Sugar Companies Said Our Investigation Is Flawed and Biased. Let’s Dive Into Why That’s Not the Case.

Series: Black Snow

Big Sugar’s Burning Problem

On July 8, The Palm Beach Post and ProPublica published an investigation into harvest practices in the nation’s largest cane sugar-producing region.

For decades, residents in Florida’s heartland have raised concerns about exposure to pollution from burning cane fields, a technique used to rid the plant of its leaves. In June 2020, The Post partnered with ProPublica’s Local Reporting Network to investigate the topic.

Over the course of the past year, the news organizations have been working together to understand the impact of cane burning on air quality and local health. We interviewed dozens of people, collected hundreds of public records and read thousands of pages of documents on the nation’s air-monitoring infrastructure.

Through installing outdoor air sensors at the homes of residents during cane burning season, we developed a better understanding of pollution patterns in the neighborhoods at the center of vast swaths of cane fields. Our analysis found repeated spikes in pollution, generally lasting less than an hour, on days when the state authorized burns and projected that smoke would blow toward our sensors. Experts said these findings strongly suggest a link between air pollution and cane burning.

Prior to publication, we sent our findings, project methodology and questions to U.S. Sugar and Florida Crystals, the largest sugar producers in the region. The companies responded by challenging our reporting and methodology. The Post and ProPublica carefully reviewed those criticisms and engaged with the companies over several weeks, clarifying our evidence and providing more information upon request, including our raw data, documents obtained through public records requests and scientific studies on the health effects of particulate matter. Below is a breakdown of the criticisms and the reporting team’s responses.

Both companies criticized our finding that the state air monitor in Belle Glade was malfunctioning as far back as eight years ago. Spokespeople for the companies called the monitor reliable.

In 2013, the Florida Department of Environmental Protection flagged the Belle Glade monitor for discrepancies, saying it wasn’t fit to determine compliance with the National Ambient Air Quality Standards, the pollution thresholds set by the Clean Air Act. The monitors used to enforce these thresholds and crack down on polluters must meet strict quality-assurance criteria, according to the Environmental Protection Agency.

Instead of replacing or repairing the monitor, the state environmental department kept running the monitor for the Air Quality Index, a tool used to broadly tell the public whether the air is good, unhealthy or hazardous. The EPA allows this, but the Air Quality Index is not an enforcement tool.

After The Post and ProPublica questioned the Florida Department of Environmental Protection about the monitor, a spokeswoman said it would be replaced “in the future” with one that meets the EPA’s stricter standards. In a June 29 air-monitoring plan, the department said the monitor was expected to be replaced this month.

A U.S. Sugar spokesperson said that “data collected and authorized by local health departments and the FDEP continue to show that the air quality in the Glades is safe, healthy and consistently better than the standards set by the EPA.”

The Belle Glade monitor routinely showed good air quality on the EPA’s Air Quality Index. Even so, the country’s air monitoring system is ill-equipped to detect pollution in the Glades for these reasons:

1. The 24-hour air-quality averages can obscure short spikes in pollution: The EPA enforces annual and 24-hour standards for a type of pollution called particulate matter. That means state agencies average out the data collected by monitors over 24 hours and a year. Shorter spikes can be obscured by the averages. Researchers have found that even short-term exposure to these pollutants, especially on a repeated basis, can be harmful.
2. Monitoring in rural areas is insufficient: When the EPA first set up the country’s air-monitoring network, agency officials decided that the more people living in an area, the more air monitors were required to measure pollution levels. This left many rural areas with no monitors. Palm Beach County has three regulatory monitors, all of which are in suburban or urban areas instead. The one monitor in the Glades could not be used to enforce air pollution standards.
3. Monitors can miss pollution from highly localized sources: Each air-quality monitor is housed in a single location and is unable to detect pollution beyond a certain range. In the Glades, sugar growers burn individual fields across a region spanning 400,000 acres. That means a single monitor in Belle Glade might not pick up on pollution across the region.

Our analysis spanned four months, whereas the EPA requires at least three years worth of data to determine Clean Air Act compliance. That’s because our goal was not to replicate regulatory monitoring. Instead, we aimed to better understand the air quality in the Glades during the burning season. The data helped us identify gaps in air monitoring that may leave residents exposed to harmful pollution.

Experts who helped us form our plan and later reviewed the data said the repeated spikes in pollution during burn season, along with residents’ health concerns, warrant more regulatory scrutiny of the area’s air quality.

After being interviewed by the reporters, Sheryl Magzamen, a Colorado State University professor, submitted a grant proposal to NASA aimed at expanding the network of air sensors in the Glades and examining health trends. “I certainly hope you highlight that the motivation for submitting the grant was your reporting on this issue,” Magzamen told The Post and ProPublica.

Both companies criticized our decision to use low-cost PurpleAir sensors instead of data from the state monitor in Belle Glade. A U.S. Sugar spokesperson said, “It is incredible that the Palm Beach Post and ProPublica would base its air quality findings on low-cost sensors they paid to have placed in backyards operated by non-professionals.”

