{"id":13956,"date":"2026-08-27T09:23:07","date_gmt":"2026-08-27T14:23:07","guid":{"rendered":"https:\/\/www.wisconsin.edu\/all-in-wisconsin\/?post_type=campus_story&p=13956"},"modified":"2026-08-27T09:23:10","modified_gmt":"2026-08-27T14:23:10","slug":"predicting-where-wildfire-smoke-will-go-next","status":"publish","type":"campus_story","link":"https:\/\/www.wisconsin.edu\/all-in-wisconsin\/story\/predicting-where-wildfire-smoke-will-go-next\/","title":{"rendered":"Predicting where wildfire smoke will go next"},"content":{"rendered":"\n
\"Photo:<\/a>
\u201cWell Red,\u201d a sculpture by artist Douwe Blumberg of UW-Madison mascot Bucky Badger, looks onto a wildfire smoke muted summer sunrise as it paints Lake Mendota in pastels near Alumni Park at the University of Wisconsin\u2013Madison on July 20, 2026. Photo: Taylor Wolfram \/ UW\u2013Madison.<\/em><\/figcaption><\/figure>\n\n\n\n

A collaboration between researchers at the University of Wisconsin\u2013Madison and Argonne National Laboratory is advancing new methods to improve air quality forecasting and public health alerts, highlighting the role of statistics in tackling real-world environmental challenges.<\/p>\n\n\n\n

In recent summers, wildfire smoke has been inescapable across Wisconsin, drifting south from Canada and leading to long stretches of poor air quality. Whether that smoke clears or lingers is often influenced in part by the atmospheric boundary layer \u2014 the lowest part of the atmosphere, where airborne particles mix and move.     <\/p>\n\n\n\n

At night or during cooler weather, the boundary layer compresses toward the ground, sometimes to heights below 100 meters above the surface. Under these conditions, pollutants, including smoke, can become trapped near the surface, increasing exposure and health risks. During the day, as the ground warms, the boundary layer rises and can reach  <\/s>heights of 2,000 meters, enhancing vertical mixing and the dispersion of pollutants. Tracking these changes in near real time is critical for predicting how pollution levels shift and how they may affect public health. <\/p>\n\n\n\n

In a new study<\/a>, University of Wisconsin\u2013Madison statistics professor Chris Geoga<\/a> and atmospheric scientist Paytsar Muradyan<\/a> of Argonne National Laboratory are tackling this challenge using machine learning and high-resolution data. Using data from the Department of Energy\u2019s CROCUS<\/a> Urban Integrated Field Laboratory \u2014 a large-scale collaboration led by Argonne \u2014 the team developed a model that can estimate boundary layer height in near real time. The work is an important step toward helping public health officials better anticipate how air quality conditions, including smoke impacts, may change \u2014 and provide earlier warnings to the public.<\/p>\n\n\n\n

The challenge of tracking a fast-changing boundary layer <\/h3>\n\n\n\n

Doppler LiDAR sends pulses of light into the atmosphere and measures how they scatter off particles, providing information used to infer boundary layer structure and height. But the data is complex and difficult to interpret. Clouds, rainfall, and even insect swarms can contaminate the signal. \u201cInsect swarms can produce absolutely outrageous data,\u201d reflects Geoga. \u201cThere\u2019s a lot of information hidden in the layers of the raw data that we\u2019re not seeing,\u201d adds Muradyan.  <\/p>\n\n\n\n

This uncertainty means the diagnosed height of the boundary layer is often misidentified, and with a significant time lag. \u201cEstimating boundary layer height isn\u2019t usually done at very high frequency,\u201d explains Muradyan. \u201cOften you need to average the data to improve the signal quality, so you might only get an estimate every 30 minutes or every hour \u2014 but many atmospheric processes happen on much shorter timescales \u2014 minutes, not hours.  And that\u2019s where we start to lose important detail.\u201d These estimates can miss rapid fluctuations in the boundary layer, such as during the morning when it begins to rapidly deepen, or from more localized phenomena such as lake breezes or urban heat. <\/p>\n\n\n\n

Applying statistical methods to an atmospheric problem <\/h3>\n\n\n\n

To address these gaps, Geoga turned to statistical modeling. Traditional meteorological approaches often rely on simplified assumptions that fail to account for boundary layer height estimates across space and time. For example, a boundary layer height estimated at 2 p.m. is often treated as independent from one taken just 30 minutes later. In reality, the underlying measurements, which are essentially snapshots of vertical wind motion, are strongly linked from one moment to the next and indirectly related to the boundary layer height. <\/p>\n\n\n\n

