Air Sensor Network Analysis Tool: R-Shiny Application

1. Introduction

Poor air quality is associated with a variety of negative health effects [1]. Air quality data are needed to understand local conditions and reduce exposure to air pollution [2]. In the U.S., air quality is measured by ambient air monitors operated by state, local, and Tribal air agencies [3]. Recent efforts supplement the national monitoring network with localized air sensor data to investigate variations in air quality at neighborhood scales. Air sensors typically cost one or two orders of magnitude less than conventional air monitors (e.g., $10-$10k) and are designed to be compact enabling additional measurements at more locations. However, air sensor data may be noisy [4,5], biased, or inaccurate [6]. Furthermore, sensor performance may vary over time [7], concentration range (e.g., nonlinear response), or the environment in which they operate (e.g., high relative humidity, RH) [8,9]. Limitations may be pollutant, sensor technology (e.g., optical, electrochemical, metal oxide), or manufacturer specific and quality assurance (QA) and quality control (QC) steps must be tailored for the specific sensor needs [10]. Sensors often must be compared with nearby measurements and corrected to improve the data quality [11,12,13,14]. The air sensor data analysis required to account for these limitations is challenging for users without extensive coding experience and for agencies faced with increased data volumes and community questions and with steady or decreasing staff time.

Many existing air data analysis software tools are costly; require coding experience to be run in an open-source environment [15,16,17]; are specific to one manufacturer’s sensor [16,18]; or are designed to focus on single collocation sites [19]. An analysis tool is needed to more easily aggregate data from multiple air quality data sources, perform standard QC, identify nearby sensors spatially for comparison, and create comprehensive visualizations. Generating air quality analysis in this form will further the capabilities of air sensors and potentially aid in protecting public health and the environment.

2. Materials and Methods

The Air Sensor Network Analysis Tool (ASNAT) is an R-Shiny [20,21] application that integrates data from multiple air quality networks (e.g., national monitoring network, sensor networks, meteorological stations) to provide a broader representation of local air quality. The tool was developed mostly in R with some code in C++ to enhance performance speed. It was designed to support air quality professionals with a base knowledge of air quality analysis to develop sensor data corrections, apply data flagging, understand the performance of new air sensor networks, and improve sensor data quality so that the comparability between networks is improved and users can better understand local air quality - including during extreme events like wildfires, dust storms, and fireworks.

Users can select data from EPA’s Remote Sensing Information Gateway (RSIG) (under “Load Web” in Figure 1) which allows users to efficiently call data based on a geographic bounding box. Users can retrieve AirNow, EPA’s Air Quality System (AQS), Meteorological Aerodrome Report (METAR), and public crowdsourced PurpleAir data [22] (https://www.epa.gov/hesc/web-access-rsig-data, last accessed 2/28/25) ). Data from AirNow and AQS available in ASNAT include fine particulate matter (PM_2.5), particulate matter 10 µm or less in diameter (PM₁₀), RH, temperature, pressure, ozone (O₃), nitrogen dioxide (NO₂), carbon monoxide (CO), and sulfur dioxide (SO₂). While the AQS and AirNow datasets are similar, the AQS dataset has gone through further QA and QC and typically it will not be available for a few months after it has been collected. In some cases, more data may be included in the AirNow dataset as not all monitors also report to AQS. Depending on the application, AirNow or AQS data may be more appropriate, and users can easily run analysis with one and then the other. The PurpleAir dataset includes the corrected PurpleAir PM_2.5 data (PurpleAir.pm25_corrected) that is corrected in a similar way as PurpleAir data on the EPA and US Forest Service AirNow Fire and Smoke Map (fire.airnow.gov, accessed: 1/28/25) including completeness criteria (70% for daily or hourly averages), exclusion when A and B channel measurements disagree, and application of the US-wide correction that accounts for nonlinearity at high smoke concentrations [23]. Users must supply a valid PurpleAir API read key (available directly from PurpleAir) to access this data. Users can also load data from standard format text files, which can be generated using the Air Sensor Data Unifier tool (https://www.epa.gov/air-sensor-toolbox/air-sensor-data-tools, last accessed 2/28/25), or from offline PurpleAir files. Data can be loaded at hourly or 24-hour averages and NowCast averaging can also be applied to visualizations and statistics (https://usepa.servicenowservices.com/airnow/en/how-is-the-nowcast-algorithm-used-to-report-current-air-quality?id=kb_article&sys_id=798ba26c1b1a5ed079ab0f67624bcb6d, last accessed 20 June, 2025).

