Archive
This archive lists my research activities over the past years. For a listing by research category or by research topic, please see below.
- Traffic Noise Estimation from Satellite Imagery with Deep Learning | 2022-06
- Contrastive Self-Supervised Learning for Multi-modal Earth Observation Data | 2022-06
- Estimating Power Plant Greenhouse Gas Emissions from Satellite Imagery | 2021-12
- Estimation of Surface Level NO2 from Remote Sensing Data | 2021-11
- Commercial Vehicle Traffic Detection from Satellite Imagery with Deep Learning | 2021-11
- Power Plant Classification from Remote Imaging with Deep Learning | 2021-06
- Characterization of Industrial Smoke Plumes from Remote Sensing Data | 2020-12
- Don Quixote and the Dormant Comets - My Astronomical Legacy | 2020-05
- Automated Cloud Detection with Machine Learning | 2020-04
- SpitzerNEOs - Diameters and Albedos for 2132 Near-Earth Objects | 2020-02
- A plethora of asteroids in TESS data | 2019-03
- Asteroid Shape Information from Gaia DR2 | 2019-01
- Are there Limits on the Applicability of Asteroid Thermal Models for Near-Earth Asteroids? | 2019-01
- sbpy - A Python module for small-body planetary astronomy | 2018-07
- Photometrypipeline | 2017-02
- How many Dead Comets are there? | 2015-08
- Rapid-Response Spectrophotometric Observations of NEOs with UKIRT and RATIR | 2014-11
- Physical Properties of two tiny Asteroids from Spitzer Observations | 2014-10
- Detection of Cometary Activity in NEO Don Quixote | 2014-10
- Herschel Observations of Plutinos | 2014-10
Applied Research Earth Observation Deep Learning Segmentation Traffic Noise Sentinel-2
Road traffic noise is a global issue that can lead to severe health effects. Despite the ubiquity of traffic noise, its quantification or estimation is complicated and detailed road traffic maps are only available for select countries or areas. We investigate whether it is possible to train a segmentation model to esimate road traffic noise from satellite imagery.
Fundamental Research Earth Observation Deep Learning Self-supervised Learnig Data Fusion Transformers Classification Segmentation
Self-supervised learning provides a powerful means to pretrain models based on un-labeled data. Un-labeled Earth observation data are abundant: this circumstance combined with the availability of multi modal data makes Earth observation a perfect playground for self-supervised learning. Our early results are very promising...
Applied Research Earth Observation Deep Learning Multi-task Learning Power Plants CO2 Sentinel-2
Tracking Greenhouse Gas emissions will be increasingly important in the future, as power generation from fossil fuels is supposed to fade out. Independent tools to monitor power plants are required for this task. We present a method that uses freely available satellite imagery to estimate power generation and CO2 emission rates on a global scale.
Applied Research Earth Observation Deep Learning Air Pollution NO2 Sentinel-2
Air pollution is a major health issue and often also contributes to climate change. Measuring air pollution is costly and therefore only available in some countries. We investigated whether freely available Earth observation data can be utilized to estimate air pollution on the surface level.
Applied Research Earth Observation Deep Learning Traffic Sentinel-2
Can we identify trucks from space and estimate truck traffic rates anywhere on the planet? Yes, we can!
Applied Research Earth Observation Deep Learning Classification Power Plants Sentinel-2
We developed a deep learning model that is able to distinguish between 10 different types of power plants in an effort to automatically identify and characterize industrial sites in satellite imagery. This work will help us to estimate greenhouse gas emission rates for individual industrial sites in the future.
Applied Research Earth Observation Deep Learning Segmentation Plumes Sentinel-2
Greenhouse gas emissions from the industrial economic sector are a major driver of the currently observed climate change. We developed a deep learning approach to identify and characterize industrial smoke plumes. In the future, we will utilize this approach to estimate greenhouse gas emissions from remote sensing data on a global scale.
Astronomy Don Quixote Dormant Comets Asteroids
Don Quixote has been considered asteroidal since its discovery in 1983, despite its comet-like orbit. We found activity in this object during its 2009 and 2018 apparitions, leading us to believe that it really is an active comet. But are there other objects like Don Quixote? I monitored a sample of ~100 asteroids that are somewhat likely to turn active, too, over a period of more than 4 years. Bottom line: Don Quixote is rather unique.
Astronomy Machine Learning Deep Learning Allsky-Camera
This is a toy project that turned into a real research project and a preparation for my new job as research scientist in computer vision: using machine learning techniques to identify clouds in all-sky camera images.
Astronomy Near-Earth Asteroids Asteroid Physical Properties Asteroid Thermal Modeling Spitzer Space Telescope
Based on Spitzer Space Telescope observations, we derived diameter and albedo estimates for 2132 asteroids in near-Earth space - the largest dataset of its kind. Our results and data are available online.
Astronomy Asteroids Asteroid Physical Properties TESS
TESS is a satellite that observes bright stars to find brightness modulations revealing exoplanet transits - but it also observes a lot of asteroids over a long period of time, which makes it a unique asset for deriving long asteroid rotation periods.
Astronomy Asteroids Asteroid Physical Properties Gaia
The Gaia mission builds a unique survey of the stars in our Milky Way - but it also observes asteroids that are crossing its field of view. We derive ensemble shape information for different asteroid populations from the first batch of asteroid data from Gaia.
Astronomy Asteroid Thermal Modeling Asteroid Thermophysical Modeling
There is a range of asteroid thermal models that are based on different physical assumptions. In this analysis, I investigate how these assumptions affect their results and applicability to different situations.
Astronomy Open-Source Software
sbpy is an astropy affiliated package for small-body planetary astronomers. We proposed this idea to NASA and were funded for developing this Python module, which is outlined here.
