Reducing Uncertainty in Aerosol-Cloud Interactions using Remote Sensing: Theory, Applications, and Climate Education

Abstract

Understanding the vertical distribution of cloud condensation nuclei (CCN) concentrations (NCCN) is crucial for reducing uncertainty in aerosol-cloud interactions (ACI) and effective radiative forcing, and yet has been an aspect missing from many ACI focused studies. Many proxies, parameterizations, and relationships between NCCN and aerosol optical properties (AOPs) have been developed and tested using widely available remote sensing observations. In Lenhardt et al. (2023), we developed a simple linear regression method to approximate vertically resolved NCCN from High Spectral Resolution Lidar 2 (HSRL-2) backscatter (BSC) and extinction profiles (EXT). This method was reliable for observations of smoke at low (≤ 50%) relative humidities (RH) from the NASA ORACLES (ObseRvations of Aerosols above CLouds and their intEractionS) campaign over the Southeast Atlantic (SEA), but is less reliable when applied to observations of mixed aerosol types in other regions. Several uncertainties and limitations exist when developing methods to relate NCCN and AOPs, generally related to high RH, environments with internal or external mixtures of several aerosol types, and differences in which parts of the aerosol size distribution are most relevant to NCCN compared to AOPs. To understand the dominant governing factors of NCCN—AOP relationships, and for which cases a simple linear approximation is and is not applicable, we use in situ observations of the aerosol size distribution and chemical composition to inform theoretical calculations of NCCN and aerosol BSC at 532 nm. These observations come from the recent NASA ACTIVATE (Aerosol Cloud meTeorology Interactions oVer the western ATlantic Experiment) campaign in the Northwest Atlantic, where there is a diverse mixture of aerosols located primarily in the marine boundary layer (MBL). Estimates from random forest models indicate that, for smoke, marine, and urban aerosols, the aerosol size distribution, as parameterized using the effective radius (Reff), is the most important predictor of the theoretical NCCN—BSC relationship. We further investigate the physical impact of Reff on the NCCN:BSC ratio and find an exponential relationship between the parameters. We find that modeling NCCN:BSC using this exponential Reff relationship can explain about 68%–79% of the variance in the theoretical NCCN—BSC relation ship. These findings suggest that including information about aerosol size is critical for future studies in constraining NCCN from AOPs. While assessing the theoretical NCCN—BSC relationship and its non-linear, multivariate dependencies, (Redemann and Gao, 2024) developed a machine learning (ML) method to predict NCCN from HSRL-2 observables and reanalysis data. Due to the high accuracy of this method to predict NCCN within about 14% uncertainty, we use this reliable and vertically resolved product (ML-CCN) to investigate ACI in the SEA. Evidence of interactions between the biomass burning smoke plume and underlying marine stratocumulus cloud deck has been found using in situ data. Motivated by these findings and equipped with this new remote sensing-based NCCN dataset, we further investigate these cloud top ACI and their dependence on lower tropospheric stability (LTS). Taking advantage of cloud edge HSRL-2 profiles, we assess the simultaneous impact of above- and below-cloud NCCN on cloud top microphysical properties. We observe a decrease in cloud droplet effective radius (Reff) and an increase in cloud droplet number concentration (Nd), both of which are associated with increasing above-cloud NCCN. These relationships are strongly dependent on LTS, with ACIREFF decreasing from 0.161 to 0.042 (-73.9%) and ACICDNC decreasing from 0.452 to 0.116 (-74.3%) as LTS increases from 10 to 22 K. Additionally, we find that above-cloud NCCN—cloud property relationships are similar for cloud edge and cloud center observations. The relationship between below-cloud NCCN and cloud top properties is strongly dependent on LTS, with ACI metrics increasing as LTS increases. This speaks to the dominance of above-cloud smoke entrainment as a modulator of stratocumulus cloud properties under unstable conditions, while below-cloud NCCN nucleation dominates in stable environments. These findings demonstrate the importance of using vertically resolved NCCN in ACI studies and establish a remote sensing-based analysis method with which future satellite studies can investigate ACI. Improving the representation of NCCN and ACI in climate models will help to better predict future climate warming in different emissions scenarios. However, climate warming cannot be viewed exclusively as a future issue, as anthropogenic emissions have already begun to impact Earth’s weather and climate. Therefore, it is also important to take action now to safeguard the future of our environment. One such route to do so is by improving climate change communication (CCC) and educating students on climate science. As a practical step towards this goal, a three-day aerosol and cloud-focused set of modules was developed and implemented in four sections of the meteorology elective at Norman High School (NHS). Responses to pre- and post surveys, as well as three reflection question worksheets, were collected to evaluate the effectiveness of the different lectures and activities. For 22 of the 24 survey statements, there was an overall shift towards more correct responses in the post-survey, and responses to the reflection questions provided context and possible explanations for why certain questions saw only a very small improvement or a shift towards decreased understanding. Student responses show the benefit of communicating complicated topics using hands-on activities and the need to balance how many topics are included in individual activities. Additionally, responses emphasized the importance of insight from the classroom teacher in terms of students’ background knowledge and what words and phrasing may confusing. The findings also demonstrated the confirmation of the primary hypothesis, which was that by focusing on teaching the mechanisms of aerosol direct and indirect effects, students would be able to better deduce their possible long term climate impacts. Several studies that took place alongside the projects described here can help place and interpret our work in a broader scientific context. We discuss the benefits of using ML (Redemann and Gao, 2024) and physics-based approaches in tandem instead of in isolation, as well as a project that evaluated more details of NCCN closure from the ACTIVATE observations than we investigated in Chapter 2 (Soloff et al., 2025). We discuss studies by Witthuhn et al. (2025) and Chang et al. (2025) that emphasize the importance of capturing spatial and temporal variations in aerosol and cloud properties for constraining global horizontal irradiance and the direct aerosol radiative effect, respectively. Lastly, we discuss recent work by Gao et al. (2026) that applies a very similar method as shown in Chapter 3 to observations of a cumulus cloud field. This study found more realistic ACI relationships using ML-predicted NCCN when compared to AOPs, which again supports the importance of and need for vertically resolved NCCN in ACI studies. The work detailed here has several implications for future remote sensing of NCCN and ACI from both aircraft and satellite platforms, as well as suggestions for outreach efforts and climate science education. Additionally, the work highlights many routes for potential future analysis that were outside the scope of these projects. We conclude by discussing several of these possible directions, such as investigating NCCN—AOP relationships for dust and examining similar theoretical relationships for aerosol mixtures in different types of environments. The ACI study could be expanded to new cloud and environmental regimes, and more meteorological variables could be included to better understand and explain the observed relationships. There are several ways in which the climate education modules could be improved and iterated upon, including implementing them in more schools and classrooms, refining some wording issues in the surveys and worksheets, and developing new hands-on activities for topics such as cloud albedo and precipitation susceptibility.