Periodic Reporting for period 2 - CloudCT (Climate CT- Cloud Tomography by Satellites for Better Climate Prediction)
Reporting period: 2021-02-01 to 2022-07-31
Combination of advanced methods in climate modelling, computed tomography and satellite system design allow in CloudCT synergies to improve climate predictions by innovative measurement approaches to characterize clouds.
The new Intergovernmental Panel on Climate Change (IPCC) report, that was just recently published, points unequivocally again that anthropogenic activity has warmed the atmosphere, ocean, and land. At this point, it is of major importance for the scientific community to be able to predict accurately climatic trends, on global and regional scales. It was also re-emphasized in this IPCC report that that clouds’ properties and their response to climate change (cloud-feedbacks) is the main uncertainty in climate prediction. Clouds play a key role in controlling the incoming and outgoing energy transfers in the climate system. Small changes in the cloud properties (such as thickness or coverage) can yield large changes to Earth’s energy balance. Cloud fields are dynamic systems, which involve complicated interactions between dynamic, thermodynamic, microphysical and radiation processes, all coupled and modulated by a network of positive and negative feedbacks. For instance, cloud optical properties, lifetime, drizzle patterns and spatial coverage all depend heavily on environmental thermodynamic factors (temperature, humidity, winds) and in return affect them. Understanding these dynamic systems and their spatial and temporal evolution in changing environmental conditions is still regarded as the toughest challenge in climate research.
Of all cloud types warm, shallow clouds contribute the largest part in the climate uncertainty. They play a lead role by affecting moisture and heat transfer to the free atmosphere. They strongly reflect incoming solar radiation. In order to better understand these cloud properties and their trends, we need an observation system that can fully describe their coverage and microphysical properties (optical properties, droplets size distribution, liquid water content and precipitation). This will allow us to better parametrize the shallow cloud properties in climate models.
However, despite their importance, a great portion of the shallow clouds subset is poorly observed. These clouds are often sparse and small and therefore overlooked by most of the current earth-observing systems. Moreover, even if the spatial resolution of some of the current satellites is below 50m (like Sentinel, Landsat), in current technology there is a physical limit to how small clouds can be while mathematical analysis of their properties is valid.
On the other hand, we leverage technological advances. Miniaturisation technologies allow implementation of cost-efficient small satellites. This can improve for the society provision of a broad spectrum of services. Especially, if coordinated in self-organizing formations, related networks enable innovative applications in Earth observation and telecommunication.
The main objective of this synergy grant is to promote scientific progress, by combining in an interdisciplinary way
- Climate prediction by improved information about clouds
- Computed tomography methods that relate to characteristics of clouds
- Control of miniaturized satellite formations for joint measurements
The mission will fill the important knowledge gap of cloud and climate models by developing a new tomography-based approach based on small satellite technology. It is based on viewing the same small clouds by several satellites from different angles, and then to use the image data to estimate the 3D properties of the main microphysical variables. This synergy project is a tight collaboration of three groups with complementary expertise. The Weizmann Institute group studies cloud physics using theoretical approaches, observations and numerical modelling. The group is responsible for developing detailed, high-resolution models of cloud fields that should be as accurate and as similar to real clouds as possible. Such simulations should provide examples of all main shallow cloud types in order to train and validate the cloud tomography algorithms that are developed by the Technion group.
To achieve these goals the Weizmann group is improving its theoretical understanding of shallow cloud structure and processes. This can enable better high-resolution numerical models of atmospheric eddies (which drive cloud formation), thus simulate more realistic shallow cloud fields. To do so, we have developed a methodology that combines theoretical approaches that are tested and updated by various sets of observations from the ground and from space.
The Technion group studies how to obtain 3D tomographic information based on scattering, as occurs in the atmosphere, and can be done using spaceborne imaging. This means determination of the most suitable viewing angles, observations times, and types of camera sensors. The latter focuses on issues as spectral (color) bands, and polarization: a property of light which to which humans are oblivious but has a strong response to properties of droplets in a cloud. Moreover, we study how to create image analysis algorithms that are tailored to this type of imaging. The main effort involves basic research on how tomography can be achieved is the randomness of scattering media. The principles are relevant not only to clouds but also to imaging in tissue, as in medical imaging and biomedical research. We believe thus, that our effort will advance biomedical research as well as climate research.
