When Hurricane Maria walloped Puerto Rico in September, it ripped off roofs, flooded neighborhoods and all but destroyed the island’s power grid, leaving a humanitarian catastrophe that Puerto Ricans are still recovering from months later.
But Maria took its toll on nature as well. Its winds of up to 155 miles an hour wrecked thousands of acres of trees, including much of El Yunque National Forest, 28,000 acres of lush tropical rain forest east of the capital, San Juan.
To a group of researchers hiking down a steep, slick mountain trail in El Yunque recently, the destruction was readily apparent. Led by María Uriarte, an ecologist at Columbia University, they were here to study the damage and better understand how an expected increase in extreme weather may undermine the ability of forests to aid the climate.
Before Maria, the mountainside here would have been in shade, a canopy of sierra palm fronds and leafy branches of yagrumo trees and others blocking much of the sunlight. But the winds knocked down many of the leafy trees, sheared off the branches of others and completely stripped the more flexible palms of their fronds. The slope was now open to the sky, and the outskirts of San Juan, normally blocked from view, were visible far in the distance.
And because climate change is expected to increase the frequency of extreme weather events in many parts of the world, the work will also help researchers understand how forests could be changed permanently as the world continues to warm.
“All the global climate models are predicting we’re going to see more severe hurricanes,” said Dr. Uriarte, who has studied the forests of Puerto Rico for a decade and a half. “So what does that mean for the composition of the forest?”
Trees are a critical part of the carbon cycle, in which carbon moves between the atmosphere, ocean and land. They remove atmospheric carbon dioxide, incorporating the carbon into their tissues as they grow. Worldwide, forests are a net storehouse, or sink, of carbon, removing one billion to two billion tons from the atmosphere each year. That’s a substantial portion of the roughly 10 billion tons of carbon pumped into the air by fossil-fuel burning and other human activities.
When a forest is damaged, the dead vegetation eventually decomposes, returning the carbon to the atmosphere. The amount can be enormous: a study of damage after Hurricane Katrina in 2005 found that the storm killed or severely damaged 320 million trees across the Gulf Coast, containing about 100 million tons of carbon.
As forests start to recover, the mix of species is often different — in a rain forest like El Yunque, for example, species that thrive in full light tend to take over until the canopy regrows. The trees are also younger and smaller, so the recovering forest stores less carbon.
“Forests take a while to recover,” said Louis Verchot, a researcher with the International Center for Tropical Agriculture in Palmira, Colombia. “And what initially recovers is not always what was there before.”
If this cycle of damage and regrowth — what ecologists call a disturbance regime — occurs more often as extreme storms become more frequent, some forests may never recover completely. Over decades, the reduction in stored carbon would likely become permanent. More carbon from human activity would remain in the atmosphere to contribute to climate change, or would have to be removed in other ways.
“If the climate warms, do we expect an increase in disturbance regimes?” said Jeffrey Q. Chambers, a geographer at the University of California, Berkeley, who led the Katrina study. “That could work on the ability of those systems to remove CO₂ from the atmosphere.”
Determining how the biomass in damaged forests changes over time is thus crucial to understanding the global carbon balance. But the work is not straightforward, as counting damaged trees in even a modest forest is a practical impossibility. Instead, researchers rely on remote sensing: satellite images to determine the presence or absence of trees, for example, or laser-based airborne measurements of a forest canopy.
Using remote-sensing data, Dr. Chambers has come up with an estimate of the trees that were killed or severely damaged by Hurricane Maria in Puerto Rico: 23 million to 31 million.
Estimates like this are necessarily preliminary. For one thing, many trees may be alive now but will eventually die from storm damage; that was the case after Hurricane Katrina. The analyses also use software to interpret the data, so the results need to be validated by the basic scientific work of assessing the damage on the ground.
“You need ground data to make sure that what you’re seeing makes sense,” Dr. Uriarte said.
That’s why she, three of her doctoral students and others were hiking down the mountain trail. Starting at about 3,000 feet elevation, near the highest point of El Yunque, their goal was to reach several study plots at intervals below.
The plots were established about two decades ago and have been studied over the years by scientists and staff at El Verde Field Station, a research outpost in the national forest that was built during the Cold War for United States government studies on the effects of radiation on vegetation.
In recent decades, with the radioactive contamination having been cleaned up, El Verde has been used for more benign research, including studies on the impact of climate change on tropical forests. Now operated by the University of Puerto Rico, the field station is home to one of the few long-term ecological monitoring programs in the tropics.
The descent was difficult, as the steep trail had been booby-trapped by downed trees and branches and covered with a thick layer of graying palm fronds.
“This is some extreme hiking,” said Andrew Quebbeman, one of the students, who was hauling a high-precision GPS device.
The idea on this day was to use the GPS instrument, which is accurate to within four inches horizontally and vertically, to match an area on the ground to the same area on a Landsat satellite image of the forest.
By the standards of digital imaging, a Landsat image is extremely low resolution. It cannot make out individual trees. Instead, software interprets the color in each small square of the image — each covering an area on the ground of about 100 feet by 100 feet — as intact or damaged forest, or bare ground.
By precisely locating one of these squares in the forest, the researchers could assess the actual condition of the trees within it and relate that to the color in the image. That will help improve the software and thus the estimates of damage across the island.
It helps that these plots, like others monitored by the field station, have been studied in the past. Each has been marked by corner stakes, and every tree with a trunk larger than about 4 inches in diameter has been identified and tagged, and its condition before the storm is known.
Once Mr. Quebbeman located the center of the square with the GPS instrument, the researchers divided the area into quarters and walked each one, stopping at each tree to assess its condition — noting whether each was alive or dead, standing or fallen, and, if a tree was damaged and standing, estimating how much of its canopy remained.
Closer to El Verde, a group of eight volunteers — recent college graduates with an interest in natural science (and in living in Puerto Rico for six months) — were doing similar work at the field station’s main study area, 40 acres of forest divided into a grid of 400 plots, each about 65 feet by 65 feet.
If anything, the damage was even worse there. Most of the palms were still standing, and in most cases two or three small fronds were already pushing out from their tops. But Maria had simply bowled over some of the other trees, leaving a jumbled mess of uprooted trunks that made it difficult to distinguish one from another. And there were so many dead palm fronds littering the ground that at times even finding the stakes marking each plot was a struggle.
While many areas were almost completely destroyed, others nearby seemed barely touched, depending on elevation, orientation and other factors, Dr. Uriarte and her team found.
They will eventually use the collected data, with the remote-sensing images and measurements, to come up with a detailed estimate of the loss from the storm.
The work is just getting started. The plan, Dr. Uriarte said, is to monitor the plots closely for several years and then less often, depending on funding.
While Dr. Uriarte doesn’t know what her final conclusions will be, she has some ideas.
“One of the things that we are seeing in forests that get more severe hurricanes is that you end up with a shorter, smaller forest, which means less carbon overall,” she said.
Whether that will happen here in Puerto Rico is one of the big questions.
“Are we going to lose some of the species? Are other species going to become more common?” she said. “We’re trying to understand this variation, and part of that is recognizing how they respond to damage.”