The Soybean Free Air Concentration Enrichment (SoyFACE) facility is the longest running open-air carbon dioxide and ozone enrichment facility in the world. For over two decades, soybean, maize, and other crops have been exposed to the elevated carbon dioxide and ozone concentrations anticipated for late this century. The facility, located in East Central Illinois, USA, exposes crops to different atmospheric concentrations in replicated octagonal ~280 m2 Free Air Concentration Enrichment (FACE) treatment plots. Each FACE plot is paired with an untreated control (ambient) plot. The experiment provides important ground truth data for predicting future crop productivity. Fumigation data from SoyFACE were collected every four seconds throughout each growing season for over two decades. Here, we organize, quality control, and collate 20 years of data to facilitate trend analysis and crop modeling efforts. This paper provides the rationale for and a description of the SoyFACE experiments, along with a summary of the fumigation data and collation process, weather and ambient data collection procedures, and explanations of air pollution metrics and calculations.
Background & Summary
Prior to the Industrial Revolution, the concentration of carbon dioxide ([CO2]) in the atmosphere did not exceed 280 parts per million (ppm) for at least 800,000 years1. The global monthly mean for ambient [CO2] was 416 ppm in October of 2022, a ~50% increase since industrialization and a ~20% increase since 1980 [https://gml.noaa.gov/ccgg/trends/global.html]. Atmospheric [CO2] is continuing to rise at an unprecedented rate and without significant emissions reductions, the concentrations will reach 700 to 1100 ppm by 21002. In addition to CO2, the tropospheric concentration of ozone ([O3]) has also increased since the Industrial Revolution. Although O3 in the stratosphere, which is about eight kilometers above the Earth’s surface, provides a useful barrier for ultraviolet radiation, O3 in the surface tropospheric layer is a toxic pollutant. The ~40% increase in tropospheric [O3] since industrialization3 has caused harm to humans, animals, and plants4. These increases in CO2 and O3 concentrations have already directly impacted plants5,6, and future atmospheric increases will only intensify impacts on agriculture.
For over a century, scientists have tried to understand the effects of atmospheric change on agriculture6,7,8,9,10. To this end, researchers have experimented with different approaches to alter greenhouse gas concentrations around plants. Controlled environmental enclosures, including growth chambers, allow for relatively precise control of light, humidity, temperature, and atmospheric composition, enabling scientists to mechanistically test how CO2 or O3 affect plant function while all other conditions are held constant. However, growth chambers are typically small which limits the number of individual plants that can be examined, and there are often inconspicuous differences in supposedly identical chambers11. Also, in addition to enriching the air with CO2 and O3, chamber experiments can have unwanted impacts on other aspects of the environment surrounding the plant12. For example, plants in chamber studies are grown in pots which can restrict root growth, altering the plant response to elevated [CO2]13,14. Greenhouses and naturally sunlit outdoor growth chambers are an alternative option to indoor chambers and can be used to study plants in natural environments9. Because the sides of outdoor chambers typically are made of acrylic plastic, they can partially block radiation and potentially increase temperature and humidity. Outdoor chambers can also increase [CO2] at the plant canopy level if the fumigation is released from the bottom of the chamber9. Additionally, open-top chambers (OTCs) shelter vegetation from the wind and force air upwards through the canopy, which alters the natural atmospheric coupling12. Thus, although outdoor chambers use aspects of the natural environment to operate, they can also modify the surrounding environment in undesirable ways similar to indoor chambers. These environmental and atmospheric impacts make the results of chamber studies less comparable to the natural response of plants to climate change, which is a clear limitation of both indoor and outdoor growth chambers.
Free Air Concentration Enrichment (FACE) facilities were developed as a ‘real-world’ approach to understanding how plants respond to altered atmospheres in fully open-air conditions9,10,15,16. An important benefit of the FACE approach is the ability to study the interaction of multiple atmospheric variables with elevated [CO2] and [O3], and the effects of these interactions on plants in nature17. Unlike growth chambers which are restricted to only 1 to 15 m2 in size, FACE plots can range in size from 100 to 300 m2, allowing for experiments that are larger in scope and more varied than OTC experiments18. A typical FACE plot consists of a circular or octagonal array of pipes that release CO2 or air enriched with CO2 or O3 at the canopy surface for small stature vegetation, or at varying levels from the ground to the top of the canopy for larger stature vegetation10. The [CO2] or [O3] is measured at the center of the plot, along with wind speed and direction. A computer-controlled system adjusts flow rates to maintain the target gas concentration in the center of the plot. While different FACE facilities have used vertical or horizontal release pipes with or without blower systems to mix the air, the general design has been used to study the response of natural and managed ecosystems to elevated [CO2] and/or [O3] around the world19,20.