The Center for Telematics (ZfT) group focuses on the formation and satellite aspects of this project, to enable a satellite formation to generate the multi-view images that are the basis for tomography. Thus, ZfT analyses possible orbits and formation topologies for the satellite formation as well as necessary formation flying strategies and maneuvres to acquire and maintain the required formation. In addition, ZfT designs the satellites including important subsystems like propulsion or attitude control / pointing to meet the requirements imposed by the formation topology and the observation / image taking scheme. Together with the groups at the Weizmann institute and at Technion, ZfT analyses possible camera payloads and acquires them. In summary, the main objectives and efforts are the technical design and development of a satellite platform and the design of innovative formations in 3D-configuration for optimum observations. Here in particular challenging attitude control capabilities for cooperative imaging needs to be implemented. Thus, a suitable formation topology and strategy to allow for the for the required observations is to be implemented.
The new Intergovernmental Panel on Climate Change (IPCC) report, that was just recently published, points unequivocally again that anthropogenic activity has warmed the atmosphere, ocean, and land. At this point, it is of major importance for the scientific community to be able to predict accurately climatic trends, on global and regional scales. It was also re-emphasized in this IPCC report that that clouds’ properties and their response to climate change (cloud-feedbacks) is the main uncertainty in climate prediction. Clouds play a key role in controlling the incoming and outgoing energy transfers in the climate system. Small changes in the cloud properties (such as thickness or coverage) can yield large changes to Earth’s energy balance. Cloud fields are dynamic systems, which involve complicated interactions between dynamic, thermodynamic, microphysical and radiation processes, all coupled and modulated by a network of positive and negative feedbacks. For instance, cloud optical properties, lifetime, drizzle patterns and spatial coverage all depend heavily on environmental thermodynamic factors (temperature, humidity, winds) and in return affect them. Understanding these dynamic systems and their spatial and temporal evolution in changing environmental conditions is still regarded as the toughest challenge in climate research.
Of all cloud types warm, shallow clouds contribute the largest part in the climate uncertainty. They play a lead role by affecting moisture and heat transfer to the free atmosphere. They strongly reflect incoming solar radiation. In order to better understand these cloud properties and their trends, we need an observation system that can fully describe their coverage and microphysical properties (optical properties, droplets size distribution, liquid water content and precipitation). This will allow us to better parametrize the shallow cloud properties in climate models.
However, despite their importance, a great portion of the shallow clouds subset is poorly observed. These clouds are often sparse and small and therefore overlooked by most of the current earth-observing systems. Moreover, even if the spatial resolution of some of the current satellites is below 50m (like Sentinel, Landsat), in current technology there is a physical limit to how small clouds can be while mathematical analysis of their properties is valid.
On the other hand, we leverage technological advances. Miniaturisation technologies allow implementation of cost-efficient small satellites. This can improve for the society provision of a broad spectrum of services. Especially, if coordinated in self-organizing formations, related networks enable innovative applications in Earth observation and telecommunication.
The main objective of this synergy grant is to promote scientific progress, by combining in an interdisciplinary way
- Climate prediction by improved information about clouds
- Computed tomography methods that relate to characteristics of clouds
- Control of miniaturized satellite formations for joint measurements
The mission will fill the important knowledge gap of cloud and climate models by developing a new tomography-based approach based on small satellite technology. It is based on viewing the same small clouds by several satellites from different angles, and then to use the image data to estimate the 3D properties of the main microphysical variables. This synergy project is a tight collaboration of three groups with complementary expertise. The Weizmann Institute group studies cloud physics using theoretical approaches, observations and numerical modelling. The group is responsible for developing detailed, high-resolution models of cloud fields that should be as accurate and as similar to real clouds as possible. Such simulations should provide examples of all main shallow cloud types in order to train and validate the cloud tomography algorithms that are developed by the Technion group.
To achieve these goals the Weizmann group is improving its theoretical understanding of shallow cloud structure and processes. This can enable better high-resolution numerical models of atmospheric eddies (which drive cloud formation), thus simulate more realistic shallow cloud fields. To do so, we have developed a methodology that combines theoretical approaches that are tested and updated by various sets of observations from the ground and from space.
The Technion group studies how to obtain 3D tomographic information based on scattering, as occurs in the atmosphere, and can be done using spaceborne imaging. This means determination of the most suitable viewing angles, observations times, and types of camera sensors. The latter focuses on issues as spectral (color) bands, and polarization: a property of light which to which humans are oblivious but has a strong response to properties of droplets in a cloud. Moreover, we study how to create image analysis algorithms that are tailored to this type of imaging. The main effort involves basic research on how tomography can be achieved is the randomness of scattering media. The principles are relevant not only to clouds but also to imaging in tissue, as in medical imaging and biomedical research. We believe thus, that our effort will advance biomedical research as well as climate research.
