Evaluating Emissions from Select Urban Parking Garages in Cincinnati, OH, Using Portable Sensors and Their Potentials for Sustainability Improvement

Abstract

Urban parking around the world faces similar challenges of inadequate space, pollution, and carbon emissions. Although various smart parking technologies have been tested and implemented, they primarily aim to reduce the time spent searching for parking, without considering the impact on air quality. In this study, the air quality in three urban garages was investigated with portable instruments at the entrance and exit gates and inside the garages. Garage emissions measured include CO₂, PM_2.5, PM₁₀, NO₂, and total VOCs. The results suggested that the PM_2.5 levels in these garages tend to be higher than the ambient levels. The emissions also exhibit seasonal variations, with the highest concentrations occurring in the summer, which are 20.32 µg/m³ in Campus Green, 14.25 µg/m³ in CCM, and 15.23 µg/m³ in Washington Park garages, respectively. PM_2.5 measured from these garages is strongly correlated (with an R² of 0.64) with ambient levels. CO₂ emissions are higher than ambient levels but within the indoor air quality limit. This suggests that urban garages in Cincinnati tend to enrich ambient air concentrations, which can affect garage users and garage attendants. Portable sensors are capable of long-term emission monitoring and are compatible with other technologies in smart garage development. With portable air sensors becoming increasingly accessible and affordable, there is an opportunity to integrate these devices with smart garage management systems to enhance the sustainability of parking garages.

1. Introduction

The US is known for high personal vehicle ownership. Based on a study in 2022, 91.7% of the US households have owned at least one light-duty vehicle (car or light-duty truck) since 2018. This raises several complex sustainability concerns, particularly in the context of growing urban populations and increasing traffic. One major issue is the oversupply and underutilization of parking infrastructure. A study indicated that Americans may have eight parking spots for each vehicle [1] at various locations. This imbalance is largely the result of mandatory parking minimums, where local zoning ordinances require municipalities to have mandatory parking minimums, regardless of actual demand. Historically, this “one-size-fits-all” approach has led to a surplus of parking in many urban areas, with surface lots taking up an estimated 5% to 30% of urban land area [2,3]. This represents missed opportunities for more productive use, especially in cities that are pursuing urban revitalization. In addition to inefficient land use, surface parking lots contribute to environmental challenges such as exacerbating stormwater runoff, increasing the risk of flooding, and contributing to the urban heat island effect [4,5]. The growth of public transit and ride-share services (such as Uber and Lyft) also reduces the need for urban parking. In response, many cities are shifting from surface lots to multi-level or underground parking garages. While this strategy conserves surface space, it introduces new challenges, such as poor air quality.

Meanwhile, rapid urbanization has intensified demands for limited transportation infrastructure, resulting in chronic parking shortages, traffic congestion, and air pollution. Parking infrastructure has a crucial role in urban mobility, but is lacking in advancements in comparison with other transportation improvements, such as alternative fuel vehicles, walkable districts, and public transportation. Traditional parking systems take up significant space but have many inefficiencies and are associated with wasted time and increased air pollution [6,7]. Another concern is greenhouse gas emissions generated by vehicles searching for available parking [8]. It is estimated that 30% of urban traffic congestion is caused by vehicles searching for parking spaces [9]. One study found that drivers took an average of nearly eight minutes to locate a parking space in a small business district of Los Angeles, CA, alone. This resulted in 47,000 gallons of fuel burned annually, producing approximately 730 tons of carbon dioxide [9].

In recent years, cities and developers have turned to high-tech ways to manage parking garages. This includes the integration of efficient ventilation systems, sensor networks, and smart parking technologies, etc. Smart garage technology, such as real-time occupancy detection, digital signage, pre-assigned parking spaces or LED lights showing spaces available, and parking guidance systems, optimizes the use of parking spaces and improves traffic flow [10], while inductive loops at entrances and exits can yield an accurate vehicle count. The benefits of smart parking systems include reduced time searching for parking, reduced fuel consumption, reduced air pollutant emissions, and health benefits from a reduction in exposure. As an example, a study focused on integrating CO sensors into the building information modeling (BIM) systems and auto-start ventilation when the concentration reached alarm levels [11]. Advanced ventilation systems will then use demand-controlled ventilation to adjust fan speeds and air exchange rates based on real-time pollution levels. Smart parking systems have been increasingly studied and practiced to combat air pollution and greenhouse gas emissions. The benefits of smart parking systems are significant, including reduced time searching for parking, reduced fuel consumption, enhanced user experience, and reduced air pollutant emissions. These innovations not only address current parking challenges but also reduce carbon footprints.