Throughout this paper we will use a 2-week case study to demonstrate the features and functionality of this tool. During June 2023, the midwestern U.S. was impacted by smoke from Canadian wildfires [24]. During uncommon events like these, state and local agencies may be interested in better understanding how sensors perform and whether any improved QA QC is needed to provide accurate data to the public. Further, the synthesis of air quality data from multiple observational networks supports current and future agency response (e.g., publicizing current air quality conditions and suggested response, documenting evidence of exceptional events).

3. Results

ASNAT provides a variety of ways for users to view and explore local air quality. Once users select and load data, monitoring sites appear on the map at the top of the map tab (Figure 1). Each data type is assigned a symbol, and users select a colormaps that is then described in the legend. Users can choose how to view the data (hourly averages, nowcast averages, daily averages, mean value over all time steps) and use the timestep slider below the map to animate the information, if applicable (Figure 1).

Currently ASNAT has six tabs that provide different ways to interact with the data. These include the map, flagging, tables, plots, corrections, and network summary tabs. After selecting data, users can decide to either move to the flagging tab or to the plots tab. In many cases the user may need to generate the tables and plots before deciding what kind of flagging and removal of problematic data is needed.

If the user has selected only an X variable, the tables tab summarizes the data by site ID including the count (i.e., number of hours or number of days available), percent of missing data, mean, minimum, 25^th percentile, median, 75^th percentile, and maximum. If the users have selected X and Y variables within a certain distance it provides the same summary table for each X and Y variables. In addition, it provides a table of the neighboring points at each time stamp, the coefficient of determination (R²) and distances between each set of points, and the number of points in each air quality index (AQI) category and the normalized mean bias error (NMBE), root mean squared error (RMSE), and normalized root mean squared error (NRMSE) in each AQI category. If the user generates these tables and then uses the save data button, these summary tables will be saved along with the raw data.

The plots and network summary tabs include additional visualizations to understand network performance and local air quality. Figure 2 presents a summary of air quality variation from the network summary tab. During our two-week case study, AQS monitors (red color) report slightly higher concentrations than the PurpleAir sensors (blue color) and show somewhat different time of day patterns. Because some differences were found, users can go on to compare nearby sensor and monitor pairs, instead of all sensors and monitors within the area, to understand whether this is due to sensor bias or because the sensors and monitors are exposed to different pollutant concentrations.

The plots tab allows users to compare two datasets by finding nearest neighbors within a user specified distance. Typically, this involves comparing a sensor dataset with more uncertainty and unknown performance to nearby reference monitor data (e.g., AirNow, AQS) to better understand sensor performance and develop corrections. ASNAT describes the sensor-monitor relationship through a series of performance metric calculated as outlined in EPA’s performance targets [25,26,27,28] (Figure 3, Figure 4).

In this case study example, sensor-monitor pairs within 250 meters were selected. Most sensor-monitor pairs are strongly correlated, meeting EPA’s performance target for coefficient of determination (R²) with only a few pairs falling below the 0.7 target. All sensor-monitor pairs meet the targets for slope (1 ± 0.35), intercept (5 ≤ b ≤ 5 mg/m³), and root mean squared error (RMSE ≤ 7 mg/m³). Some sensors have normalized root mean squared error outside of the target (>30%) however, EPA’s performance targets only require meeting either the RMSE or the NRMSE target.

Using the corrections tab, sensor corrections can also be developed to adjust for bias between the sensor and the monitor (or whatever two comparison variables the user selects). Sometimes, a simple linear correction allows two datasets to become more comparable. However, much past work with air sensors has shown that RH or temperature can influence sensor measurements and that many gas sensors may have cross sensitivities to other gases [9] so a third variable may be needed to account for some of these influences or interferences. In ASNAT, currently available corrections include single variable, multivariable additive, or multivariable interactive in either linear, quadradic, or cubic forms.

