Leak Affects Cloud Condensation Nuclei Data From 2 ARM Campaigns

Published: 2 June 2025

A sample line leak affected the usability of cloud condensation nuclei (CCN) counter data from the 2021–2022 TRacking Aerosol Convection interactions ExpeRiment (TRACER) and the 2023–2024 Eastern Pacific Cloud Aerosol Precipitation Experiment (EPCAPE).

The Atmospheric Radiation Measurement (ARM) User Facility deployed the same Aerosol Observing System (AOS) but different CCN counters for TRACER in La Porte, Texas, and EPCAPE in La Jolla, California. The issue appeared in CCN number concentration (N_CCN) data from both field campaigns.

Users should not use CCN counter data from either campaign. The affected datastreams are:

• EPCAPE: epcaosccn200M1.a1, epcaosccn2colaavgM1.b1, epcaosccn2colaM1.b1, epcaosccn2colaspectraM1.b1, epcaosccn2colbM1.b1, and epcaosccnsmpskappaM1.c1.
• TRACER: houaosccn200M1.a1, houaosccn2colaavgM1.b1, houaosccn2colaM1.b1, houaosccn2colaspectraM1.b1, houaosccn2colbM1.b1, and houaosccnsmpskappaM1.c1.

Uncovering and Untangling the Issue

Maria Zawadowicz, a scientist at Brookhaven National Laboratory in New York, raised some concerns about the CCN counter data while working on a Scientific Focus Area project supported by the U.S. Department of Energy’s Atmospheric System Research (ASR) program area. Her unique position as both an ASR-supported scientist and ARM’s lead instrument mentor for the aerosol chemical speciation monitor (ACSM)—one of the AOS instruments overseen by Brookhaven—was instrumental in bringing attention to the problem. This dual role exemplifies the strong synergy between ASR and ARM and how it can lead to important advancements in data quality and instrument performance.

For the ASR project, Zawadowicz compared N_CCN measured by the CCN counter with N_CCN predicted from ACSM and scanning mobility particle sizer (SMPS) measurements at TRACER, EPCAPE, and ARM’s Southern Great Plains (SGP) atmospheric observatory in Oklahoma.

Zawadowicz found that the measured and predicted N_CCN were comparable for the SGP, but the measured N_CCN was substantially lower (35%) than the predicted N_CCN for TRACER and EPCAPE. These differences between the measured and predicted N_CCN data for both field campaigns made ARM AOS staff realize there was an issue.

The root cause was traced to a leak in the CCN sample line due to a faulty pressure, temperature, and relative humidity sensor. These sensors, which were installed and used for the first time at TRACER, are placed close to the instrument in each sample line to characterize the sample conditions. The leak allowed cabin air to enter, diluting the aerosol sample and rendering the measurements inaccurate.

This issue was uncovered during onsite troubleshooting at the end of the EPCAPE campaign by Janek Uin, the lead mentor for the CCN counter at that time.

Because the chemical composition of the particles in the cabin air at the time was unknown, there is no straightforward way to correct the CCN data for the leak. There were no effects on other instruments.

Working to provide an alternative to the CCN data, Zawadowicz is calculating an SMPS estimate that could be made available to ARM users as a principal investigator product.

During TRACER’s summer 2022 intensive operational period, a Texas A&M University team operated a CCN counter as part of a mobile laboratory van that collected measurements in the greater Houston area. The CCN counter data from the Texas A&M TRACER campaign are available in the ARM Data Center.

AOS Platform Complexity and Response

The sample line leak highlights the need for more frequent checks of the AOS infrastructure and more complex cross-instrument comparisons to find issues sooner, especially as the AOS platform grows more complex. Each AOS container integrates 12 to 15 instruments, along with a central inlet, distribution manifold, and numerous sample lines, all of which must work in harmony to ensure accurate data collection.

It is not just the performance of individual instruments that matters; it is also about delivering the same aerosol sample to each instrument. The inclusion of pressure, temperature, and relative humidity sensors in the sample lines, based on community requests, added valuable metadata but also introduced additional potential failure points.

To prevent future issues, the ARM AOS team established a protocol for checking sample line integrity at key stages: before deployment; during field operations; and before and after making significant changes in the line, such as changing a tube or adding/removing any component. The goal is to identify and correct any leaks early.

Each check involves removing the sample line from the aerosol stack and placing a HEPA filter on the stack end. All connected instruments should then read zero; any deviation triggers further investigation. Every connection point along the sample line is considered a potential single point of failure and is checked accordingly.

The AOS team first implemented the sample line integrity checks at the Bankhead National Forest observatory in Alabama and the Coast-Urban-Rural Atmospheric Gradient Experiment (CoURAGE) in Maryland in October 2024, then at the SGP in January 2025. In the summer of 2025, the team plans to implement the protocol during its annual calibration activities at the Eastern North Atlantic (ENA) observatory in the Azores.

In April 2025, ARM published updated AOS instrument handbooks for the SGP, ENA, and each of its three mobile facilities, including the one that ran at TRACER and EPCAPE, with a new section about the sample line integrity checks.

Questions? Feedback?

If you have questions or comments about the data or instruments in this article, please contact the associated mentors through the instrument mentor page: AOS lead mentor Olga Mayol-Bracero, CCN counter mentors Ogochukwu Enekwizu and Janek Uin, ACSM lead mentor Maria Zawadowicz, or SMPS lead mentor Ashish Singh.