Emissions from biomass burning in the Yucatan

Two small research planes are circling above the Yucatan Peninsula in March 2006, with their noses pointed into columns of smoke rising from burning fields and forests. One is an NCAR C-130, and the other is a University of Montana Twin Otter. They fly directly into the plumes, sampling air that no research aircraft had ever sampled in this part of the world. This was the first time anyone had made detailed measurements of biomass burning emissions in the Northern Hemisphere tropics. What they found rewrote the picture of how fire smoke behaves — not just at the moment of burning, but in the hours that follow. Biomass burning is the second-largest trace gas source on Earth and the largest source of primary fine carbonaceous particles. Most of that burning happens in the tropics. However, before 2006, nearly all detailed airborne fire studies had been done in the Southern Hemisphere during its dry season. The Northern Hemisphere tropics — where burning peaks from February through May — were poorly characterized. The MILAGRO campaign, short for Megacity Initiative: Local and Global Research Observations, was designed to study how Mexico City's emissions and regional fires together shape atmospheric chemistry. Yokelson and colleagues used it to fill the Northern Hemisphere tropics gap, sampling 20 fires across the Yucatan and measuring up to 49 trace gas and particle species per fire. The scale of burning they were studying was not trivial: a regional model estimates roughly 44 teragrams of biomass burned in the Yucatan during the spring of 2006 alone. Trajectory analyses showed that about half of those emissions disperse northwesterly, with the capacity to mix with Mexico City outflow and reach much of Mexico and the United States. The study focused on two dominant fire types: deforestation fires and crop residue fires. To characterize them, the researchers used a metric called modified combustion efficiency, or MCE — essentially the fraction of burned carbon that comes out as carbon dioxide rather than carbon monoxide. High MCE means more flaming, while lower MCE means more smoldering. Yucatan deforestation fires averaged an MCE of 0.927, consistent with previous regional measurements. The two fire types turned out to be broadly similar in their emissions, but crop residue fires emitted notably more organic acids and ammonia. Acetic acid emission factors were 4.8 grams per kilogram for crop residue fires versus 2.9 for deforestation fires. Ammonia was 1.38 grams per kilogram for crop residue versus 0.775 for deforestation. Formic acid showed up in crop residue smoke but was below detection in deforestation smoke. Beyond those differences, the chemical profiles largely overlapped. The instrument setup on the Twin Otter centered on an airborne Fourier transform infrared spectrometer, which measured gases including carbon monoxide, carbon dioxide, methane, nitrogen oxides, ammonia, hydrogen cyanide, formaldehyde, methanol, and both organic acids, in a flow-through cell sampling directly from the plume. Whole-air canisters captured a wider range of non-methane hydrocarbons. The C-130 brought additional instruments for aerosol chemistry and continuous trace gas measurements. Together, the two platforms increased the number of species quantified in these plumes from four to 51. What they found in fresh smoke was chemically rich in ways that surprised even experienced atmospheric chemists. The Yucatan fires emitted unusually high amounts of sulfur dioxide: the molar sulfur dioxide to carbon monoxide emission ratio was 0.0173 — roughly seven times the tropical forest average. Particle chloride was also elevated, with an emission factor of about 0.51 grams per kilogram. Yokelson and colleagues attribute both to the strong marine influence on the Yucatan Peninsula, where soils, fertilizers, or marine deposition appear to load vegetation with sulfur and chloride before it ever burns. Hydrogen cyanide and acetonitrile, the two classic tracers for biomass burning, were present at ratios consistent with previous fieldwork. But the finding that stood out most was nitrous acid, or HONO. In the freshest samples, HONO accounted for roughly 10 percent of initial NOy — the total reactive nitrogen pool. That is an extraordinary proportion, and it has direct consequences for what happens next. Because once smoke leaves the fire, it is not done reacting. It is just getting started. Yokelson and colleagues tracked what happened to one C-130 plume — Fire 3 — as it aged. The ratio of excess ozone to excess carbon monoxide, written as delta O3 over delta CO, climbed from near zero to about 15 percent in less than one hour. Formaldehyde increased. Hydrogen peroxide quadrupled