The Center for Telematics (ZfT) group focuses on the formation and satellite aspects of this project, to enable a satellite formation to generate the multi-view images that are the basis for tomography. Thus, ZfT analyses possible orbits and formation topologies for the satellite formation as well as necessary formation flying strategies and maneuvres to acquire and maintain the required formation. In addition, ZfT designs the satellites including important subsystems like propulsion or attitude control / pointing to meet the requirements imposed by the formation topology and the observation / image taking scheme. Together with the groups at the Weizmann institute and at Technion, ZfT analyses possible camera payloads and acquires them. In summary, the main objectives and efforts are the technical design and development of a satellite platform and the design of innovative formations in 3D-configuration for optimum observations. Here in particular challenging attitude control capabilities for cooperative imaging needs to be implemented. Thus, a suitable formation topology and strategy to allow for the for the required observations is to be implemented.
The first period of this project, was dedicated to update and correct schemes in the numerical large eddy simulator (LES) model and to tune it to super-high-resolution grids. We have produced for the first time LES runs of shallow clouds in 10m resolution. These improved models allow us to resolve small scale processes that until now were only parametrized. They capture well fine-scale features in the cloud’s dynamics and microphysics. This enables detailed examination of the microphysical processes in the core and margins of the clouds, as well as better understanding of the challenging regimes of mixing and entrainment within a convective cloud.
In addition, we determined a new type of shallow cumulus clouds: green Cu. These continental shallow Cu clouds form over vegetated regions and they share similar properties even though they form all over the globe, from the tropics to mid-latitudes. We investigated their common properties, organization patterns, global distribution and their governing environmental thermodynamic conditions. We have created a novel global dataset of green Cu that was used and will be used for studying these clouds. Using our analyses of satellite data, we have shown important contribution of the mostly overlooked twilight clouds to the longwave budget of the climate system. Such contribution of areas that are regarded as cloud-free were not attributed to clouds. Such a finding emphasises the need to consider a 3rd class in the atmosphere (the twilight zone) which is not cloud-free, but instead is affected by the presence of clouds and therefore has different optical properties. In addition, our observational analysis yielded a research project that showed that deep-convective clouds have transported heavy smoke from the Australian fires, to the stratosphere, yielding a volcanic-eruption-like effect of long-lived stratospheric aerosols that reduce the solar energy influx to the lower atmosphere.
Our theoretical studies allow us to set the foundations for a new data-driven stochastic modeling of variety of cloud fields. We predict that data-driven stochastic modeling is going to be one of the most promising fronts of alternative data models. In addition, we explored the fundamental mechanisms governing open-closed cells transitions such as occurring in marine stratocumulus which form the most important reflecting cloud decks over the subtropical oceans.
We thoroughly studied and determined the characteristics of a desirable and achievable payload for the CloudCT mission, and a desired orbit for the satellite formation. Algorithmically, we developed ways to make the recovery more efficient and accurate. These include introduction of polarization to scattering tomography, use of a prior on cloud structure (extinction tending to grow with altitude in a cloud), coarse-to-fine analysis of large fields, initialization schemes of optimization, stochastic (Monte-Carlo) learning-based analysis and spatiotemporal tomography.
We exploited analogies of scattering in the atmosphere to tissue, to suggest a way of improving X-ray CT, by multiplexing X-ray sources (thus speeding up acquisition) while removing an anti-scatter grid (thus reducing dose on patients). Moreover, we found that our understanding of tomography in random media can significantly aid in-situ imaging and analysis of plankton. Plankton have major ecological implications, depending on their density, size distribution and 3D structure. However, in-situ they appear randomly in a sensor. Thus we showed how stochastic tomography can help finding these unknowns.
In addition, we established the CloudCT website, and advanced in setting up infrastructure for the project, including computing and calibration hardware.
