# Improving Low-Cost Optical PM Sensor Accuracy in Humid Conditions via Aerosol Liquid Water Estimation Using U.S. EPA CSN Data

## TL;DR

This paper improves the accuracy of low-cost PM2.5 sensors in humid conditions by estimating aerosol liquid water using EPA data.

## Contribution

A novel framework for correcting PM2.5 sensor measurements in high humidity using aerosol liquid water estimation and optical calibration.

## Key findings

- Corrected PM2.5 data align well with EPA reference measurements.
- Aerosol liquid water quantification aids aqueous-phase aerosol chemistry studies.
- Practical guidance is provided for regions without EPA CSN data.

## Abstract

High ambient relative humidity (RH) poses a substantial
challenge
to the accuracy of low-cost optical sensors used for measuring the
fine particulate matter (PM2.5) concentration. In this
study, we developed a novel, practical, and feasible framework for
mechanistically correcting low-cost PM2.5 sensor measurements
under high-humidity conditions by quantitatively separating aerosol
liquid water mass (ALW) using the widely available EPA Chemical Speciation
Network (CSN) data set, after accounting for the necessary optical
calibration procedures that affect sensor performance at elevated
RH. We introduced two key correction processes for a low-cost optical
PM2.5 measurement system comprising a nephelometer and
an optical particle counter: (1) optical calibration grounded in Mie
theory to account for variations in sensor performance driven by aerosol
size distribution, refractive index, and hygroscopic growth, and (2)
determination of ALW to estimate dry-equivalent PM2.5 mass
concentrations under high RH conditions. The corrected PM2.5 data exhibit strong agreement with EPA reference measurements, affirming
the robustness of the proposed correction framework. Furthermore,
the quantification of ALW offers valuable insights for advancing aqueous-phase
aerosol chemistry and secondary aerosol formation studies. For regions
without colocated CSN data, we provide practical guidance for applying
these correction methods using surrogate information. Overall, the
methodologies developed in this work are expected to significantly
enhance the accuracy and applicability of low-cost optical PM2.5 sensors in humid environments.