much is uncertain.

The laboratory has modeled what would happen during flash floods from 100-year storms, but the model assumed the forest in the mountains would be intact. The lab has never looked at the much larger floods that could come down from mountain areas stripped of vegetation by a forest fire-or done much to prepare for such an event.

But preventative measures might be pointless. "There would be nothing anyone could do that would be effective [at reducing the flooding] the first year," Allen says.

The thought of a wall of fire bearing down on a nuclear weapons facility may seem the stuff of nightmares, but in fact it's already happened. Late on the morning of June 18, 1977, two days after the La Mesa fire had broken out near Bandelier, the several hundred firefighters who were battling the blaze witnessed something they would never forget. The air that day was almost devoid of moisture-the relative humidity was 3 percent-and the winds were persistently strong-from twenty to thirty miles per hour. These factors played upon a fire that had already consumed 2,000 acres to produce what John Lissoway, a fire management officer with Bandelier, calls "one of the most spectacular, high-intensity, sustained crown-fire runs in recent history."

Across a front more than a mile wide, the fire threw up columns of flame 200 to 300 feet high. Panicked fire crews fled, and an engine crew member died of a heart attack. Perhaps his distress had something to do with the fact that the fire was charging directly at the southwestern flank of Los Alamos lab.

Fiery debris as far as two miles in advance of the front began to rain down on the lab in the early afternoon, igniting multiple spot fires within laboratory technical areas. Meanwhile, the main fire invaded the lab's southern border. But the fire began to lose intensity as the wind slackened in the late afternoon and as it entered less densely wooded areas of the laboratory. The La Mesa fire would not be contained for another four days, but it had showed everyone its worst.

The extraordinary explosiveness of the La Mesa blaze led to the convening of two scientific symposiums and a slew of studies on the effect of high-intensity wildfires on the environment. But as frightening as the fire was-and as close a call as it was for the laboratory-it oddly did not spur any serious efforts to reduce the fire hazard in and around the sprawling facility. It took the Dome fire in April 1996 to do that. The fire was in many respects a replay of La Mesa. It erupted a few miles southwest of the lab, near Bandelier. It burned fiercely in heavily forested, almost impossible-to-access canyon bottoms. And it consumed about the same acreage-16,500 acres compared to La Mesa's 15,400.

It also blew up one day like La Mesa-so quickly and unexpectedly that a staggering number of firefighters, forty-eight, had to deploy fire-resistant shelters, the most in the annals of western firefighting. The fire burned directly over three of them while they huddled in their shelters. Miraculously, no one died.

The fire burned so demonically that day that it was literally invincible. Attacking it from the ground was futile. Attacking it from air with helicopter drops of slurry-fire-retardant chemicals mixed with water-was ineffective. Dense smoke hid the blaze and made it impossible for pilots to see. Even in those instances when they could see well enough to make a drop, the intense heat evaporated the slurry before it reached the ground. As one firefighter who was there that day recalls, "All you could do was watch-from a safe distance."

The fire never burned onto the lab, but probably would have had it occurred in June, the peak of the fire season. "The only reason the lab didn't burn down during Dome was that the wind was blowing from the west,'' says Pat Valerio, a fire management official with the lab. "If it had been blowing from the southwest [the norm in June], the lab and the townsite would have received major damage."

Following Dome, fire officials from the Santa Fe National Forest, the National Park Service (which operates Bandelier), the Energy Department, the laboratory and Los Alamos County started collaborating on ways to protect both the laboratory and the town of Los Alamos, home to 11,830 people.

The effort, by all accounts, is daunting. "After 50 years of basically doing nothing (the lab was founded in the 1940s), we've got a lot of work to do," says Valerio.

The laboratory has cut a firebreak by thinning a 550-foot-wide area on either side of State Road 501, which runs along the lab's heavily forested western border. It is also cutting trees near buildings that are in wooded areas or that contain particularly dangerous combustible materials. Trees are also being cleared near outdoor detonation sites-which harbor significant soil contamination-and near a nuclear waste dump and a weapons facility identified in a recent Energy Department study as possible sources of radioactive releases during a wildfire.

According to Valerio, the lab's thinning efforts will continue for the foreseeable future. The lab received $300,000 to fund the work in the 1999 fiscal year, he says, and hopes to get about twice that much in the next fiscal year.

The National Park Service has been steadily burning areas in Bandelier for the past several years, which lowers the risk to Los Alamos because any fire in or near Bandelier is likely to be pushed north onto the lab by the prevailing winds-as the La Mesa fire was twenty-two years ago. Ironically, because the La Mesa fire burned so hot, that area has been slow to recover and serves as a significant buffer zone between the lab and the forest. The problem area, everyone agrees, is west of the laboratory, on Santa Fe National Forest land, where a virtually unbroken, thickly wooded ridgeline forms a scenic but ominous backdrop to the lab and the community of Los Alamos. The Forest Service has done some thinning and burning, but the "vast majority is untreated,'' according to Rich Wands, a fire management officer with the Santa Fe forest.

In his 1998 report, Armstrong lays out his prescription: thin and burn 5,000 acres in a 19,000-acre area west of the lab. The project, estimated to cost between $1 million and $1.5 million, is set to be implemented next year and could be completed in six to eight years, according to Armstrong and Wands.

Whether that goal will be met remains to be seen, but Armstrong said the chances would be improved were the Energy Department to chip in some money. The department currently provides funds for fire risk reduction efforts carried out by the Forest Service at another government nuclear weapons facility, the Savannah River Plant in South Carolina.

A factor that might delay the effort, or make it more time consuming, is the large size of the Santa Fe National Forest. Covering about 1.5 million acres, there are lots of areas on the forest in need of thinning and burning. Next year, Wands says, Santa Fe officials hope to conduct thinning and burning on 27,000 acres-only 3,000 of which are close to the lab's western border.