Figure 4a shows a sensor with a weak correlation (R² = 0.57 < 0.7). There are a variety of reasons why this sensor may have weaker performance than the typical sensor during this period (Figure 3). For instance, this sensor may be located closer to a source leading to the higher PurpleAir concentrations reported when the nearby reference monitor (AQS) measurements are low, the user may have improperly reported the latitude and longitude of the sensor meaning the sensor may be more than 250 m from the AQS monitor, differences in federal equivalent method (FEM) may lead to differing performance, this sensor may have hardware or software differences leading to different performance, or there may be other issues. Figure 4b shows a different sensor has strong agreement and low normalized mean bias error (NMBE) when compared to the nearby reference monitor. Depending on the project objectives, this performance may be adequate. If higher accuracy is needed, the individual sensor correction generated (y = -1.4597 + 1.1643x) could be applied.

Performance of the sensor can also be explored by considering how frequently the sensors and monitors report the same AQI category (Figure 5) also using the plots tab. This example shows strong agreement with hourly averages agreeing at least 76% of the time across categories. NowCast and 24-hour averages are more frequently used for AQI display but it is typically more challenging for sensors and monitors to agree at shorter averaging intervals (e.g., hourly).

The flagging tab provides a variety of options to flag and/or remove anomalous data based on several categories such as: exceedance of a threshold, agreement with a nearest neighbor, repeated values, outlier detections (using several methods), a user specified period (e.g., a day with a known sensor issue), and more (Figure 6). More than one flag can be applied. Performance plots on the plots tab (i.e., overall scatterplot, AQI agreement plot, boxplots of performance targets) are shown with all data and only the unflagged data and flagged points are indicated in lighter shades on the timeseries. When data is downloaded the flags will be removed. In this case study example, there are no clear outliers to remove.

4. Discussion

The Air Sensor Network Analysis Tool (ASNAT) is an R-Shiny [20,21] application designed to ease the burden of combining, comparing, and summarizing discrete air monitoring networks operating in the same time and place making it easier for air monitoring agencies to understand local air quality. Limitations of this tool include data size limits and the need for users to have some background air quality knowledge to appropriately use and interpret the available tools. Functionality is limited to the most requested analyses. Dataset size limits are dependent on local computing resources. Users may need to limit longer time analysis to smaller spatial ranges and/or longer averaging intervals. Some background air quality knowledge is needed to successfully use this tool and make informed decisions based on ASNAT outputs. For example, nearest neighbor radius will depend on pollutant chemistry, local sources, and geography. Local knowledge may be required to determine if outliers are due to sensor malfunctions or real short-term pollutant events. In addition, more complicated corrections (e.g., machine learning, additional variables) are sometimes required for adequate sensor performance but these more complicated corrections are not currently possible in ASNAT. These more complex corrections can increase the risk of over fitting (i.e., generating a correction model that will not perform adequately outside of the calibration period) [29,30], are not easy to interpret [29], and require longer calibration periods [31]. For these reasons we have not included more complicated corrections currently to hopefully prevent users from generating unhelpful corrections. There will always be additional analysis, quality assurance, or plots that could be helpful to understand local air quality issues, however, this tool gives users without data analysis expertise and those who do a quick way to perform some helpful analysis.

This project is ongoing, and we hope to add additional features and improvements based on the feedback from the initial users. In addition, since the code is open-source and publicly available, we hope that users will take it and modify it as needed for their own uses adding additional functionality and customized displays. So far more than two hundred air quality professionals have been trained to use this tool including staff from state, local, and tribal agencies, EPA, other federal agencies, academia, consulting companies, and other organizations. The high engagement in tool training and office hours highlight the strong need for this type of tool.

This tool contributes to increasing the utility of air sensor data. For example, state agency staff plan to use the tool to help support a community monitoring project using PurpleAir sensors and to perform regression analysis on multiple PurpleAir sensors collocated with an air monitor. Also, they plan to use the tool to use this tool to understand how custom-built sensors compare to nearby PurpleAir sensors. Local agency staff plan to use the tool to understand how local sources (e.g., incinerators) impact air quality at schools. By reducing the time it takes to complete analysis, resources are saved that can be put into furthering these projects or other agency priorities. In addition, this data can be used to take informed actions to better protect public health. This may include providing more accurate data to the public or better understanding local sources.