The Soybean Free Air Concentration Enrichment (SoyFACE) facility in Illinois was first developed in 2001 and has been operational for over two decades, making it the longest-running FACE facility (Fig. 1). SoyFACE contains replicated elevated [CO2] and elevated [O3] plots, along with ambient plots, within a 32 ha farm. Initially, experiments focused on understanding the responses of soybean and maize to elevated [CO2] and [O3]21,22,23,24,25,26 and later experiments investigated the interaction of CO2 with temperature27,28 and drought stress29,30,31. More recently, the facility has been used to study genetic variation in crop responses to atmospheric change32,33,34 and genetically engineered adaptation to rising [CO2] and temperature35,36. During the growing season, ambient [CO2] and [O3] and FACE [CO2] and [O3] data are collected every four seconds along with the wind speed and wind direction. These data are averaged to produce 1-minute fumigation37 (File 9) and ambient data files, which can be used to calculate fumigation efficiency statistics and hourly, daily, monthly, and seasonal fumigation and ambient data metrics. As the longest running FACE facility in the world, it is fitting that we provide broad access of our data to the scientific community. The abundance of fumigation data from experiments executed over two decades make the SoyFACE fumigation data particularly useful for modeling the impacts of rising CO2 and O3 on crop physiology and agronomy, as well as ecosystem function.
Field site & experimental design
The 32 ha SoyFACE farm is located on the south side of the University of Illinois Urbana-Champaign campus (228 meters above sea level; 40° 02′ North, 88° 14′ West). Each year, approximately half of the field is planted with soybean (Glycine max) and half with maize (Zea mays). The maize crop is fertilized with ~200 kg N ha−1 and the soybean crop is not fertilized. Lime has been periodically applied to the crops over the past 20 years, with pre- and post-emergent herbicides applied to the crops per common practice in the region. The two predominant soil types at the farm are Drummer silty clay loam and Flanagan silt loam. Within the 32 ha field, there are fixed locations for 32 octagonal plots. Soybean and maize are each grown over roughly half of the facility in a given year, and rotate positions in successive years. Each FACE plot is 20 meters in diameter with an area of ~280 m2 (Fig. 1a,b). In a given year, experiments within a specific crop have used 8–16 plots in randomized complete block designs. Most experiments have been designed with replication at the plot level of n = 4 (Tables 1, 2). In some cases, split-plot treatments of drought or temperature have also been applied27,38. The crops are typically planted between mid-May and mid-June, with fumigation starting within two weeks of the planting date and crop harvests occurring in September and October.
Outside of the six-year period between 2003 and 2008, any given treatment plot was fumigated with only elevated [CO2] or elevated [O3] treatments (Tables 1, 2). Between 2003 and 2008, some plots were simultaneously fumigated with both CO2 and O3 in a ‘combined’ treatment that allowed testing of interaction effects when studied in conjunction with the CO2-only and O3-only plots. Wind speed and [CO2] and [O3] are measured from the center of each plot (Fig. 1b) and sent to a computer control system in the field, which inputs the measurements into the Proportional Integral Differential (PID) algorithm (Eq. 1). The PID algorithm is a commonly used control process that uses key sensor inputs and setpoints to calculate the output variable. The SoyFACE PID algorithm includes a wind speed component that is not found in other PID algorithms and was developed by Lewin et al.39. In the case of the SoyFACE experiments, the sensor inputs are the wind speed and current [CO2] and [O3], the setpoints are the target [CO2] and [O3], and the output variable is the voltage that should be used for the gas valves in order to maintain the setpoints. The setpoints for [CO2] and [O3] in the SoyFACE plots for each experiment between 2001 and 2021 are given in Table 1 (2001 through 2010) and Table 2 (2011 through 2021). The process of adjusting the CO2 and O3 levels is inexact since it is impossible to predict exactly which valve setting should be used to add a specific amount of gas to each plot. However, by using the PID algorithm at frequent intervals, the fumigation system is able to use the valve settings to quickly correct gas level adjustments that are above or below the desired amount. The goal of the SoyFACE system is to maintain the gas levels as closely as possible to the setpoints by continually measuring the current levels, implementing the PID algorithm, and releasing additional gases into the plots as needed.
The variables in Eq. 1 represent the components of the PID algorithm, as defined below.
Gcv (Control/Output Variable): The voltage applied to the valve that controls release of CO2 or O3 into the plot.
Gstpt (Setpoint): The setpoint of CO2 or O3 in the plot.
Gpv (Process Variable/Sensor Input): The current level of CO2 or O3 in the plot.
ws (Wind Speed/Sensor Input): The current wind speed, measured from the center of the plot.
Kp: Proportional coefficient; indicates that the control variable should be adjusted proportionally to the error in the system, or how much the process variable differs from the setpoint: error = (Gstpt – Gpv).
Ki: Integral coefficient.
Kd: Differential coefficient.
Kw: Wind coefficient.
Int(): Integral function; measures the accumulation of the error over time.
Der(): Differential function; compensates for sudden changes in the error.