However, the focus of smart parking systems is to reduce time searching for parking, save fuel, and improve users’ experience, etc., with air pollutant reduction as an indirect benefit. Air pollutant sensors have not yet been reported in these smart parking garage initiatives. Existing studies indicated that air pollutant levels in urban garages tend to be higher than in the ambient environment. A study conducted in an eight-story garage in Baltimore, Maryland, USA, during June and July 2002 found that the levels of carbon monoxide, particle-bound PAHs, and benzene were 2–9 times higher during weekdays than weekends, and the concentrations were correlated with traffic volumes [12].

Parking management is particularly challenging in urban university settings due to limited parking spaces and continuously growing student populations [10]. Universities often face inadequate student parking capacities and fail to balance demand, pricing, and accessibility [13]. As a result, parking garages can act as localized pollution “hot-spots”, where pedestrians, garage attendants, and other users are exposed to elevated levels of air pollution [12]. The combination of enclosed structural design and localized vehicle activity, especially during peak hours, creates conditions that can lead to the accumulation of air pollutants that exceed outdoor concentrations. Studies show that pollutant concentrations inside parking garages can be up to five times higher than the surrounding urban ambient air [12]. This problem is particularly acute in underground and or enclosed structures, where limited natural ventilation allows emissions such as particulate matter, carbon monoxide, and nitrogen dioxide to concentrate. These pollutants stem primarily from idling car engines and stop-and-go driving.

Carbon monoxide measurements were conducted in the underground College Conservatory of Music (CCM) garage at the University of Cincinnati in 2010 using a portable CO measurer (Lagan T15). Higher concentrations were observed during the evening rush hours (16–18:00), with an average CO concentration of 4.4 ppm for spring, 4.7 ppm during the summer, and 3.9 ppm during the fall. The sports events usually occurred during nonpeak hours, and the long line of vehicles exiting the garage tends to result in higher CO concentrations. The average hourly CO concentrations ranged from 2.9 to 15.7 ppm, the latter was observed at the exit gate, between 22:30 and 23:30 on 11 November 2010, after a football game [14]. Although these values are within the 1 h limit of 35 ppm based on the US EPA’s National Ambient Air Quality Standards, they are much higher than the ambient values, which are mostly below 1 to 2 ppm. Up to 600 ppb of CO was reported from another garage study at the University of British Columbia in Vancouver, Canada [15]. CO, NOx, and PM_2.5 concentrations were measured with low-cost sensors. Higher concentrations were observed at the center of the garage instead of at the gate, but the levels are lower than the EPA’s standards.

During peak traffic periods, such as rush hour or university course times, parking garages can experience a high volume of vehicle turnover, further exacerbating pollution levels. Drivers often circle the garage multiple times searching for a space, which significantly contributes to both congestion and emissions. Additionally, vehicle idling is a common issue, particularly in university garages where students may wait in their cars until class time. However, there have not been studies of other pollutants in urban garages in Cincinnati, OH. Meanwhile, there is a growing interest in identifying more sustainable approaches to garage design and operation to reduce carbon footprints. This includes the integration of high-efficiency ventilation and air filtration systems, real-time air quality monitoring, and smart parking technologies. Improving the environmental performance of parking garages is essential for public health and better sustaining urban infrastructure.

Therefore, this study aims to collect and compare emissions data from a range of parking garages within a mid-sized and growing city in the Midwest USA, with a focus on understanding air pollutant emissions from urban garages and associated contributing factors. This study examined two university garages and one downtown garage with differing ventilation configurations. Seasonal garage emission data were systematically collected with portable instruments (also known as air sensors) and then compared with urban air quality data obtained from a local monitoring site. This comparison aims to understand the extent to which pollution within parking garages deviates from general outdoor conditions and contributing factors. These are important in identifying high-pollution scenarios and creating strategies for improving air quality within parking garages and the sustainability of the parking system.

2. Materials and Methods

2.1. Portable Instruments Used

Traditional methods of monitoring vehicle emissions have relied on reference-grade instrumentation, which is highly accurate but expensive and complex to operate. In contrast, the development of low-cost sensors (LCSs) has emerged as a promising alternative for monitoring air quality and emissions more dynamically. These compact devices are capable of detecting and recording a range of pollutants. While they may not yet match the precision of reference-grade instruments, LCSs have been shown to correlate well with high-end instruments [16]. This study uses the advantages of LCS technology to measure and analyze vehicular emissions within parking garage environments.

