Highlighted Studies in a Nutshell

Quantifying the contribution of the sky near clouds to Earth’s energy budget 

Clouds play a crucial role in Earth's climate system, significantly altering the radiative flux at the top-of-the-atmosphere and creating large deviations from clear-sky fluxes. As a result, atmospheric science often separates the sky into cloudy and clear regions, studying each independently. Here we analyze multi-spectral images of cloud fields measured from a satellite, alongside co-located measurements of radiative flux at the top of the atmosphere (TOA) from a different, lower-resolution instrument. We re-cast the equation for radiative fluxes at TOA and use machine learning to estimate cloud effects in the “clear sky” for the first time. Our findings indicate that the local RE of marine clouds on their surroundings is very large. In the solar part of the spectrum, clouds significantly enhance the aerosol RE. In the longwave infrared, we confirm that the near-cloud RE is equivalent to a substantial concentration of CO2. Considering these results, we advocate for categorizing the sky into three regimes — cloudy, cloud-influenced-clear-sky, and far-field clear-sky. 

Highlighting the important role of ring vortices in shallow Cumulus clouds' mixing. 

Shallow convective clouds play a crucial role in climate as they transfer heat in the vertical dimension and affect radiation transfer in the atmosphere. These clouds are much smaller than climate models' resolution, and so they are represented by simplified equations in weather and climate models (parameterizations). This simple representation of such important processes is one of the largest sources of uncertainty in climate models. In particular, the process of clouds mixing with their dry surroundings is known to have a large contribution to the uncertainty reflected by clouds in climate prediction. In this work, we use high-resolution simulations of cumulus clouds to investigate the toroidal (ring) vortex that is located at the top of a rising thermal (a known property of cumulus dynamics). We show that cloud-scale vortices dominate cloud dilution and are at least as important as stochastic turbulent motions which are often considered in mixing parameterizations. These ideas can serve for future parameterizations of shallow cumulus clouds in coarse-resolution models.

Biomass burning aerosol are presented as analogs to extreme greenhouse gas concentrations in forming a hot-house climate and extreme precipitation. 

According to the latest IPCC report, aerosol, clouds and aerosol-cloud interactions are the largest source of uncertainty in climate prediction. Aerosol species that absorb solar radiation generate local warming of the atmosphere that changes the vertical profile of temperature and by that affects cloud and precipitation development. In this paper we used idealized computer simulations to investigate the effect of absorbing aerosols on precipitation, and specifically on extreme precipitation events in the tropics. We demonstrate that under certain conditions, absorbing aerosols can strongly enhance extreme precipitation even despite reducing the mean rain amount. We show that this trend can be explained by a mechanism previously reported for much warmer climate conditions than currently found on Earth, involving heating by radiation of the lower part of the troposphere. These results have implications for climate change mitigation and disaster risk management.

High resolution simulations of clouds and flow tracers are used to test methods for estimating the basic adiabatic condensation approximation.

The process of mixing in warm convective clouds and its effects on microphysics are important for predicting weather and climate. Still, they remain open questions in the field of cloud physics. Adiabatic regions in the cloud could be considered diluted areas and for this reason, the adiabatic fraction (AF) is an important parameter that estimates the mixing level in the cloud in a simple way. Here, we test different methods of AF calculations using high-resolution (10 m) simulations of isolated warm cumulus clouds. The calculated AFs are compared with a normalized concentration of a passive tracer, which is an accurate measure of dilution by mixing. This comparison enables the examination of how well the AF parameter can determine mixing effects and the estimation of the accuracy of different approaches used to calculate it. Comparison of three different methods to derive AF, with the passive tracer, shows that one method is much more robust than the others. Moreover, this method's equation structure also allows for the isolation of different assumptions that are often practiced when calculating AF. The use of a detailed spectral bin microphysics scheme allows an accurate description of the supersaturation field and demonstrates that the accuracy of the saturation adjustment assumption depends on aerosol concentration, leading to an underestimation of AF in pristine environments.

Thermal infrared radiation is used to give "more colors" to the ill-defined sky near clouds.

The binary delineation of cloudy and clear sky is not clearly defined due to the presence of a transitionary zone, known as the cloud twilight zone, consisting of humidity, cloud droplets and humidified to dry aerosols. The twilight zone is an inherent component of cloud fields, yet its influence on longwave-infrared radiation remains unknown. This study present analysis spectral data from global satellite observations of shallow cloud fields over the ocean to estimate a lower bound on the twilight zone’s effect on longwave radiation. We find that the average longwave radiative effect of the twilight zone is ~0.75 W m–2, which is equivalent to the radiative forcing from increasing atmospheric CO2 by 75 ppm. We also find that the twilight zone in the longwave occupies over 60% of the apparent clear sky within the analyzed cloud fields. 

Machine learning is used for a remote sensing retrieval of clouds in 3D, including the uncertainty per volume sample (voxel)

Understanding clouds is one of the biggest challenges in predicting climate change, especially because we don’t fully know what’s happening inside them. This is particularly true for small, scattered clouds that form in the lower atmosphere. These clouds are shaped by turbulent air movement, making their insides highly variable and constantly changing. Aircraft in-situ measurements are too course to capture their internal structure that occurs on small spatial-temporal scales. Moreover, their heterogenous internal structure means that the radiative transfer through them needs to consider all three dimensions, though most atmospheric application simplified it to the vertical dimension alone.  In this study, we simulate ten tiny satellites that fly in formation and take pictures of the same clouds from different angles at the same time. We then use machine learning to recover the 3D internal structure of shallow clouds and derive statistics such as uncertainty. We apply this on real-world data to demonstrate proof of concepts for remote sensing key features for predicting precipitation and renewable energy.

Lagrangian tracking in high resolution shows the spatio-temporal scales of three distinct flow regimes in shallow Cumulus.

Understanding how shallow clouds mix with the surrounding air is a major challenge in atmopsheric science, yet it’s essential for improving weather and climate predictions. This process is difficult to study because cloud behavior is complex and hard to observe or simulate accurately. In this study, we use high-resolution computer simulations and a technique called Lagrangian tracer tracking, which follows the paths of individual air particles, to explore how mixing happens inside cumulus clouds. By analyzing when and where these particles enter and exit the cloud, and how long they stay, we identified three distinct mixing patterns that occur during a cloud’s life. We also found a fourth pattern that appears when the cloud begins to dissipate.

Wavelet analysis is used to separate features of small-scale turbulent flow from larger cloud-scale flows.

Atmospheric motions in clouds and cloud surroundings have a wide range of scales, from several kilometers to centimeters. These motions have different impacts on cloud dynamics and microphysics. Larger-scale motions (hereafter referred to as convective motions) are responsible for mass transport over distances comparable with cloud scale, while motions of smaller scales (hereafter referred to as turbulent motions) are stochastic. The present study aimed to describe the method for separating the motion scale into a convective component and a turbulent component. A wavelet method is used to separate the velocity field into the convective and turbulent components. The efficiency of the method is demonstrated by an example of a vertical velocity field of a cumulus cloud simulated with detailed microphysics and high resolution of 10 m. It is shown that vertical velocity in clouds can indeed be represented as a sum of convective velocity (forming zone of cloud updrafts and subsiding shell) and a stochastic velocity obeying laws of homogeneous and isotropic turbulence.

The time dimension is used to improve 3D cloud tomography.

While traditional cloud tomography assumes a static cloud, this work introduces a method that incorporates time, capturing both cloud evolution and camera motion. Spatiotemporal CT is achieved through an optimization framework that leverages the cloud’s correlation time. The approach is demonstrated using both simulated clouds and real-world observational data.