-
The Impact of Cumulus Clouds and CCNs Regeneration on Aerosol Vertical Distribution and Size
Authors:
Yael Arieli,
Alexander Khain,
Ehud Gavze,
Orit Altaratz,
Eshkol Eytan,
Ilan Koren
Abstract:
This study employs a high-resolution (10m) System for Atmospheric Modeling (SAM) coupled with the Spectral Bin Microphysical (SBM) scheme to thoroughly investigate the processes governing the evolution of aerosol properties within and outside a shallow cumulus cloud. The model encompasses the complete life cycle of cloud droplets, starting from their formation through their evolution until their c…
▽ More
This study employs a high-resolution (10m) System for Atmospheric Modeling (SAM) coupled with the Spectral Bin Microphysical (SBM) scheme to thoroughly investigate the processes governing the evolution of aerosol properties within and outside a shallow cumulus cloud. The model encompasses the complete life cycle of cloud droplets, starting from their formation through their evolution until their complete evaporation or sedimentation to the ground. Additionally, the model tracks the aerosols' evolution both within droplets and in the air. Aerosols are transported within the droplets, grow by droplet coalescence, and are released into the atmosphere after droplet evaporation (regeneration process). The aerosol concentration increases by droplet evaporation and decreases along with falling drops. So, the effects of clouds on the surrounding aerosols depend on the microphysical and dynamic processes, which in turn depend on the amount of background aerosols; here, we compare clean and polluted conditions. It is shown that clouds significantly impact the vertical profile of aerosol concentration in the lower troposphere, as well as their size distribution, and can serve as a source of large (giant) cloud condensation nuclei. Furthermore, it is shown that both precipitating and non-precipitating boundary layer clouds contribute to a substantial increase in aerosol concentration within the inversion layer due to intense evaporation.
△ Less
Submitted 27 May, 2024;
originally announced May 2024.
-
Learned 3D volumetric recovery of clouds and its uncertainty for climate analysis
Authors:
Roi Ronen,
Ilan Koren,
Aviad Levis,
Eshkol Eytan,
Vadim Holodovsky,
Yoav Y. Schechner
Abstract:
Significant uncertainty in climate prediction and cloud physics is tied to observational gaps relating to shallow scattered clouds. Addressing these challenges requires remote sensing of their three-dimensional (3D) heterogeneous volumetric scattering content. This calls for passive scattering computed tomography (CT). We design a learning-based model (ProbCT) to achieve CT of such clouds, based o…
▽ More
Significant uncertainty in climate prediction and cloud physics is tied to observational gaps relating to shallow scattered clouds. Addressing these challenges requires remote sensing of their three-dimensional (3D) heterogeneous volumetric scattering content. This calls for passive scattering computed tomography (CT). We design a learning-based model (ProbCT) to achieve CT of such clouds, based on noisy multi-view spaceborne images. ProbCT infers - for the first time - the posterior probability distribution of the heterogeneous extinction coefficient, per 3D location. This yields arbitrary valuable statistics, e.g., the 3D field of the most probable extinction and its uncertainty. ProbCT uses a neural-field representation, making essentially real-time inference. ProbCT undergoes supervised training by a new labeled multi-class database of physics-based volumetric fields of clouds and their corresponding images. To improve out-of-distribution inference, we incorporate self-supervised learning through differential rendering. We demonstrate the approach in simulations and on real-world data, and indicate the relevance of 3D recovery and uncertainty to precipitation and renewable energy.
△ Less
Submitted 9 March, 2024;
originally announced March 2024.
-
3D Scattering Tomography by Deep Learning with Architecture Tailored to Cloud Fields
Authors:
Yael Sde-Chen,
Yoav Y. Schechner,
Vadim Holodovsky,
Eshkol Eytan
Abstract:
We present 3DeepCT, a deep neural network for computed tomography, which performs 3D reconstruction of scattering volumes from multi-view images. Our architecture is dictated by the stationary nature of atmospheric cloud fields. The task of volumetric scattering tomography aims at recovering a volume from its 2D projections. This problem has been studied extensively, leading, to diverse inverse me…
▽ More
We present 3DeepCT, a deep neural network for computed tomography, which performs 3D reconstruction of scattering volumes from multi-view images. Our architecture is dictated by the stationary nature of atmospheric cloud fields. The task of volumetric scattering tomography aims at recovering a volume from its 2D projections. This problem has been studied extensively, leading, to diverse inverse methods based on signal processing and physics models. However, such techniques are typically iterative, exhibiting high computational load and long convergence time. We show that 3DeepCT outperforms physics-based inverse scattering methods in term of accuracy as well as offering a significant orders of magnitude improvement in computational time. To further improve the recovery accuracy, we introduce a hybrid model that combines 3DeepCT and physics-based method. The resultant hybrid technique enjoys fast inference time and improved recovery performance.
△ Less
Submitted 10 December, 2020;
originally announced December 2020.
-
Spatiotemporal tomography based on scattered multiangular signals and its application for resolving evolving clouds using moving platforms
Authors:
Roi Ronen,
Yoav Y. Schechner,
Eshkol Eytan
Abstract:
We derive computed tomography (CT) of a time-varying volumetric translucent object, using a small number of moving cameras. We particularly focus on passive scattering tomography, which is a non-linear problem. We demonstrate the approach on dynamic clouds, as clouds have a major effect on Earth's climate. State of the art scattering CT assumes a static object. Existing 4D CT methods rely on a lin…
▽ More
We derive computed tomography (CT) of a time-varying volumetric translucent object, using a small number of moving cameras. We particularly focus on passive scattering tomography, which is a non-linear problem. We demonstrate the approach on dynamic clouds, as clouds have a major effect on Earth's climate. State of the art scattering CT assumes a static object. Existing 4D CT methods rely on a linear image formation model and often on significant priors. In this paper, the angular and temporal sampling rates needed for a proper recovery are discussed. If these rates are used, the paper leads to a representation of the time-varying object, which simplifies 4D CT tomography. The task is achieved using gradient-based optimization. We demonstrate this in physics-based simulations and in an experiment that had yielded real-world data.
△ Less
Submitted 6 December, 2020;
originally announced December 2020.