Leveraging this dependence, Geoga developed a machine learning model that treats the boundary layer as something that evolves over time, revealing \u201cbeautiful fluctuations\u201d in the data. By taking into account relationships between successive measurements of vertical wind, the approach can track both gradual shifts and sudden changes, capturing patterns that would otherwise be smoothed away.  <\/p>\n\n\n\n

Instead of analyzing each moment in isolation, the model learns from how conditions change over time to infer boundary layer height. And because this approach doesn\u2019t treat these measurements as independent of each other, another advantage is that it reduces the time lag for identifying boundary layer changes from hours to minutes; a shift that dramatically improves how quickly scientists can detect meaningful changes. The approach, Geoga explains, \u201cis far more statistically and computationally demanding than it might seem.\u201d<\/p>\n\n\n\n

Planes, pollen, and pollution <\/h3>\n\n\n\n

More accurate tracking of the boundary layer could improve air quality forecasts and support earlier warnings during events such as wildfire smoke episodes. In practice, this means communities could receive more timely alerts about dangerous air conditions. And because its height is a critical parameter in weather forecasting, it can also help researchers and meteorologists better anticipate a host of other events influenced by the boundary layer, including aircraft turbulence and the spread of airborne particles like pollen. <\/p>\n\n\n\n

The\u00a0work illustrates how statistics and data science\u00a0can unlock new insights\u00a0in\u00a0complex, real-world systems\u00a0spanning\u00a0atmospheric science, meteorology, and public health.\u00a0By making boundary layer changes visible in near real time, the research offers a powerful new tool for understanding, and responding to, air quality challenges.\u00a0Georga\u2019s work\u00a0also reflects a broader approach\u00a0in the\u00a0newly launched\u00a0College of Computing\u00a0&\u00a0Artificial Intelligence<\/a>, where computational methods and domain\u00a0expertise\u00a0come together to\u00a0address challenges\u00a0in\u00a0environmental monitoring, forecasting, and public health.\u00a0<\/p>\n\n\n\n

<\/p>\n\n\n\n

\n\n\n\n

Written by Emma Frankham, UW-Madison<\/p>\n\n\n\n

Link to original story: https:\/\/cai.wisc.edu\/2026\/07\/22\/predicting-where-wildfire-smoke-will-go-next\/<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"

A collaboration between researchers at the University of Wisconsin\u2013Madison and Argonne National Laboratory is advancing new methods to improve air quality forecasting and public health alerts, highlighting the role of statistics in tackling real-world environmental challenges. In recent summers, wildfire smoke has been inescapable across Wisconsin, drifting south from Canada and leading to long stretches […]<\/p>\n","protected":false},"author":15,"featured_media":13957,"comment_status":"closed","ping_status":"closed","template":"","institution":[103],"story_category":[147,150,146],"class_list":["post-13956","campus_story","type-campus_story","status-publish","has-post-thumbnail","hentry","institution-uw-madison","story_category-community","story_category-featured","story_category-research-innovation"],"_links":{"self":[{"href":"https:\/\/www.wisconsin.edu\/all-in-wisconsin\/wp-json\/wp\/v2\/campus_story\/13956","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.wisconsin.edu\/all-in-wisconsin\/wp-json\/wp\/v2\/campus_story"}],"about":[{"href":"https:\/\/www.wisconsin.edu\/all-in-wisconsin\/wp-json\/wp\/v2\/types\/campus_story"}],"author":[{"embeddable":true,"href":"https:\/\/www.wisconsin.edu\/all-in-wisconsin\/wp-json\/wp\/v2\/users\/15"}],"replies":[{"embeddable":true,"href":"https:\/\/www.wisconsin.edu\/all-in-wisconsin\/wp-json\/wp\/v2\/comments?post=13956"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.wisconsin.edu\/all-in-wisconsin\/wp-json\/wp\/v2\/media\/13957"}],"wp:attachment":[{"href":"https:\/\/www.wisconsin.edu\/all-in-wisconsin\/wp-json\/wp\/v2\/media?parent=13956"}],"wp:term":[{"taxonomy":"institution","embeddable":true,"href":"https:\/\/www.wisconsin.edu\/all-in-wisconsin\/wp-json\/wp\/v2\/institution?post=13956"},{"taxonomy":"story_category","embeddable":true,"href":"https:\/\/www.wisconsin.edu\/all-in-wisconsin\/wp-json\/wp\/v2\/story_category?post=13956"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}