Coefficients Kp, Ki, Kd, and Kw are constant values determined through analysis of iterative test runs of the fumigation process in early SoyFACE experiments. Currently the coefficients are as follows:
Value for CO2 Experiments
Value for O3 Experiments
The fumigation system at the SoyFACE farm was initially designed to distribute CO2 to the treatment plots. The early SoyFACE model was based upon the system used at a poplar FACE plantation (POPFACE) in Tuscania, Italy. This fumigation system was designed to release pure CO2 into the atmosphere at a high velocity through a large number of small regularly spaced air jets to create a shock wave and turbulence at the gas exit point. This enhanced the mixing of CO2 into the ambient air and since the jets face outward, the CO2 is mixed with the air prior to being carried by the wind back into the plot. This provides a relatively uniform elevated [CO2] within the plot16.
In addition to the air jet configuration, determining the ideal pressure settings for each step of the CO2 fumigation process was critical. To this end, a manually controlled pressure generator with narrow perforations was used to modify the pressure of the CO2 gas flow within the pipes in eight steps, from 0.15 MPa to 0.45 MPa. The ability to change the air pressure allowed for better regulation of the CO2 flow rate as the gas was transported through an underground HDPE pipeline to horizontal pipes along the sides of the octagonal plots. Additionally, the voltage calculated by the PID algorithm regulated the pressure inside the horizontal pipes and controlled the amount of CO2 entering the plots. Following the release of CO2, natural wind currents facilitated the distribution of CO2 throughout each plot.
The CO2 fumigation system at SoyFACE retains much of the same design of the original POPFACE model. Liquid CO2 is stored in a 50-ton vertical tank at the SoyFACE facility and passed through vaporizing equipment to produce gaseous CO2, which is delivered to specific locations in the field through underground pipes. CO2 is transported to a manifold (gas delivery system; Fig. 1c) outside of the plot and tubing delivers the CO2 to the tubes surrounding the treatment plots. CO2 is released through a linear flow valve (SMC pressure controllers16) and pure CO2 is released into the wind through 350 or 500 small air jets placed 15 mm apart and drilled into 8-meter-long pipes that surround the SoyFACE plots (Fig. 1). The flow valves have settings between 0 and 10, with a setting of 0 indicating a completely closed valve and a setting of 10 indicating a completely open valve. The control computer system (formerly Z-World Inc. Model SR9000; currently Rabbit BLS4200 series) calculates the amount of gas that should be released into each plot based on information from wind sensors (R.M. Young Model 12005), CO2 analyzers (PP Systems SBA series), and the PID algorithm. The high jet velocity of the CO2 gas stream creates rapid dilution with the ambient air17. Throughout the past 20 years, as seen in Fig. 2a,b, the mean wind speed varied slightly over the site, and was 2.0 and 1.7 m/s in Plot 14 and Plot 3, respectively. These two plots exemplify the variation in wind speed and direction measured at the site and were used in most of the past 20 years (Tables 1, 2). In the instances when the wind speeds dropped below 0.2 m/s, the CO2 fumigation system cycled CO2 gas around the plots to maintain the setpoint as closely as possible.
Testing of different aspects of the CO2 fumigation system at POPFACE and other FACE sites assisted with the development of the CO2 fumigation process used at SoyFACE. However, CO2 was not the only greenhouse gas of interest, and the fumigation system was modified to deliver O3 treatments in addition to CO2. In preparation of the O3 fumigation process, liquid oxygen is stored in a 450-liter cryogenic tank at SoyFACE. The liquid O2 is passed through vaporizing equipment to produce gaseous O2. Then, a generator is used to pass gaseous O2 through a high-voltage dielectric field (~6000 volts) inside a generator (PCI-Wedeco Model GA40 prior to 2005; currently Ozonia Ozat Model CFS-3 2 G). The dielectric field forces some of the O2 molecules to disassociate and recombine to form O3, producing up to 3.5 kg of O3 per day. Due to the toxicity of O3 and the fact that the generator produces the gas at a low pressure, it must be pressurized and mixed with compressed air before being transported to the SoyFACE plots. This is accomplished with a bypass venturi differential pressure injector (Mazzei Injectors Model 384-X), which forces O3 to mix into the higher-pressure air stream. The compressed air stream enters the bypass pressure injector at 90 PSI, O3 enters the pressure injector at a low pressure of 8 PSI, and the resulting O3-air mixture has a pressure of 35 PSI. The change in pressure allows the O3 gas mixture to be delivered from the pressure injector to the manifold. Computer-controlled linear flow valves (PCI-Wedeco Model GA40 prior to 2005; currently Teledyne Hastings Model HCF-302) control the release of O3-enriched air into the wind. The concentration of O3 in the center of each octagonal plot is monitored with an O3 analyzer (Thermo Fisher Scientific Model 49 C/49I), and that information is used to control the setpoint with the PID algorithm as described for CO2 fumigation.
The gas concentrations and wind data are transmitted to a central computer located in an onsite trailer for general data storage and performance analysis. The control computer uses the wind direction measurements to control which main sector of the octagonal treatment plot releases CO2 or O3-enriched air, with the two neighboring octagon sectors releasing a smaller amount of CO2 or O3. Since the three fumigation entry sectors are most directly upwind, following the high-pressure valve release, the gases distribute evenly throughout the octagon and dilute to the background gas concentrations within ~100 meters of the plot.