over roughly one point three hours. These are not slow, steady changes — they are rapid photochemical transformations driven by an extraordinarily reactive chemical environment. The engine was hydroxyl radical, OH — the primary oxidant of the atmosphere. Yokelson and colleagues made the first in-situ OH measurement ever recorded in a biomass burning plume. Over a 29-second average in smoke estimated to be 22 to 43 minutes old, OH was one point fourteen times ten to the seventh molecules per cubic centimeter. That is about 520 times higher than nearby background hydroxyl radical levels. The source of that radical burst traces directly back to the high initial HONO: HONO photolyzes rapidly in sunlight to produce hydroxyl radical and nitrogen oxide, and when 10 percent of your reactive nitrogen starts as HONO, you generate a surge of radicals that accelerates every subsequent reaction. Within about 20 minutes, the HONO signal had fallen to roughly one-tenth of its initial value — it had already been consumed to fuel the chemistry downstream. Nitrogen oxides themselves were rapidly lost. Total nitrogen oxides loss reached 54 percent within about one point two to one point four hours. About 31 percent of that went into peroxyacetyl nitrates, or PANs — organic nitrate compounds that can transport reactive nitrogen far from the source before releasing it. An additional 30 percent of the nitrogen oxides loss could be attributed to gas-phase hydroxyl radical chemistry, based on the measured hydroxyl radical concentrations and the reaction rate constant for hydroxyl radical plus nitrogen dioxide. Particles were changing just as fast. The mass ratio of secondary particulate matter to carbon monoxide increased by a factor of 2.6 within about one point four hours. Aerosol mass spectrometer measurements showed sulfate rising by a factor of 8.6, nitrate by 5.4, and organic aerosol by 2.3. Black carbon's mass fraction fell by a factor of 2.6 as non-absorbing material was added around it, and single-scattering albedo — the fraction of light scattered rather than absorbed — rose from about 0.75 to 0.93. Transmission electron microscope images confirmed the transformation at the particle level: particles shifted from potassium chloride compositions to sulfate-dominated chemistry, and tar-ball organics appeared within 10 to 30 minutes of emission. This matters because models of fire impacts typically use emission factors — the chemical snapshot of fresh smoke — as their primary input. However, these measurements show that a plume one hour old looks fundamentally different from a plume one minute old. It has more ozone, more secondary particles, and more oxidized nitrogen compounds. The atmosphere near a major fire source is not just receiving emissions; it is processing them at a rate that current models struggle to reproduce. Attempts to model the observed ozone growth, Yokelson and colleagues note, only work if you add heterogeneous reactions or increase initial organic compound estimates by around 30 percent to account for species that remain unmeasured. The regional picture makes this more than an academic concern. The NCAR model driven by MODIS satellite fire detections estimates nearly one million hectares burned in the Yucatan in spring 2006. Comparing the resulting emissions to the Mexico City metropolitan area inventory, Yucatan fires appear to emit roughly 2 to 4 times more sulfur dioxide and primary particulate matter annually than the entire Mexico City urban area — accounting for inventory biases in both directions. And those emissions do not stay local. Forward trajectory analyses show them reaching broadly across Mexico and into the midwestern and eastern United States during the March through May season. Here is what makes this study unusual: it did not just fill in a blank on the map. It revealed how large the blank actually was. Roughly 50 percent of the non-methane organic compounds emitted by these fires remain chemically unidentified. Secondary formation of sulfate, organic aerosol, and nitrate was already well underway within the first hour. Hydroxyl radical concentrations exceeded anything previously measured in a fire plume. The chemistry responsible — triggered by unexpectedly high HONO, amplified by marine-enhanced sulfur and chloride, driven by radical bursts that models had not anticipated — is still not fully understood. For the second-largest trace gas source on Earth, that is a significant gap. Yokelson and colleagues argue plainly that more research is needed to understand these post-emission processes. The two planes circling above the Yucatan in March 2006 found answers. They also mapped the shape of what we still do not know. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.