LCSs by the names of Temtop M2000 (Elitech Technologies Inc., San Jose, CA, USA) multifunctional air quality detector and Plume Labs Flow-2 (Plume Labs, Paris, France) were used in the study. Temtop M2000 is a handheld device that measures particulate matter (PM_2.5 and PM₁₀), carbon dioxide (CO₂), and formaldehyde (HCHO). It is equipped to alert if concentrations exceed permissible limits. This device can also measure temperature, relative humidity, and particle count. Temtop PM is measured with a laser sensor where particles pass through the light source, combined with a scattering principle and a smart particle inversion algorithm to determine PM_2.5 and PM₁₀ mass concentrations. Its PM_2.5 measurement has an accuracy of ±10 µg/m³ in the 0–100 µg/m³ range and ±15 µg/m³ (0–100 µg/m³) for PM₁₀, both with a resolution of 0.1 µg/m³. CO₂ is measured with a Sense Air NDIR sensor, which uses non-dispersive infrared gas sensing technology based on the Lambert–Beer Absorption Law, combined with built-in CO₂ gas analysis to determine concentrations [17]. Its CO₂ sensor has an accuracy of ± 50 ppm with a resolution of 1 ppm. Based on our previous experience with this device [18], results from the PM_2.5, PM₁₀, and CO₂ concentrations are used in this study.

Flow-2 measures nitrogen dioxide (NO₂), volatile organic compounds (VOCs), and particulate matter (PM₁, PM_2.5, and PM₁₀). It reports both the concentrations and in the form of an air quality index (AQI) [19]. The Flow PM sensor uses a laser particle counter to measure particulate matter concentrations through light scattering, and a heated oxide sensor for gases.

The Flow-2 has GPS capability, which tracks the location in real time via the Plume Labs app. Both pollution and position data are stored in the cloud and can be downloaded locally. It complements the Temtop unit with NO₂ and VOC concentrations, while the PM measured from this sensor was not used due to lower accuracies reported by our group [18]. A brief summary of sensor properties is also included in Table S1.

2.2. Garage Selections

Three garages were selected in this study. The two garages on the University of Cincinnati (UC) campus are called the Campus Green, which has no exhaust system but does have a carbon monoxide detection system, and the College Conservatory of Music (CCM), an underground garage that has an exhaust system that runs on 25% power during fall, winter, and spring, and 5% in summer. The university garages are used heavily by commuter students, with Campus Green being the largest on West Campus and the most often used garage. UC is in an urban community with approximately 53,200 total students and 12,000 faculty and staff, a typical size for a public higher education institution in the U.S [20]. The percentage of commuters at the University of Cincinnati’s main campus is 82.90% as of 2022, many of whom use the university parking facilities [21].

The third one is the Washington Park garage, located in downtown Cincinnati, and is fully enclosed. The Washington Park garage is used mainly by commuters to office buildings in downtown Cincinnati. Garages in the downtown area have a wide range of parking spots, ranging from 165 to 1100. However, most garages in the area average roughly 400–600 spaces [22].

The Campus Green garage uses natural ventilation and has two gate locations. The underground gate has three entrance lanes and three exit lanes, while the ground level gate has two entrance lanes and two exit lanes. The main gate at the CCM garage has two entrance lanes and two exit lanes, while the other two are one-way gates, with either one lane entrance or one exit lane. There is a fan at the bottom level that can be activated by high CO levels (35 ppm). The Washington Park garage has two gates that lead to different streets. One gate contains one entrance and one exit lane, while the other has one entrance, one exit, and one reversible lane, which changes direction throughout the day. Details of these garages are summarized in

Table 1. Details of the three garages.

Garage Name	Location	Garage Type	Number of Parking Spaces	Number of Gates	Underground or Ground Level
Campus Green	UC	Semi-Enclosed	1594	2	1 Ground, 1 Underground
CCM	UC	Fully Enclosed	404	3	Underground
Washington Park	Downtown	Fully Enclosed	450	2	Underground

.

Like most urban universities, UC garages often face inadequate student parking capacities and fail to balance demand, pricing, and accessibility [13]. Commuter students comprise the largest share of parking users, together with faculty, staff, and visitors. Likewise, implementation of smart technologies has been slow, and limited to signage outside the garage to indicate “LOT FULL” and a data collection system at the entrance [10,21]. The downtown garage also has “LOT FULL” signage, but it is not an automatic digital sign. Instead, it is a free-standing sign that needs to be put out by employees. There is room for more sustainable technologies compared with other garages in the US and around the world.