Data collection & processing methods
Wind serves a crucial role in the fumigation experiments by dispersing CO2 and O3 throughout each treatment plot, and therefore accurate wind data are important. To this end, wind data for each plot are recorded at 4-second intervals along with the CO2 and O3 fumigation levels, setpoint gas levels, and flow valve settings. In particular the wind direction data have been analyzed at SoyFACE, with the prevailing wind direction at the experiment site found to be South/Southwest with some variation between treatment plots due to the difference in location (Figs. 2, 3). The central computer at SoyFACE receives the 4-second and 1-minute fumigation data files (which are averaged from the 4-second data files), and stores both sets of data. Occasionally, there have been data losses and errors at SoyFACE due to extreme weather or technological issues with the analyzers or computer systems. The use of customized Matlab computer code and functions (detailed in the Matlab Files sub-section of the Data Records section) allows the identification of gas measurements that are outside the expected threshold, or ‘filter window’, along with possibly erroneous repeated values. The filter window for [CO2] was determined to be 250–1500 ppm, while the filter window for [O3] was 0–500 parts per billion (ppb). After the processing and quality control of the 1-minute fumigation data files, the fumigation measurements were averaged (i.e., the mean and median values were calculated from the 60 1-minute measurements over each hour) via computer code to produce hourly fumigation files37 (File 11). The hourly fumigation files have proven to be useful, as they are more commonly used in plant growth models than the 4-second and 1-minute fumigation data files. Ambient [CO2] and [O3] are measured at a central location in the SoyFACE field and stored in 1-minute ambient files in the central computer, which have also been consolidated into hourly files.
Precipitation and solar radiation data were recorded at the Water and Atmospheric Resources Monitoring (WARM) station in Champaign (https://warm.isws.illinois.edu/warm/), while other weather metrics such as temperature, wind speed, and relative humidity were recorded at the Surface Radiation (SURFRAD) station about 16 km southwest of the SoyFACE farm (https://gml.noaa.gov/grad/surfrad/bondvill.html). Collectively, the WARM and SURFRAD meteorological data are referred to as CMI weather data due to the stations’ proximity to the University of Illinois-Willard Airport (CMI). Computer code that is publicly available on GitHub (https://github.com/eloch216/oscillator-based-circadian-clock-analysis/) can be used to process and consolidate the CMI weather data collected from the WARM and SURFRAD stations into hourly files from 1995 through 2021. These hourly CMI weather data files have been combined with hourly ambient [CO2] and [O3] data into ‘Hourly Weather and Ambient Data’ files37 (File 6), which provide a summary of the atmospheric and meteorological conditions for each year of the SoyFACE experiments.
Additional useful metrics include the O3 exposure indices AOT40, SUM06, and W126 (Eqs. 2–4), which are included in the ‘SoyFACE Hourly Fumigation Data’ files. In these equations, i = 1, …, n represent the daylight hours between 8:00 AM and 7:00 PM. The O3 exposure indices were measured in parts per billion initially and then converted to parts per million. Unlike the AOT40 and SUM06 indices, which ignore all O3 concentrations below a certain threshold, the W126 index is a sigmoidal weighted function that gives preferential treatment to higher concentrations of O3 up to 100 ppb (0.1 ppm) without ignoring the lower concentrations40. All three O3 indices measure the accumulated daytime O3 exposure throughout the entire growing season.
Another important metric that can be computed from the SoyFACE experimental data is the fumigation target percentages (Table 3). This metric calculates the proportion of minutes that the CO2 or O3 level is within 10% or 20% of the setpoint when the fumigation system is turned on, which provides insight into the efficiency and accuracy of the SoyFACE experiments. These data can also be found in the ‘Fumigation Target Percentages’ file; the code that generates the data is described in the supplementary explanatory file37 (Files 2-3).
The data records cited in this work are stored in the Illinois Data Bank, which is a public access repository for publishing research data from the University of Illinois at Urbana-Champaign (https://databank.illinois.edu/). This data set consists of 8 files and 4 zipped folders37. Descriptions of these data records are as follows.
SoyFACE plot information 2001–2021 File
This file describes the SoyFACE experiments between 2001 and 2021, including the fumigation treatment type (CO2, O3, or a combination treatment), crop species, the plot (also referred to as ‘ring’) numbers used for each experiment, planting, treatment, and harvesting dates, and the gas setpoints37 (File 12). The full data are also contained in Tables 1–2.
SoyFACE 1-minute fumigation data files
The raw fumigation data at SoyFACE are initially recorded as 4-second data files, and subsequently averaged to create 1-minute data files. The 1-minute raw data have been quality controlled to remove erroneous repeated values (Data_Issues_Finder custom code37 (Files 1 and 9).
The quality controlled 1-minute data files are named ‘Avg_MMDDYY’ and contained in the ‘SoyFACE 1-Minute Fumigation Data Files’ folder. The ‘SoyFACE 1-Minute Fumigation Data Explanation’ file contains the column descriptions, units of measurement, and other important notes37 (File 8).