In addition, we determined a new type of shallow cumulus clouds: green Cu. These continental shallow Cu clouds form over vegetated regions and they share similar properties even though they form all over the globe, from the tropics to mid-latitudes. We investigated their common properties, organization patterns, global distribution and their governing environmental thermodynamic conditions. We have created a novel global dataset of green Cu that was used and will be used for studying these clouds. Using our analyses of satellite data, we have shown important contribution of the mostly overlooked twilight clouds to the longwave budget of the climate system. Such contribution of areas that are regarded as cloud-free were not attributed to clouds. Such a finding emphasises the need to consider a 3rd class in the atmosphere (the twilight zone) which is not cloud-free, but instead is affected by the presence of clouds and therefore has different optical properties. In addition, our observational analysis yielded a research project that showed that deep-convective clouds have transported heavy smoke from the Australian fires, to the stratosphere, yielding a volcanic-eruption-like effect of long-lived stratospheric aerosols that reduce the solar energy influx to the lower atmosphere.
Our theoretical studies allow us to set the foundations for a new data-driven stochastic modeling of variety of cloud fields. We predict that data-driven stochastic modeling is going to be one of the most promising fronts of alternative data models. In addition, we explored the fundamental mechanisms governing open-closed cells transitions such as occurring in marine stratocumulus which form the most important reflecting cloud decks over the subtropical oceans.
We thoroughly studied and determined the characteristics of a desirable and achievable payload for the CloudCT mission, and a desired orbit for the satellite formation. Algorithmically, we developed ways to make the recovery more efficient and accurate. These include introduction of polarization to scattering tomography, use of a prior on cloud structure (extinction tending to grow with altitude in a cloud), coarse-to-fine analysis of large fields, initialization schemes of optimization, stochastic (Monte-Carlo) learning-based analysis and spatiotemporal tomography.
We exploited analogies of scattering in the atmosphere to tissue, to suggest a way of improving X-ray CT, by multiplexing X-ray sources (thus speeding up acquisition) while removing an anti-scatter grid (thus reducing dose on patients). Moreover, we found that our understanding of tomography in random media can significantly aid in-situ imaging and analysis of plankton. Plankton have major ecological implications, depending on their density, size distribution and 3D structure. However, in-situ they appear randomly in a sensor. Thus we showed how stochastic tomography can help finding these unknowns.
In addition, we established the CloudCT website, and advanced in setting up infrastructure for the project, including computing and calibration hardware.
In the cloud physics front we expect to set the foundations and to develop a new physical description of the formation and life cycle of a convective cloud. It will be a more realistic description of the clouds’ vertical movement, the role of entrainment and mixing and how the clouds microphysics is affected by it. Combining the new CloudCT measurements, theoretical studies, and updating the LES model will yield a much deeper understanding of shallow clouds, their coverage, lifetime, optical properties and role in climate. Moreover, we expect our measurements and theoretical models to provide a more realistic way to parametrize such clouds in climate model and therefore reducing the uncertainty. In parallel our new measurements will allow us to further develop alternative approaches of simulations that their governing rules are all more constrained by data-driven trends. Finally, combining deeper understanding of cloud observations, reanalysis data, LES simulations and the new CloudCT data will allow us to better capture climate trends that are hard to capture due to the complexity and the high natural variance of cloud systems.
In the coming period of the project we look forward to the building, launch and operations of the CloudCT satellites. Prior to this stage, we anticipate an intense period of testing and calibration. These include, actually, research into new types of calibration tasks, such as polarimetric calibration in orbit. After launch, we foresee a much more intense effort on analysis of spaceborne data. We further expect that the tomographic analysis methods we derive will keep improving regarding efficient use of computing resources, accuracy of results and application domains.
We expect further progress in the field of distributed control of small satellite formations. These are meant to pursue a common goal and perform joined target pointing and joined observations. These may include further developments and results in control theory in the sense of distributed formation (position) control, as well as precise cooperative attitude control (pointing towards a joint observation target), collaborative navigation in distributed systems, as well as in the real-world demonstration of such a self-organizing satellite formation.
In the coming period of the project we look forward to the building, launch and operations of the CloudCT satellites. Prior to this stage, we anticipate an intense period of testing and calibration. These include, actually, research into new types of calibration tasks, such as polarimetric calibration in orbit. After launch, we foresee a much more intense effort on analysis of spaceborne data. We further expect that the tomographic analysis methods we derive will keep improving regarding efficient use of computing resources, accuracy of results and application domains.
We expect further progress in the field of distributed control of small satellite formations. These are meant to pursue a common goal and perform joined target pointing and joined observations. These may include further developments and results in control theory in the sense of distributed formation (position) control, as well as precise cooperative attitude control (pointing towards a joint observation target), collaborative navigation in distributed systems, as well as in the real-world demonstration of such a self-organizing satellite formation.