2.3. SWOAQA Taft NCore Site

Ambient air quality data were acquired from the Southwest Ohio Air Quality Agency (SWOAQA) Taft National Core (NCore) Site located at 250 William Howard Taft Road, Cincinnati, Ohio. The Taft NCore site provides regulatory-grade air quality data in compliance with U.S. EPA standards. Specifically, the data used in this study included daily and hourly average concentrations of PM_2.5 and PM₁₀.

This monitoring site was selected due to its proximity to the study area, 0.67 miles away from the UC campus and 1.45 miles away from the downtown garage (Figure 1). The central location of this site among the garages ensures that its air quality data are representative of the ambient atmospheric conditions for comparison with the garages.

2.4. Sampling and Analytical Methods

For sampling, monitoring instruments were placed on top of ticket boxes in each garage. Figure 2a and Figure 2b illustrate the sensor placement at the parking garage gate, where the arrow indicates the location of the ticket box. Vehicles stop at the box to either obtain a ticket, scan the parking validation, or wait for the gate to open if the vehicle has automatic validation. The number of vehicles entering and exiting the garage was recorded in 10-min intervals to better understand how traffic influences pollutant concentrations at gate locations. These data were then used to make correlations between fluctuations in pollutant concentrations and changes in traffic volume to determine how emissions are affected by traffic patterns. In addition to gate monitoring, this study also included a “walking” segment, where the tester carried the sensors while walking inside the garages at various locations.

Garage emission and vehicle volume data were collected on weekdays that aligned with typical commuter activity patterns. The majority of the measurements were taken during rush (peak) hours, where the morning rush hour was from approximately 8:00 AM to 9:00 AM and the evening rush hour was from 4:00 PM to 5:00 PM. In theory, these timeframes correspond to peak vehicle movements from students, faculty, and staff. Other monitoring times outside of rush hours were grouped into broader AM or PM hour categories. Temporal classifications enable analysis of how traffic-related emissions fluctuate throughout the day. Measurements were carried out from June 2023 to April 2025 and were randomly selected every month to ensure representation of the data. A total of 85 measurements were taken at the Campus Green garage, 45 at CCM, and 34 at the Washington Park garage.

This study also investigated seasonal trends of garage emissions. Seasons were determined based on a previous study in the Cincinnati area [23] as follows: winter includes December, January, and February; spring consists of March, April, and May; summer includes June, July, and August; and fall includes September, October, and November. Statistical analysis (such as the mean, median, and standard deviation) and data regression were performed with Microsoft Excel and R (2024.12.0+467).

3. Results

3.1. Emissions from All Garages

Figure 3 shows the average PM_2.5 and PM₁₀ averages distributed monthly. Seasonal trends suggest that the highest particulate concentrations were reported during the summer months. Both PM_2.5 and PM₁₀ concentrations increase steadily starting from February (winter), with the most significant jump occurring between May and June. In June, average PM_2.5 concentrations reach a high of 21.73 µg/m³ and 37.46 µg/m³ for PM₁₀. These PM_2.5 concentrations are higher than the primary annual EPA National Ambient Air Quality Standard (NAAQS) of 9.0 μg/m³ and the secondary standard of 15.0 μg/m³, but less than the 24 h limit of 35.0 μg/m³.

Table 2. Summary of annual PM_2.5 measurements.

Garage	Sensor Orientation	Average Temtop M2000 PM2.5 (μg/m3)	Average Taft Site PM2.5 (μg/m3)
Campus Green	Ground Gate	12.55	10.00
	Underground Gate	10.72
	Walking	14.34
CCM	Underground Gate	10.95	10.19
	Walking	10.33
Washington Park	Underground Gate	10.99	9.78
	Walking	10.90

shows the annual PM_2.5 concentrations at the three garages, together with the concentrations observed at the Taft station during the same time as the garage measurements. Emissions at the garages are higher than those from the ambient levels. The PM_2.5 levels when walking in the Campus Green garage were the highest, which is consistent with [16], while PM_2.5 exposure when walking in other garages was close to those at the gates. All values are slightly higher than the NAAQS annual standard of 9.0 μg/m³.

Figure 4 shows PM_2.5 concentrations by garages and sensor locations. Campus Green ground level measurements show high variability with concentrations ranging from 1.80 µg/m³ to 28.59 µg/m³ with a median of 9.72 µg/m³. Underground gate concentrations are lower with less variation, with a similar median of 9.63 µg/m³. Walking measurements show a broader spread of concentrations representing the dataset’s highest median and maximum concentrations. The mean walking concentration is 14.34 µg/m³ for Campus Green, 10.33 µg/m³ for CCM, and 10.90 µg/m³ for Washington Park. Both underground and walking measurements for CCM show lower overall concentrations and less variability, with medians from 7 to 9 µg/m³. However, outliers are more frequent within the CCM dataset than in the other two garages. Washington Park underground measurements show slightly higher concentrations than walking measurements.