SoyFACE hourly fumigation data files
The hourly SoyFACE fumigation files are generated by averaging the CO2 and O3 fumigation data from the quality controlled 1-minute data files, ignoring values outside the filter window as described in the Data Collection & Processing Methods sub-section. The hourly fumigation files also include ozone exposure metrics AOT40, SUM06, and W126.
The hourly fumigation files are named ‘YYYY_HrlyFumData_ByRing’ and contained in the ‘SoyFACE Hourly Fumigation Data Files’ folder. The ‘SoyFACE Hourly Fumigation Data Explanation’ file contains the formulas for ozone exposure indices AOT40, SUM06, and W126, details about the custom code used to create the hourly fumigation files, and column descriptions for the files37 (File 10).
Hourly weather and ambient data files
The 1-minute ambient CO2 and O3 data are used to generate hourly ambient data files using the same methods that generate the hourly fumigation data files. The hourly weather data collected from the WARM and SURFRAD stations are combined with the hourly ambient SoyFACE data into single files for each year of the experiments.
The hourly weather and ambient data files are named ‘YYYY_HrlyWeatherData’ and contained in the ‘Hourly Weather and Ambient Data Files’ folder. The ‘Hourly Weather and Ambient Data Explanation’ file contains the column descriptions, units of measurement, and other important notes37 (File 4).
Fumigation target percentages file
The target fumigation percentages file shows the proportion of minutes during each growing season that the fumigation CO2 and O3 levels are within 10% and 20% of the target concentrations (setpoints) for the SoyFACE experiment when the fumigation system is turned on. The ‘Fumigation Target Percentages Explanation’ file contains details about the custom code used to create the ‘Fumigation Target Percentages’ file, and column descriptions for the file37 (File 2). The full data from this file are also contained in Table 3.
There are several custom Matlab files37 (File 7) that were created to process and quality control the ‘SoyFACE 1-Minute Fumigation Data’ files, and to generate the ‘SoyFACE Hourly Fumigation Data’ and ‘Fumigation Target Percentages’ files, as enumerated below:
rings_for_year: The rings_for_year function takes a specific year as user input and generates a list of the rings (plots) used in that year’s SoyFACE experiments as the output variable.
Data_Issues_Finder: The rings_for_year function must be run prior to running this code; the user inputs a specific year into that function, and the output is stored as a variable. This output variable is then used as an input for the Data_Issues_Finder code, which loops through the SoyFACE 1-minute raw data files for the year and identifies fumigation measurements that are potentially erroneous by keeping a record of all values that are repeated from one minute to the next. Once the output file has been generated, the user must use qualitative analysis to determine which fumigation measurements are actually erroneous (by comparing the repeated fumigation values to the ambient CO2 and O3 concentrations, considering the number of repeats in a row, etc.). Usage details can be found in the ‘Data_Issues_Finder Code Explanation’ file37 (File 1).
fum: The fum function stores user input details about a specific fumigation experiment as variables, which can then be accessed by other Matlab functions.
batch: The batch code allows the user to run the HourlyDataFunction function in bulk, for all dates within the growing period (5/1 through 10/15).
HourlyDataFunction: The HourlyDataFunction function takes the output from the fum function as input, along with additional user-provided input. The function uses these inputs to generate output variables such as the file names of the 1-minute fumigation data files, and also calls the HourlyData code so that it does not need to be run separately.
HourlyData: The HourlyData code generates the hourly mean and median fumigation metrics from the quality controlled SoyFACE 1-minute fumigation data files, generating output that is used to create the ‘SoyFACE Hourly Fumigation Data’ files. Values outside of the filter window are ignored in calculations of the hourly mean and median fumigation metrics. Further details about items 3–6 can be found in the ‘SoyFACE Hourly Fumigation Data Explanation’ file.
Target_yearly: The rings_for_year function must be run prior to running this code. The Target_Yearly code calculates the proportion of minutes that the CO2/O3 level is within either 10% or 20% of the setpoint when the fumigation system is turned on, generating output that is used to create the ‘Fumigation Target Percentages’ file. Usage details can be found in the ‘Fumigation Target Percentages Explanation’ file.
Technical validation of the fumigation data set was achieved by regular maintenance and calibration of equipment. CO2 analyzers were calibrated before the start of each growing season and regularly throughout the season using certified gases from an ISO/IEC 17025:2017 accredited source per the manufacturer’s instruction. CO2 calibration is at a single point, 750 ppm, and a verification of 0 ppm. The analyzers self-zero approximately every hour. Ozone analyzers were calibrated before the start of the season with a Thermo Scientific 49 C PS or Thermo Scientific 49i PS ozone transfer standard. The standard was either verified by the United States Environmental Protection Agency as a ‘Level 2’ standard or verified by Illinois Environmental Protection Agency as a ‘Level 3’ standard41. The US EPA ozone verification program is part of a larger program managed by the National Institute for Standard and Technology.
