This document serves as an overview to Clarity's approach for ensuring the accuracy of air quality sensor data. It outlines the importance of calibration and provides an overview of the calibration process used by Clarity.
Importance of Calibration
Calibration plays a crucial role in obtaining accurate measurements of air pollutant concentrations. It ensures that the data collected by the air quality sensors is reliable and can be used for informed decision-making. Calibration helps to minimize measurement errors and ensures that the sensor readings are aligned with reference instruments.
Furthermore, Calibration performance metrics can be used to estimate measurement performance and measurement uncertainty, to inform data analysis throughout the monitoring project.
Why PM2.5 Calibration
Clarity utilizes low-cost optical particle counter (OPC) sensors to measure PM2.5 mass concentration. However, these sensors do not directly measure PM2.5 mass concentration but count and size particles instead. Calibration is necessary to estimate PM2.5 mass concentration accurately by accounting for the composition of particulate matter at the project site.
Why NO₂ Calibration
Clarity uses electrochemical cell (ECS) sensors to measure NO₂ concentrations. These sensors can be influenced by various factors, such as temperature and humidity changes, leading to baseline shifts and reduced accuracy. Calibration is essential to correct for these interferences and ensure that the raw NO₂ signal from the ECS sensors aligns with the output of reference instruments.
The Calibration Process
Clarity's approach for obtaining reliable data from low-cost sensors diverges from considering these sensors as standalone products. Instead, we approach air quality monitoring using low-cost sensors as comprehensive projects. Through Sensing-as-a-Service, Clarity provides customer organizations with more than just hardware and software; we guide them through the entire project setup. This approach encompasses crucial stages to obtain data accuracy, including collocation-based calibration, described below.
Preset Global Calibrations
Clarity nodes come equipped with preset calibrations: the PM2.5 and NO2 Global calibrations.
These calibrations are the result of rigorous development, involving collocating of Clarity devices with reference monitors at numerous global sites, covering a broad spectrum of pollution and environmental conditions: 6,000,000+ collocated measurements from 450+ sensors deployed in 45+ different cities.
The preset PM2.5 calibration minimizes errors related to OPC sensors estimates of aerosol composition, while the preset NO2 calibration profile uses advanced statistical methods to counteract temperature and relative humidity fluctuations affecting ECS sensors baseline.
These calibrations significantly boost accuracy compared to raw sensor data - which is always available alongside calibrated measurements. However, Clarity recommends organizations take an extra step by performing collocation-based calibration as described below. This further improves accuracy and allows measurement performance quantification for all nodes.
For organizations opting not to conduct collocation-based calibration, or lacking access to a reference monitor, we communicate the limitations of calibration without collocation. If the organization later decides to carry out collocation-based calibration, or a reference monitor becomes accessible, retroactive calibration can be performed.
Collocation
Collocation is a critical step in the collocation-based calibration process. It involves installing Clarity Nodes next to reference monitors to ensure exposure to the same pollutant concentrations. The purpose of collocation is:
- To collect data that will be used to develop custom calibrations for PM2.5 and NO₂ measurements: To do this, all Clarity Nodes are deployed next to a reference monitor in the project region for at least one month, prior to the field deployment. The collocation data collected needs to satisfy requirements related to pollutant concentration range and temperature and relative humidity range.
- To monitor long-term performance of the calibrated Clarity Nodes: To do this, Clarity advises customer organizations to keep at least one Clarity Node permanently collocated with the reference monitor for the entire duration of the monitoring project.
Table: collocation requirements
Custom Collocation-Based Calibrations
Collocation data plays a crucial role in assessing the effectiveness of both PM2.5 and NO2 preset calibrations. If the preset calibrations meet predefined performance standards, they are validated for use. However, if their performance falls short of expectations, the collected collocation data is leveraged to create custom calibrations. For PM2.5, a regional calibration may be developed based on the collected data, while for NO2, individual sensor-specific custom calibrations are generated.
Importantly, during the calibration process, each calibrated node's measurement performance is quantified. This is accomplished using a subset of the collocation dataset specifically reserved for evaluation. The nodes performance metrics provide further assurance of data accuracy and reliability, and are useful to quantify uncertainty during air quality data analysis.
Typical performance
Setting performance goals for low-cost air quality sensors is a continuous effort involving governments and academics. This comes with challenges. Different use cases require different sensor performance. While calibration helps, there's a balance between enhancing accuracy and managing operational costs. Sensors also work differently under different environmental conditions. Comparing sensor performance is challenging due to differences in testing protocols and scenarios. Clarity tackles this by implementing a performance metric-based decision framework, leveraging our understanding of typical performance on the field. This ensures reliable data for real situations.
Image: performance metric-based decision framework
Calibration Process Communication
Throughout the calibration process, clear communication is maintained with the customer organization. The calibration results, including the selected calibrations, are presented to the organization for approval. Any challenges or issues encountered during calibration are addressed and discussed with the organization. Clarity Nodes performance metrics are highlighted and made available to the customer organization, to be used as estimates of measurement performance and uncertainty for data analysis throughout the monitoring project.
Long-term performance monitoring
The Clarity Nodes performance metrics calculated during calibration are expected to hold true as long as the sensors are operated within the same environmental and pollutant composition conditions that were experienced during calibration.
In order to better monitor the performance of the Clarity Nodes long-term, Clarity advises customer organizations to keep at least one Clarity Node permanently collocated with the reference monitor for the entire duration of the monitoring project. The long-term collocation data is made available to customer organizations through the Clarity Air Monitoring API and specific collocation-focused sections of the web-based Clarity Dashboard. If the nodes performance metrics drop below the calibration levels, the customer organization can open a support ticket with Clarity. Clarity will review long-term collocation data and advise on either a recalibration or device substitutions.
Conclusion
Calibration is a critical aspect of ensuring accurate air quality sensor data. Through the process described, Clarity strives to minimize measurement errors and provide reliable air quality information. By following this data quality assurance process, Clarity aims to deliver high-quality and accurate air quality data to its customer organizations.
Useful resources
Clarity Learning Center: Device Collocation and Calibration for Improved Data Quality
Clarity Learning Center: Understanding Device Calibration and Sensor Performance
Resources: Guide to Low-Cost Sensor Networks
Resources: Air Quality Measurement Series
Resources: A Guide to Accurate Particulate Matter Measurements with Air Sensors