Table 3. Seasonal trends of PM_2.5 concentrations in three garages.

Season	Garage	Temtop M2000 PM2.5 (μg/m3) (Mean ± SD)	Taft Site PM2.5 (μg/m3) (Mean ± SD)
Fall	Campus Green	10.37 ± 6.35
	CCM	8.10 ± 2.60	9.47 ± 2.60
	Washington Park	9.00 ± 4.31
Winter	Campus Green	7.52 ± 2.98
	CCM	10.55 ± 3.18	9.88 ± 3.68
	Washington Park	10.82 ± 4.16
Spring	Campus Green	9.32 ± 5.14
	CCM	6.93 ± 3.68	7.11 ± 2.78
	Washington Park	8.34 ± 3.00
Summer	Campus Green	20.32 ± 7.19
	CCM	14.25 ± 9.64	14.20 ± 6.17
	Washington Park	15.23 ± 8.13

shows the seasonal trends of PM_2.5 concentrations in the three garages in comparison with those from the Taft site in means and standard deviations. The highest concentrations were observed during the summer for all three garages and Taft, while the traffic volumes tend to be the lowest due to the break. This seasonal trend is consistent with a previous study at the Taft site [23] and is likely due to both meteorological factors and the hot soak of vehicles parked in garages. Campus Green has the highest values in the fall, spring, and summer among the three garages. This is likely due to the much higher traffic volumes. The underground garages (CCM and the Washington Park) did not lead to higher emissions, which can be due to the much smaller sizes or the exhaust fans. The lowest concentrations were observed in Winter in Campus Green and in Spring in CCM, Washington Park, and Taft.

Data variations, as indicated by the standard deviation, are the highest during the summer. The variations are generally higher in garage emissions than in the ambient, which may be due to less stability of the sensors.

These findings are largely consistent with the results reported by Gonzales et al., 2022, where PM_2.5 emissions in garages were higher than those observed from ambient monitoring stations, suggesting that garages can be local air quality hotspots for higher exposures [24].

Figure S1 presents a box plot of the seasonal PM₁₀ trends of the three garages in comparison with those from the Taft site. Consistent with that of PM_2.5, the highest coarse particle level also occurred in the summer, with higher values in the garages. However, the ambient PM₁₀ levels are higher in fall and winter.

Table 4. High PM_2.5 emission days at each garage location.

Garage	Date	Average Temtop M2000 PM2.5 (μg/m3)	Average Taft Site PM2.5 (μg/m3)	Traffic Volume (Cars per Hour)	AQI Alert
Campus Green	21 June 2024	28.59	23.20	47	No
	20 June 2024	27.53	21.50	62	No
	9 July 2024	26.70	14.50	60	No
CCM	17 July 2023	32.44	32.20	51	Yes
Washington Park	19 June 2024	25.36	19.50	26	No

summarizes the days with the highest PM_2.5 emissions at each garage, and all occurred in the summertime. All except one are much higher than the ambient values, while traffic volume was lower than the median. The PM_2.5 concentration observed on 17 July 2023 in CCM is likely due to the impact of the Canadian wildfires, which occurred from May through July of 2023, where smoke was dispersed by the North American trough and severely impacted PM concentrations [15]. Time series of PM_2.5 and CO₂ levels of these three garages are shown in Figures S2–S4, which suggest that these two components are generally not correlated. In addition, an average PM_2.5 concentration of 10.33 µg/m³ was observed on 14 April 2025, one of the highest in spring, during the new control console installation at the CCM garage.

Table 5. Summary of PM_2.5/PM₁₀ ratios in garages and the ambient values.

Season	Measurement Site	PM2.5/PM10 Ratio
Fall	Garage	0.59
	Taft	0.51
Winter	Garage	0.57
	Taft	0.52
Spring	Garage	0.59
	Taft	0.52
Summer	Garage	0.59
	Taft	0.57

summarizes PM_2.5/PM₁₀ ratios between the garage and Taft site by season. All the ratios (at all seasons) are greater than 0.5, which is an indication that fine particles are major contributors to PM emissions. The ratios are higher at garage sites than in the ambient conditions, since traffic emissions contribute more fine particles at garage sites. Consistent with another study [25], the ratio was the lowest during winter, which may be related to humidity.

CO₂ levels in the three garages are shown in

Table 6. Summary of garage CO₂ concentrations.