Ozone analyzers were calibrated during the season, using linear regression at 0 ppb, 50 ppb, 100 ppb, 150 ppb, and 200 ppb. Ozone generators, compressors, and pressure regulators were serviced each field season according to their manuals, and parts were replaced as needed. Data from each of the fumigation plots were compared to test for outliers and the need to calibrate sensors. Wind sensors were maintained each field season and calibrated as needed. The custom software system alerted the FACE site managers of communication and electrical problems.
The fumigation target percentages for each year of the SoyFACE experiments between 2001 and 2021 provide a quantitative measurement of the accuracy of the fumigation process at SoyFACE (Table 3). In particular, Table 3 shows that the CO2 fumigation process maintained 20% accuracy (i.e., the measured CO2 value was within 20% of the setpoint) for at least 90% of the time for 17 out of 21 years of CO2 experiments and maintained 20% accuracy at least 80% of the time for all years. The O3 fumigation process maintained 20% accuracy at least 90% of the time for 6 out of 20 years of O3 experiments, at least 80% of the time for 17 out of 20 years, and at least 70% of the time for all years. It is clear that the CO2 fumigation process is able to achieve greater precision than the O3 fumigation process, likely because the target elevated concentration for CO2 was an approximate 50% increase over ambient, whereas the elevated O3 concentration in recent years was a 150% increase over ambient.
This SoyFACE fumigation data set can be used as an input for studies that aim to model the impacts of atmospheric change on crop productivity at field, landscape, or regional scales42,43,44,45. For example, a recent semi-mechanistic model of soybean biochemistry and growth was developed using data from a few years of the SoyFACE experiment45. Having 20 years of SoyFACE data compiled and accessible could enhance the development and testing of such a model. Jin et al.44 investigated the interaction of rising [CO2] and drought stress on regional soybean production, again using only a few years of data from SoyFACE to parametrize the CO2 response. Having the full set of fumigation data from SoyFACE could improve scenario simulations, yielding more useful results. The physiological and agronomic data describing crop responses to elevated [CO2] and [O3] at SoyFACE have been previously published in both original manuscripts25,31,32 and meta-analyses26,46 and are available as supplemental files in those studies. Here, for the first time, we provide the complete hourly and seasonal fumigation information for 20 years of SoyFACE experiments.
The fumigation data set can also be used to explicitly test how wind speed, wind direction, and other environmental factors impact the precision and efficiency of fumigation. Recent studies have hypothesized that rapid fluctuations in CO2 concentration in FACE experiments may reduce the photosynthetic, growth, and yield response of crops to elevated CO2 concentrations47. Compiled data from both ambient and elevated [CO2] plots at SoyFACE may provide additional data to test that hypothesis, and to identify parts of the PID algorithm that could be altered to improve fumigation accuracy and precision.
The Matlab version used for this work is MATLAB R2022b. The custom Matlab code and functions used to generate several of the supplementary files associated with this work are described in the Matlab Files sub-section of the Data Records section. The Matlab code and functions are described in further detail in the ‘Explanation’ files, which are publicly accessible via the Illinois Data Bank37. The Matlab code and functions are contained in the ‘Matlab Files’ folder (File 7), and the underlying data set is contained in the ‘SoyFACE 1-Minute Fumigation Data Files’ folder (File 9).
The R statistical programming language and openair package are required in order to use the windRose function (Figs. 2, 3). The R version used for this work is Rx64 4.1.2, which is free for all users to download. The underlying data set is contained in the ‘SoyFACE 1-Minute Fumigation Data Files’ folder (File 9).
Tripati, A. K., Roberts, C. D. & Eagle, R. A. Coupling of CO2 and ice sheet stability over major climate transitions of the last 20 million years. Science. 326, 1394–1397 (2009).
Liu, P. R. & Raftery, A. E. Country-based rate of emissions reductions should increase by 80% beyond nationally determined contributions to meet the 2 °C target. Commun. Earth Environ. 2, 29 (2021).
Yeung, L. Y. et al. Isotropic constraint on the twentieth century increase in tropospheric ozone. Nature. 570, 224–227 (2019).
Ainsworth, E. A., Yendrek, C. R., Sitch, S., Collins, W. J. & Emberson, L. D. The effects of tropospheric ozone on net primary productivity and implications for climate change. Annu. Rev. Plant Biol. 63, 637–661 (2012).
McGrath, J. M. et al. An analysis of ozone damage to historical maize and soybean yields in the United States. Proc. Natl. Acad. Sci. USA (PNAS). 112, 14390–14395 (2015).
Ainsworth, E. A., Lemmonier, P. & Wedow, J. M. The influence of rising tropospheric carbon dioxide and ozone on plant productivity. Plant Biol. 22, 5–11 (2019).
Mangin, L. et al. Sur la vegetation dans une atmospherie viciée par la respiration. C. R. Acad. Sci. Paris. 122, 747–749 (1896).
Strain, B. R. & Bazzaz, F. A. in CO2 and Plants: The Response of Plants to Rising Levels of Atmospheric Carbon Dioxide (ed. Lemon, E. R.) Ch. 7, 177–222 (Westview Press, 1983).