When asked which part of the Santa Fe forest was the top priority in terms of fire risk reduction, Wands cited the Jemez District, located well to the southwest of the lab in the southern portion of the Jemez range. "That's probably the driest and it's heavily used for recreation by people in Albuquerque," he explains.

That prioritizing doesn't sit well with Valerio, who used to work for the Santa Fe forest. "What could be more important than reducing the fire risk near a nuclear weapons facility?" he asks.

Another difficulty is that considerable preparation must take place before a thinning or burning project can be undertaken. In the case of Armstrong's proposal, surveys must be done to ensure that archaeological sites-of which there are a staggering number in the Jemez Mountains-will not be damaged. Steps must also be taken to prevent harm to rare wildlife, such as the Jemez Mountains salamander, which lives nowhere else. All of this costs money.

"Planning for these burns is getting almost as expensive as pulling them off," Wands says.

In fact, the agency spends more time and money planning the projects than doing the on-the-ground work of cutting trees and burning. Consider next year's budget request: $2.1 million is being sought to pay for planning efforts for 75,000 acres of the forest, and $1 million has been requested to cut and burn 27,000 acres.

Armstrong's proposal is likely to face a challenge from environmentalists. Citing a recent Forest Service study that found that the best way to protect structures from wildfires is to clear immediately surrounding vegetation-not to thin forest on adjacent public land-Bryan Bird of the Santa Fe-based Forest Conservation Council said he doubts Armstrong's project would be effective.

However, the author of the study, Jack Cohen, a scientist with the agency's Fire Sciences Laboratory in Missoula, Montana, says his analysis should not be construed to mean that thinning and burning near Los Alamos is unneeded-particularly since the lab harbors so many outdoor contaminated areas in need of protection from fire.

"There may be all kinds of reasons to do vegetation management in that particular area. The last thing you want is a whole bunch of movement (of contaminants) off property because of a severe fire event," Cohen says.

If a major fire did burst onto Los Alamos, what might happen? It's not at all clear. However, a recent Energy Department study provides clues. The study postulates the following scenario. An out-of-control forest fire much like the La Mesa blaze burns onto the lab, consumes everything combustible in its path and eventually torches 27,000 acres-8,000 of which are within the boundaries of the nuclear facility. The fire is out of control, and it doesn't slow down until it reaches lower-elevation areas of the lab that do not contain dense forests.

Protective action taken by laboratory personnel-such as relocating hazardous materials and barricading windows-significantly reduces the risk of major releases of radioactive substances and toxic chemicals, but doesn't remove it.

In the end, tritium, a radioactive form of hydrogen, escapes from a weapons facility on the lab's wooded western fringe. Meanwhile, drums containing plutonium-contaminated waste at a nuclear dump burst from the heat of the blaze and their contents catch fire.

Vegetation at the laboratory, which is known to contain elevated levels of radioactivity, also contributes radiation to the smoke plume. But burning plutonium and tritium deliver the lion's share of the estimated total radiation dose of 675 "person-rems" delivered to the public within a fifty-mile radius of the fire.

That's 135 times the annual allowable exposure limit for nuclear workers, which may sound like a lot. But no single member of the public comes close to receiving such a dose, or even the 5 rem annual worker limit, according to the study, because most of the radiation is dispersed into the atmosphere by the smoke plume. In fact, assuming that the population in the fifty-mile radius area is 100,000-which is roughly correct-the average exposure for each person, according to a calculation by one of the authors, Doug Minnema, is only 6 millirems-a fraction of the 325 millirems of radiation that people living in Los Alamos are exposed to from natural sources such as sunlight and cosmic rays.

Minnema, who works for the Energy Department and is an expert on the health effects of radiation, acknowledges that in reality things would be more complex. Some people, depending on their proximity to tainted smoke from the fire, would receive larger doses of radioactivity than others. Someone standing directly downwind of the nuclear dump when it burned, for example, could absorb significant amounts of radiation if they remained there for the duration of the blaze. Finally, a portion of the laboratory's vast stores of combustible chemicals could ignite in a forest fire and release their deadly contents into the air.

Firefighters would be in more danger than anyone else. To reduce exposures, Valerio says specially trained teams equipped with respirators would be sent in to fight fires in areas known to be especially contaminated.

There's something laboratory and Energy Department officials don't like to talk about when it comes to the subject of an out-of-control wildfire on the lab: the chaos factor.

Sure, lab workers may have time to relocate hazardous materials to fireproof areas. But if the fire surprises them, they may flee in panic before they get the job done. Sure, firefighters wearing respirators may be protected from toxic fumes in the smoke. But what if so many firefighters are needed that there aren't enough respirators to go around? Armstrong spoke of chaos that May day as he drove into the town of Los Alamos in his pickup. He pointed out that the two-lane road he traveled on is the only major route off the plateau. He pointed to a nearby canyon and said, "You'd have smoke coming out of there." And then he gestured to the thickly forested mountains above the town and the lab and said, "You'd have spot fires raining down from a fire over there.

"Everybody's trying to get out down this one little road," he continued. "You've got fire below you and fire above you and some guy panics and gets his vehicle side-crossed and the way out is blocked. Then what do you got?" he asked, his pale blue eyes suddenly flashing.

This, then, is the reason for Armstrong's urgency. He sees it. He sees the mountain on fire. He sees the canyons belching up tainted smoke onto the plateau where most of the people are (so much for smoke wafting harmlessly upward). He sees biblical-sized floods flushing God-knows-what into the Rio Grande.

He drove on, through the town, through the lab and into the thick forest.

-- Keith Easthouse is associate editor of Forest Magazine.