Season	Garage	Temtop M2000 CO2 (ppm) (Mean ± SD)
Fall	Campus Green	572.9 ± 122.3
	CCM	549.9 ± 57.7
	Washington Park	526.3 ± 57.9
Winter	Campus Green	548.8 ± 131.3
	CCM	526.2 ± 79.3
	Washington Park	517.2 ± 107.9
Spring	Campus Green	515.7 ± 91.0
	CCM	610.3 ± 110.3
	Washington Park	684.9 ± 131.8
Summer	Campus Green	540.9 ± 151.4
	CCM	606.0 ± 143.1
	Washington Park	570.5 ± 131.6

. Although the seasonal mean CO₂ concentrations are higher than that of ambient (400 ppm levels), they are much less than the ASHRAE guideline of 1000 ppm [26]. Higher standard deviations during the summer months indicate more CO₂ fluctuation compared to other seasons.

CO₂ levels in Campus Green are relatively consistent with the least variations, which is largely due to the natural ventilation. Washington Park in the spring showed the highest average CO₂ emissions. The highest emission day was on 16 March 2024, with an average CO₂ level of 909.4 ppm. The highest CO₂ emissions in CCM were also in spring. The highest emission day was on 5 June 2024, with an hourly average of 1050.2 ppm. A CO₂ increase in spring and summer in the two enclosed garages is likely the result of the temperature increase. During the days when CO₂ levels are close to 1000 ppm, it is desirable to start the ventilation and replenish fresh air.

NO₂ and VOC data were also measured with Flow2 sensors and are shown in Tables S2 and S3. Table S2 summarizes seasonal NO₂ concentrations among the three garages, with the highest NO₂ concentrations occurring in the summer and the lowest in winter for all garages, which is expected. The summer concentrations were comparable to those reported in [24], which were 15.4 ± 3.6 ppb. NO₂ levels observed are far less than the NAAQS 1 h limit of 100 ppb. Since Flow-2 tends to underestimate NO₂ values [27], direct comparison with FEM/FRM data should be cautioned.

Table S3 summarizes seasonal VOC concentrations among the three garages. No clear seasonal trends were observed, and the variations were high, which may be due to the emission characteristics or instability of the VOC sensor.

3.2. Emission Characteristics of Individual Garages

PM_2.5 concentrations grouped by different traffic periods at each garage are shown in Figure 5a–c, and the averages are summarized in

Table 7. Average PM_2.5 concentrations during different traffic periods (μg/m³).

Garage	AM Rush	AM Non-Rush	PM Rush	PM Non-Rush
Campus Green	15.33	7.44	11.07	7.07
CCM	7.42	7.42	7.74	13.04
Washington Park	11.51	7.01	9.25	15.78

. These data reflect the usage characteristics of the garages.

At Campus Green Garage (Figure 5a), PM_2.5 concentrations during rush hours are higher than those of non-rush hours, which indicates the garage is mainly used by commuters on regular schedules. PM_2.5 concentrations at the AM rush hour had the highest mean value of 15.33 µg/m³ and maximum concentrations reaching 30 µg/m³, which is followed by those of the PM rush hour with a mean of 11.07 µg/m³. PM_2.5 concentrations at non-rush hours are 51% and 35% lower than AM and PM rush hours, respectively.

During air quality sampling of the Campus Green Garage, it is frequently observed that drivers spend extended time circling in search of available parking spots. This repeated motion of cars contributed to increased vehicle idling and stop-and-go movement, which are conditions known to significantly increase pollutants such as PM, VOCs, and NO_x.

In contrast, a different trend is observed at the two enclosed garages. PM_2.5 concentrations at PM non-rush hours are the highest.

Figure 5b presents PM_2.5 concentrations for the CCM Garage across different traffic times of the day. PM_2.5 concentrations during the PM non-rush hour period had the highest average concentration of 13.04 µg/m³ and the broadest range, with a maximum above 30 µg/m³. PM_2.5 concentrations at PM rush hour display a tight spread and a mean around 7 µg/m³, which could suggest improved ventilation or less vehicle congestion in the evening.

At the Washington Park Garage (Figure 5c), PM non-rush hours had the highest median concentration within the data of 15.78 µg/m³ and are tightly clustered. AM rush hour reported the second highest in concentrations with a mean of 11.51 µg/m³ and a maximum value of 25.36 µg/m³.