Allen, L. H. Jr., Drake, B. G., Rogers, H. H. & Shinn, J. H. Field techniques for exposure of plants and ecosystems to elevated CO2 and other trace gases. CRC Crit. Rev. Plant Sci. 11, 85–119 (1992).
Long, S. P., Ainsworth, E. A., Rogers, A. & Ort, D. R. Rising atmospheric carbon dioxide: Plants FACE the future. Annu. Rev. Plant Biol. 55, 591–628 (2004).
Porter, A. S., Evans-FitzGerald, C., McElwain, J. C., Yiotis, C. & Elliott-Kingston, C. How well do you know your growth chambers? Testing for chamber effect using plant traits. Plant Methods. 11, 44 (2015).
McLeod, A. R. & Long, S. P. Free-air carbon dioxide enrichment (FACE) in global change research: A review. Adv. Ecol. Res. 28, 1–56 (1999).
Arp, W. J. Effects of source-sink relations on photosynthetic acclimation to elevated CO2. Plant Cell Environ. 14, 869–875 (1991).
Thomas, R. B. & Strain, B. R. Root restriction as a factor in photosynthetic acclimation of cotton seedlings grown in elevated carbon-dioxide. Plant Physiol. 96, 627–634 (1991).
Harper, L. A., Baker, D. N., Box, J. E. Jr. & Hesketh, J. D. Carbon dioxide and the photosynthesis of field crops: A metered carbon dioxide release in cotton under field conditions. Agron. J. 65, 7–11 (1973).
Miglietta, F. et al. Free-air CO2 enrichment (FACE) of a poplar plantation: The POPFACE fumigation system. New Phytol. 150, 465–476 (2001).
Long, S. P., J.G. R.F. et al.) Ch. 1, 3 – 13 & Ch. 2, 15–43 (Springer, 2006).
Messerli, J. et al. Performance of low-cost open-top chambers to study long-term effects of carbon dioxide and climate under field conditions. Agron. J. 107, 916–920 (2015).
Leakey, A. D. B., Bishop, K. A. & Ainsworth, E. A. A multi-biome gap in understanding of crop and ecosystem responses to elevated CO2. Curr. Opin. Plant Biol. 15, 228–236 (2012).
Montes, C. M., Demler, H. J., Li, S., Martin, D. G. & Ainsworth, E. A. Approaches to investigate crop responses to ozone pollution: from O3-FACE to satellite-enabled modeling. Plant J. 109, 432–446 (2022).
Rogers, A. et al. Leaf photosynthesis and carbohydrate dynamics of soybeans grown throughout their life-cycle under Free-Air Carbon dioxide Enrichment. Plant Cell Environ. 27, 449–458 (2004).
Bernacchi, C. J., Morgan, P. B., Ort, D. R. & Long, S. P. The growth of soybean under free air [CO2] enrichment (FACE) stimulates photosynthesis while decreasing in vivo Rubisco capacity. Planta. 220, 434–446 (2005).
Leakey, A. D. B., Bernacchi, C. J., Dohleman, F. G., Ort, D. R. & Long, S. P. Will photosynthesis of maize (Zea mays) in the US Corn Belt increase in future [CO2] rich atmospheres? An analysis of diurnal courses of CO2 uptake under free-air concentration enrichment (FACE). Glob. Chang. Biol. 10, 951–962 (2004).
Leakey, A. D. B. et al. Photosynthesis, productivity, and yield of maize are not affected by open-air elevation of CO2 concentration in the absence of drought. Plant Physiol. 140, 779–790 (2006).
Morgan, P. B., Mies, T. A., Bollero, G. A., Nelson, R. L. & Long, S. P. Season-long elevation of ozone concentration to projected 2050 levels under fully open-air conditions substantially decreases the growth and production of soybean. New Phytol. 170, 333–343 (2006).
Long, S. P., Ainsworth, E. A., Leakey, A. D. B., Nösberger, J. & Ort, D. R. Food for thought: Lower-than-expected crop yield stimulation with rising CO2 concentrations. Science 312, 1918–1921 (2006).
Ruiz-Vera et al. Global warming can negate the expected CO2 stimulation in photosynthesis and productivity for soybean grown in the Midwestern United States. Plant Physiol. 162, 410–423 (2013).
Ruiz-Vera, U. M., Siebers, M. H., Drag, D. W., Ort, D. R. & Bernacchi, C. J. Canopy warming caused photosynthetic acclimation and reduced seed yield in maize grown at ambient and elevated [CO2]. Glob. Chang. Biol. 21, 4237–4249 (2015).
Leakey, A. D. B. Rising atmospheric carbon dioxide concentration and the future of C4 crops for food and fuel. Proc. R. Soc. B. 276, 2333–2343 (2009).
Markelz, R. J. C., Strellner, R. S. & Leakey, A. D. B. Impairment of C4 photosynthesis by drought is exacerbated by limiting nitrogen and ameliorated by elevated [CO2] in maize. J. Exp. Bot. 62, 3235–3246 (2011).