PurpleAir sensors are an accessible solution for everyday people interested in measuring their air quality. Their low cost, portability and ease of operation are attractive features to state and federal governments too. In fact, the EPA uses PurpleAir sensors as part of a loan program across the Midwest, in tribal communities and in California to get a more detailed picture of air quality.

Though they are not used for enforcement purposes, the EPA also recently began integrating real-time data from PurpleAirs in its AirNow Fire and Smoke Map, which tracks fires and air quality nationwide. PurpleAir initiatives have been successful at providing the public with real-time air-quality data to help protect its health, an EPA spokesperson told The Post and ProPublica.

That said, researchers need to make some adjustments when using the sensors. The equipment can be affected by high levels of humidity, PurpleAir creator and founder Adrian Dybwad said in a 2020 panel discussion on air-quality research. He added, however, that significant effects are rare, usually seen when humidity levels surpass 80%. Even so, the EPA developed a correction formula for PurpleAir monitor readings that accounts for relative humidity, among other factors. In consultation with air-quality experts who have used PurpleAir sensors and EPA researchers, The Post and ProPublica applied the EPA formula to the data. Our methodology details the quality-assurance steps we took to vet our sensor data.

It’s true that The Post and ProPublica paid for the PurpleAir sensors, but all of our sensor hosts were volunteers and none were paid or otherwise compensated.

Florida Crystals criticized our use of 10-minute averages to measure pollution instead of the 24-hour averages used by state and federal regulators.

Experts suggested we do this because of the short-term nature of cane burns, which typically last less than an hour.

Although the EPA doesn’t require the reporting of shorter time periods like 10-minute averages for particulate matter, it has shifted to shorter averages for other pollutants. For instance, the agency used to measure sulfur dioxide, a byproduct of burning oil and coal, using 24-hour averages. After multiple studies tied short-term exposure to the gas to breathing problems, and after a successful lawsuit against the EPA over short-term measurements, the agency adjusted its standards in 2010. It now measures sulfur dioxide using one-hour and five-minute averages.

Both companies said spikes in pollution could not be tied to sugar cane because they occurred outside the time frame in which burns are performed.

One of the challenges with tracking cane burning is there’s no way to know exactly when each burn took place. Permits issued by the state don’t dictate the exact time growers must burn their cane. Instead, the state issues permits using time brackets. For example, one permit might authorize a grower to burn cane anytime between noon and 8 p.m.

There were some spikes in pollution outside the time frame that burns are authorized. But the majority of the spikes fell within the time frame when cane burning permits are authorized (between 9 a.m. and two hours after sunset, or about 8 p.m., for certified burners).

To strengthen our analysis, we used modeled smoke plume data from the state. Pollution spikes were more pronounced and concentrated on days when smoke was projected to move toward our sensors versus away from them. For this reason, experts told us the analysis strongly suggests the spikes were coming from nearby cane burns.

U.S. Sugar questioned how we found our sensor hosts, pressing us on whether they were activists with the Sierra Club.

They are not. None of the residents who hosted the monitors have worked with environmental advocacy organizations in the Glades. And none have taken part in the pending federal lawsuit against sugar companies over cane burning.

In fact, most of the residents we spoke with had no ties to environmental groups or the lawsuit. When the news organizations started reporting a year ago, we set out to speak with as many Glades residents as possible. We wanted to hear from people who had not taken an active role in the public battle over cane burning. “We just want to know: What’s in the air? What are we breathing?” said Jose Fonseca, a parks worker who grew up in the Glades and hosted a PurpleAir sensor for this project.

We used a number of tools to reach hundreds of residents, including sending letters to public school teachers and custodians across the Glades; knocking on doors in neighborhoods in Belle Glade, Pahokee and South Bay; attending a virtual church service; canvassing food distribution sites; delivering flyers to businesses and nonprofits; and calling local doctors and nurses. We also set up an automated text bot to interview residents in real time when we witnessed pollution spikes on our PurpleAir sensors. Ultimately, we connected with dozens of residents who spoke about the air quality in the agricultural community.

A U.S. Sugar spokesperson said: “The Palm Beach Post and ProPublica are selectively presenting their preferred interpretation of studies to support a biased conclusion for this story.” Florida Crystals also alleged that our newsrooms cherry-picked studies that align with our findings.

While there are disagreements among policymakers about how polluters should be regulated, there is little dispute among scientists about the harms associated with particulate matter.

The companies specifically criticized The Post and ProPublica’s references to a 2020 EPA rule proposal that examined research into long- (annual) and short-term (daily or less) exposure to particulate matter. The proposal reads that “several key epidemiologic studies report positive and statistically significant PM2.5 health effect associations based largely, or entirely, on air quality likely to be allowed by the current primary PM2.5 standards.”

A seven-member EPA committee, selected by the Trump administration, ultimately concluded that current standards “sufficiently” protect health.

In mid-June, the EPA under President Joe Biden announced it would reconsider the previous decision to maintain the standards. The agency cited the prior proposal’s “strong body of scientific evidence” showing that long- and short-term exposures to PM2.5 can lead to heart attacks, asthma attacks and premature death.