These two underground garages reported the highest emissions during the PM non-rush hour, which may be due to traffic leaving the garage ahead to avoid rush hour traffic. Different trends are shown in different garages, which warrants monitoring of each one individually. Contrary to anticipations, PM_2.5 concentrations during rush hours are not consistently higher than those during non-rush hours. This trend highlights the importance of continuous monitoring of garage emissions, rather than limiting the effort only to rush hours. PM₁₀ concentrations at each garage grouped by different traffic periods are shown in Figures S5–S7, which present similar trends to those of PM_2.5.

3.3. Contributing Factors of Garage Emissions

Different regression models were used to evaluate the correlation of PM_2.5 concentrations between garages and the ambient site (Taft), as shown in Figure 6. A positive slope indicates that as Taft PM_2.5 concentrations increase, garage PM_2.5 concentrations also increase. A strong linear correlation of R² = 0.6364 suggests that approximately 63.6% of the variation in garage PM_2.5 can be explained by changes in Taft PM_2.5. Additionally, a p-value of 0.04 (less than 0.05) concludes that there is statistical significance between the garage and Taft sites. The significant relationship implies that ambient air quality is likely influencing air quality in parking garages. The strong correlation with ambient values is unique to this study, which has not been reported in other university garage emissions [12,14,24]. The strong correlation with ambient air quality also suggests that control and mitigation practices can be implemented based on local air quality forecasts.

High correlations are also obtained from logarithmic and exponential regressions. However, some scatter and variability indicate that other factors, such as traffic and garage structure, may also affect pollution in garages. Similarly, a linear regression of PM₁₀ concentrations between garages and the ambient is also shown in Figure S8, with a slightly lower R² of 0.44.

Table S4 shows hourly averaged traffic volumes from the three garages. The summer days with the highest emissions did not have higher traffic, which is also an indication that PM emissions at these garages do not correlate with traffic. The median 1 h car count was 98 vehicles for Campus Green, 67 for CCM, and 54 for Washington Park. With the exception of Campus Green, the largest garage, this volume is close to another college garage study, with a median of 71 cars per hour on weekdays [12]. Table S4 also indicates that the traffic volume is not very different between rush hour and non-rush hours, which is likely due to the work-from-home practices post-pandemic.

The correlation between traffic volume and PM_2.5 was low, with an R² value of 0.09 for Campus Green, 0.21 for CCM, and 0.10 for Washington Park. This suggests that the correlation between PM_2.5 and hourly traffic volumes is not linear. One of the reasons may be the much lower average traffic volume compared with busy roads. This result is consistent with Gonzalez et al., 2022 [24]. However, the same study found that CO and NO were strongly associated with car volume.

4. Discussion on Potential Sustainability Improvement for Urban Garages

Urban garages can act as both localized hot spots for exposure to vehicle emissions and also increase the carbon footprint of the community, both of which are unsustainable.

In this study, particulate matter and CO₂ concentrations are elevated across all three parking garages compared to corresponding ambient concentrations, which suggests that these garages tend to accumulate vehicle emissions. These results reinforce the need for constant air quality monitoring. This study indicated that continuous measurements with portable instruments are feasible with meaningful trends. This can help the public understand air quality trends in the garages they frequently use and are exposed to. These sensors use laser or electrochemical technologies and can produce digital signals of the pollutant measured. Therefore, they can be easily integrated into various smart parking garages as the physical layer, similar to space detection sensors [10]. In addition, many of these sensors are also affordable. However, without the actual use of air quality sensors, the air quality benefits of smart parking systems can only be indirectly assumed.

Meanwhile, the urban garages in Cincinnati, OH, are also behind in smart parking technologies, which can impede the sustainability scores of the university and the city. The garages studied are equipped with very basic space management technology to show only “garage full”. There are gaps to fill for better garage management with advanced technologies for future consideration. Parking garages where drivers must search for unoccupied spaces are time-consuming and inefficient, especially in garages with many floors. Smart parking systems directly display information to drivers, such as spot availability status or the number of available spaces. Such measures will minimize time looking for spaces and provide real-time information. Real-time parking guidance systems reduce vehicle circling and idling time, while signage and efficient traffic paths minimize congestion and reduce stop-and-go emissions. Conventional ventilation systems are often inefficient and costly in terms of electricity consumption, which emphasizes the need for real-time air monitoring and automated control systems [11]. These systems also feature smart automation, which activates ventilators and alarms when pollutant levels exceed safety thresholds. This again calls for air monitoring data in garages.

Integration of air quality sensors to smart parking systems can help better protect people’s health, provide garage managers direct support for effective mitigation, and lead to more sustainable garages.