Gray, S. B. et al. Intensifying drought eliminates the expected benefits of elevated [CO2] for soybean. Nat. Plants. 2, 16132 (2016).
Betzelberger, A. M. et al. Effects of chronic elevated ozone concentration on antioxidant capacity, photosynthesis, and seed yield of 10 soybean cultivars. Plant Cell Environ. 33, 1569–1581 (2010).
Yendrek, C. R. et al. High-throughput phenotyping of maize leaf physiological and biochemical traits using hyperspectral reflectance. Plant Physiol. 173, 614–626 (2017).
Choquette, N. E. et al. Uncovering hidden genetic variation in photosynthesis of field-grown maize under ozone pollution. Glob. Chang. Biol. 25, 4327–4338 (2019).
Rosenthal, D. M. et al. Over-expressing the C3 photosynthesis cycle enzyme Sedoheptulose-1-7 Bisphosphatase improves photosynthetic carbon gain and yield under full open air CO2 fumigation (FACE). BMC Plant Biol. 11, Art. 123 (2011).
Köhler, I. H. et al. Expression of cyanobacterial FBP/SBPase in soybean prevents yield depression under future climate conditions. J. Exp. Bot. 68, 715–726 (2017).
Aspray, E. K. et al. SoyFACE Fumigation Data Files. Illinois Data Bank https://doi.org/10.13012/B2IDB-3496460_V1 (2023).
Gray, S. B. et al. Minirhizotron imaging reveals that nodulation of field-grown soybean is enhanced by free-air CO2 enrichment only when combined with drought stress. Func. Plant Biol. 40, 137–147 (2013).
Lewin, K. F., Hendrey, G. R. & Kolber, Z. Brookhaven National Laboratory Free-Air Carbon Dioxide Enrichment Facility. CRC Crit. Rev. Plant Sci. 11, 135–141 (1992).
Musselman, R. & Lefohn, A. The use of critical levels for determining plant response to ozone in Europe and in North America. Sci. World J. 7(Supp. 1), 15–21 (2007).
EPA (2017) Quality Assurance Document. Quality Assurance Project Plan for the Standard Reference Photometer Program. Environmental Protection Agency Office of Air Quality Planning and Standards, Research Triangle Park, NC.
Twine, T. E. et al. Impacts of elevated CO2 concentration on the productivity and surface energy budget of the soybean and maize agroecosystem in the Midwest USA. Glob. Chang. Biol. 19, 2838–2852 (2013).
Fu, T., Ha, B. & Ko, J. Simulation of CO2 enrichment and climate change impacts on soybean production. Int. Agrophys. 30, 25–37 (2016).
Jin, Z., Ainsworth, E. A., Leakey, A. D. B. & Lobell, D. B. Increasing drought and diminishing benefits of elevated carbon dioxide for soybean yields across the US Midwest. Glob. Chang. Biol. 24, e522–e533 (2018).
Matthews, M. L., Marshall-Colon, A., McGrath, J. M., Lochocki, E. B. & Long, S. P. Soybean-BioCro: A semi-mechanistic model of soybean growth. is P. 4, Art. diab032 (2022).
Bishop, K. A., Leakey, A. D. B. & Ainsworth, E. A. How seasonal temperature or water inputs affect the relative response of C3 crops to elevated [CO2]: a global analysis of open top chamber and free air CO2 enrichment studies. Food Energy Security 3, 31–45 (2014).
Allen, L. H. et al. Fluctuations of CO2 in Free-Air CO2 Enrichment (FACE) depress plant photosynthesis, growth, and yield. Agric. For. Meteorol. 284, 107899 (2020).
Carslaw, D. C. & Ropkins, K. “openair – an R package for air quality data analysis”. Environmental Modelling & Software 27-28(0), 52–61, https://doi.org/10.1016/j.envsoft.2011.09.008 (2012).
This research was supported by the USDA ARS and a Headquarters post-doctoral award to EKA. SoyFACE was supported by the Illinois Council for Food and Agricultural Research (CFAR), Archer Daniels Midland Company, Argonne National Laboratory, the International Arid Land Consortium, the Midwestern Center of the National Institute for Climate Change Research, the US Department of Energy (DE-AC02-98CH10886; DE-FG02-04ER63849; DE-FC02–06ER64158, DE-SC0018420), the US Department of Agriculture (Agricultural Research Service, Cooperative State Research, Education and Extension Service National Research Initiative 2002-02723; National Institute of Food and Agriculture 2010–65114–20343, 2010-65114-20355, 2014-67013-21783, 2015-67013-22836), and the National Science Foundation (IBN 0236053; PGR-1238030).
The authors declare no competing interests.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Aspray, E.K., Mies, T.A., McGrath, J.A. et al. Two decades of fumigation data from the Soybean Free Air Concentration Enrichment facility. Sci Data 10, 226 (2023). https://doi.org/10.1038/s41597-023-02118-x