Comment on acp-2021-999

Hailing Jia, Johannes Quaas, Edward Gryspeerdt, Christoph Böhm, Odran SourdevalOpenalex Unknown 2022 · DOI · OpenAlex

Abstract

Aerosol–cloud interaction is the most uncertain component of the overall anthropogenic forcing of the climate, in which the Twomey effect plays a fundamental role. Satellite-based estimates of the Twomey effect are especially challenging, mainly due to the difficulty in disentangling aerosol effects on cloud droplet number concentration ($N_d$) from possible confounders. By combining multiple satellite observations and reanalysis, this study investigates the impacts of a) updraft, b) precipitation, c) retrieval errors, as well as (d) vertical co-location between aerosol and cloud, on the assessment of $N_d$-to-aerosol sensitivity ($S$) in the context of marine warm (liquid) clouds. Our analysis suggests that $S$ increases remarkably with both cloud base height and cloud geometric thickness (proxies for vertical velocity at cloud base), consistent with stronger aerosol-cloud interactions at larger updraft velocity. In turn, introducing the confounding effect of aerosol–precipitation interaction can artificially amplify $S$ by an estimated 21%, highlighting the necessity of removing precipitating clouds from analyses on the Twomey effect. It is noted that the retrieval biases in aerosol and cloud appear to underestimate $S$, in which cloud fraction acts as a key modulator, making it practically difficult to balance the accuracies of aerosol–cloud retrievals at aggregate scales (e.g., 1° × 1° grid). Moreover, we show that using column-integrated sulfate mass concentration ($SO_4C$) to approximate sulfate concentration at cloud base ($SO_4B$) can result in a degradation of correlation with $N_d$, along with a nearly twofold enhancement of $S$, mostly attributed to the inability of $SO_4C$ to capture the full spatio-temporal variability of $SO_4B$. These findings point to several potential ways forward to account for the major influential factors practically by means of satellite observations and reanalysis, aiming at an optimal observational estimate of global radiative forcing due to the Twomey effect.

Introduction

The "Twomey effect" describes how an increase in anthropogenic aerosols leads to a higher concentration of smaller cloud droplets ($N_d$), increasing cloud albedo and exerting a cooling effect on the planet. Quantifying this effect is a cornerstone of climate modeling, yet satellite-based estimates of the sensitivity ($S$) of $N_d$ to aerosol changes remain plagued by uncertainty. The primary challenge lies in isolating the causal impact of aerosols from environmental confounders such as updraft velocity, precipitation, and retrieval errors. This study aims to disentangle these factors to provide a more robust observational estimate of the Twomey effect.

Methodology

The research utilizes a multi-sensor approach, integrating data from the A-Train satellite constellation (Aqua and CloudSat) and the Terra satellite.

  • Updraft Proxies: In the absence of direct vertical velocity measurements, Cloud Base Height (CBH) from MISR and Cloud Geometric Thickness (CGT) were used as proxies for updraft strength.
  • Precipitation Filtering: CloudSat radar data (2B-CLDCLASS) allowed for the separation of non-precipitating clouds from those undergoing droplet depletion via rain.
  • Vertical Co-location: To improve upon standard column-integrated metrics, MERRA-2 reanalysis was used to estimate sulfate mass concentrations specifically at the cloud base ($SO_4B$).
  • Statistical Analysis: Sensitivity ($S$) was determined through linear regression of $\ln N_d$ against $\ln (\text{Aerosol Proxy})$, employing binning methods to ensure statistical robustness.

Key Findings

The Role of Updrafts and Dynamics

Sensitivity ($S$) was found to increase significantly with cloud base height and thickness. Stronger updrafts generate higher supersaturation, which activates a larger fraction of aerosol particles into droplets. This confirms that failing to account for dynamical constraints leads to a fundamental misinterpretation of aerosol-cloud interactions.

The Impact of Precipitation

Including precipitating clouds in the analysis artificially inflates the calculated sensitivity by approximately 21%. Precipitation acts as a sink for droplets; at low aerosol concentrations, clouds are more prone to raining, which further depletes droplets and creates a misleadingly steep regression slope. Removing these "raining" cases reduced the diagnosed sensitivity ($S_{AI}$) from 0.68 to 0.56.

Vertical Co-location and Column Bias

Most previous studies rely on column-integrated aerosol data (e.g., Aerosol Optical Depth). This study demonstrates that column-integrated sulfate ($SO_4C$) fails to capture local variability at the cloud base. Using $SO_4C$ resulted in a nearly twofold enhancement of $S$ (0.88) compared to using cloud-base sulfate $SO_4B$ (0.47), suggesting a significant overestimation in prior literature.

Retrieval Bias Paradox

A "tug-of-war" exists in satellite retrievals: high cloud fractions improve $N_d$ accuracy but degrade aerosol retrievals due to "near-cloud" effects like aerosol swelling. This makes balancing retrieval accuracy at standard aggregate scales (e.g., 1° × 1° grids) extremely difficult.

Conclusion

This study identifies critical factors that skew satellite-based estimates of the Twomey effect. By accounting for updraft proxies, filtering for precipitation, and ensuring proper vertical co-location of aerosols and clouds, researchers can move toward a more accurate global radiative forcing estimate. Future improvements may come from polarimetric satellite retrievals for direct CCN measurements and better accounting for entrainment mixing processes.

Summary updated

Abstract

Aerosol–cloud interaction is the most uncertain component of the overall anthropogenic forcing of the climate, in which the Twomey effect plays a fundamental role. Satellite-based estimates of the Twomey effect are especially challenging, mainly due to the difficulty in disentangling aerosol effects on cloud droplet number concentration (Nd) from possible confounders. By combining multiple satellite observations and reanalysis, this study investigates the impacts of a) updraft, b) precipitation, c) retrieval errors, as well as (d) vertical co-location between aerosol and cloud, on the assessment of Nd-toaerosol sensitivity (S) in the context of marine warm (liquid) clouds. Our analysis suggests that S increases remarkably with both cloud base height and cloud geometric thickness (proxies for vertical velocity at cloud base), consistent with stronger aerosol-cloud interactions at larger updraft velocity. In turn, introducing the confounding effect of aerosol–precipitation interaction can artificially amplify S by an estimated 21 %, highlighting the necessity of removing precipitating clouds from analyses on the Twomey effect. It is noted that the retrieval biases in aerosol and cloud appear to underestimate S, in which cloud fraction acts as a key modulator, making it practically difficult to balance the accuracies of aerosol–cloud retrievals at aggregate scales (e.g., 1° × 1° grid). Moreover, we show that using column-integrated sulfate mass concentration (SO4C) to approximate sulfate concentration at cloud base (SO4B) can result in a degradation of correlation with Nd, along with a nearly twofold enhancement of S, mostly attributed to the inability of SO4C to capture the full spatio-temporal variability of SO4B. These findings point to several potential ways forward to account for the major influential factors practically by means of satellite observations and reanalysis, aiming at an optimal observational estimate of global radiative forcing due to the Twomey effect.
BibTeX
@inproceedings{jia2022comment,
  author = {Hailing Jia and Johannes Quaas and Edward Gryspeerdt and Christoph Böhm and Odran Sourdeval},
  title = {{Comment on acp-2021-999}},
  booktitle = {Openalex Unknown},
  year = {2022},
  doi = {10.5194/acp-2021-999-rc2},
}
Figure