The role of parking garages should be upgraded from vehicle storage to include safety and security, and from car-centric to people-centric. We are confident that air sensors, together with smart parking technologies, can result in a more user-friendly parking experience.

5. Conclusions

Air pollutant emissions were measured in three urban garages in Cincinnati, OH, using portable instruments, each equipped with multiple sensors. PM_2.5 emissions in the garages were found to be consistently higher than ambient levels during all seasons, suggesting that garages can be local air quality hotspots for higher pollutant exposures. Each garage has its emission characteristics at different traffic periods of the day, which may be related to its structural configuration, traffic volume, and user types. These warrant continuous air pollutant monitoring. The Campus Green garage is the largest among the three and is semi-enclosed with multiple levels. It has the highest median traffic volume of 98 vehicles per hour and the highest annual average PM_2.5 concentrations of 12.55 µg/m³ among the three garages. PM_2.5 exposure levels when walking in the garages ranged from higher to comparable to gate levels. Seasonally, the highest average PM_2.5 concentrations were observed in the summer, which are 20.32 µg/m³ in Campus Green, 14.25 µg/m³ in CCM, and 15.23 µg/m³ in Washington Park garages, respectively. This seasonal increase may be due to higher ambient concentrations, evaporative emissions from vehicles, and stagnant air inside garages. The lowest concentrations were found in the spring in the two enclosed garages, while during winter it was in Campus Green. A strong linear correlation coefficient of 0.64 was found between the gate-level PM_2.5 and the ambient concentrations, but the correlation with traffic volume was not demonstrated. Although higher than ambient levels, CO₂ emissions were found to be largely within the 1000 ppm limit of ASHRAE in all three garages during all seasons. NO₂ emissions are also within the limit of NAAQS.

This study suggests that continuous air pollutant monitoring with portable instruments is feasible and especially helpful to better understand long-term trends. Real-time air quality monitoring can serve as guidelines for effective reduction in pollution and carbon footprint, especially in urban or high-demand parking garages. Integrating air sensors into smart garage management will enhance the sustainability of the garages, which will lead to better health protection and a better user experience.

6. Limitations

Low-cost devices, though very convenient in operation and quick to set up, have limitations and uncertainties that should be estimated and compared with standard instruments. The portable devices used in this study may be affected by temperature, relative humidity, and other environmental conditions since these devices do not have the sophisticated conditioning mechanisms of the FEM/FRM grade devices [27]. The South Coast Air Quality Management District (AQMD) in California tested both sensors and found that the Temtop M2000 has been regarded as fairly accurate, especially for its PM_2.5 mass concentrations. An R² of 0.999 in correlation to FEM GRIMM mass concentrations was observed in chamber calibration, and an R² of 0.77–0.82 was obtained for calibrations in the ambient environment. For PM₁₀, an R² of 0.18–0.28 was obtained, which is a partial reason our study mainly focused on PM_2.5 [28]. Temtop M2000 has been used as a reference sensor for PM [29,30] and CO₂ [31] measurements, which are indications of its data quality.

However, Flow-2 underestimated PM mass concentrations as measured by FEM GRIMM and FEM T640 and did not seem to track diurnal PM variations [27], and Flow-2 sensors overestimated the NO₂ concentrations. They did not seem to track diurnal variations as recorded by the FRM instrument. R² values representing the 1 h correlation between Flow-2 and the FEM instrument (GRIMM) were 0.11, 0.12, and <0.09 in average for PM₁, PM_2.5, and PM₁₀, respectively.

Although the Temtop M2000 has been regarded as fairly accurate [28], in this study, we did not do side-by-side calibration with FEM devices, which is a limitation. Studies have indicated that vehicle emissions are among the major sources of ultrafine particles (UFP, particles less than 0.1 µm), which can penetrate deeply into the respiratory system [32]. Another study showed that peak ultrafine particle concentrations were observed during traffic volume surges [33]. However, there were only two UFP measurements in this study. One was conducted at the CCM garage on 28 July 2023 during AM peak time, and the P-Trak (TSI) was carried by the tester while walking in the garage together with other devices. The UFP count ranged from 7068 to 20,178 particles/cm^3, and the mean was around 10,176 particles/cm³. This is about 1.37 times the ambient level, which had a mean of 7436 particles/cm³. This observation is consistent with that of Zhao et al., 2018 [34]. Another CNC measurement in the CCM garage was taken on 13 September 2023 during PM rush hour, where the P-Trak was placed at the garage entrance together with other devices and obtained a much lower mean count of 5345 particles/cm³, which is closer to ambient values. The preliminary results warrant further UFP measurements, especially inside garages. This is another limitation of this study.