Astronomy Open-Source Software Photometry Pipeline
photometrypipeline is a Python software package for automated image registration, calibration, and extraction of photometry tailored to the needs of asteroid observer. However, pp can also be applied to other imaging observations.
Astronomy Near-Earth Asteroids Dormant Comets
Dead comets are small bodies that appear as inactive asteroids, but have a cometary origin. Some of these objects might still harbor ices and can still activate and appear comet-like. Since these objects might signifcantly contribute to the volatile reservoir of the asteroid population, it is important to understand how many of these objects there are.
Astronomy Near-Earth Asteroids Rapid-Response Observations Spectrophotometry UKIRT
With rapid-response observations using UKIRT, we are able to observe newly discovered near-Earth asteroids when they are still close to Earth and thus bright. Our observations enable a probabilistic taxonomic classification of asteroids with typically small sizes and the determination of their compositional distribution.
Astronomy Near-Earth Asteroids Asteroid Physical Properties Spitzer Space Telescope
We observed two very small asteroids with the Spitzer Space Telescope to actually measure their physical properties for the first time. The constraints that we were able to place on their properties do not agree with the standard picture of small asteroids.
Astronomy Don Quixote Dormant Comets Spitzer Space Telescope
Near-Earth asteroid Don Quixote has long since been considered a good candidate for an inactive comet nucleus due to its comet-like orbit. Since its discovery in 1983, comet-like has never been observed - until we observed this object with the Spitzer Space Telescope...
Astronomy Plutinos Asteroid Physical Properties Herschel Space Observatory
My first research paper. Using Herschel Space Observatory observations, we investigated the physical properties of 18 Plutinos - small bodies at the outskirts of the Solar System that have orbital properties similar to those of Pluto. What we found? That Pluto is pretty unique...
Research categories
Astronomy
- Don Quixote and the Dormant Comets - My Astronomical Legacy | 2020-05
- Automated Cloud Detection with Machine Learning | 2020-04
- SpitzerNEOs - Diameters and Albedos for 2132 Near-Earth Objects | 2020-02
- A plethora of asteroids in TESS data | 2019-03
- Asteroid Shape Information from Gaia DR2 | 2019-01
- Are there Limits on the Applicability of Asteroid Thermal Models for Near-Earth Asteroids? | 2019-01
- sbpy - A Python module for small-body planetary astronomy | 2018-07
- Photometrypipeline | 2017-02
- How many Dead Comets are there? | 2015-08
- Rapid-Response Spectrophotometric Observations of NEOs with UKIRT and RATIR | 2014-11
- Physical Properties of two tiny Asteroids from Spitzer Observations | 2014-10
- Detection of Cometary Activity in NEO Don Quixote | 2014-10
- Herschel Observations of Plutinos | 2014-10
Astronomy Don Quixote Dormant Comets Asteroids
Don Quixote has been considered asteroidal since its discovery in 1983, despite its comet-like orbit. We found activity in this object during its 2009 and 2018 apparitions, leading us to believe that it really is an active comet. But are there other objects like Don Quixote? I monitored a sample of ~100 asteroids that are somewhat likely to turn active, too, over a period of more than 4 years. Bottom line: Don Quixote is rather unique.
Astronomy Machine Learning Deep Learning Allsky-Camera
This is a toy project that turned into a real research project and a preparation for my new job as research scientist in computer vision: using machine learning techniques to identify clouds in all-sky camera images.
Astronomy Near-Earth Asteroids Asteroid Physical Properties Asteroid Thermal Modeling Spitzer Space Telescope
Based on Spitzer Space Telescope observations, we derived diameter and albedo estimates for 2132 asteroids in near-Earth space - the largest dataset of its kind. Our results and data are available online.
Astronomy Asteroids Asteroid Physical Properties TESS
TESS is a satellite that observes bright stars to find brightness modulations revealing exoplanet transits - but it also observes a lot of asteroids over a long period of time, which makes it a unique asset for deriving long asteroid rotation periods.
Astronomy Asteroids Asteroid Physical Properties Gaia
The Gaia mission builds a unique survey of the stars in our Milky Way - but it also observes asteroids that are crossing its field of view. We derive ensemble shape information for different asteroid populations from the first batch of asteroid data from Gaia.
Astronomy Asteroid Thermal Modeling Asteroid Thermophysical Modeling
There is a range of asteroid thermal models that are based on different physical assumptions. In this analysis, I investigate how these assumptions affect their results and applicability to different situations.
Astronomy Open-Source Software
sbpy is an astropy affiliated package for small-body planetary astronomers. We proposed this idea to NASA and were funded for developing this Python module, which is outlined here.
Astronomy Open-Source Software Photometry Pipeline
photometrypipeline is a Python software package for automated image registration, calibration, and extraction of photometry tailored to the needs of asteroid observer. However, pp can also be applied to other imaging observations.
Astronomy Near-Earth Asteroids Dormant Comets
Dead comets are small bodies that appear as inactive asteroids, but have a cometary origin. Some of these objects might still harbor ices and can still activate and appear comet-like. Since these objects might signifcantly contribute to the volatile reservoir of the asteroid population, it is important to understand how many of these objects there are.
Astronomy Near-Earth Asteroids Rapid-Response Observations Spectrophotometry UKIRT
With rapid-response observations using UKIRT, we are able to observe newly discovered near-Earth asteroids when they are still close to Earth and thus bright. Our observations enable a probabilistic taxonomic classification of asteroids with typically small sizes and the determination of their compositional distribution.
Astronomy Near-Earth Asteroids Asteroid Physical Properties Spitzer Space Telescope
We observed two very small asteroids with the Spitzer Space Telescope to actually measure their physical properties for the first time. The constraints that we were able to place on their properties do not agree with the standard picture of small asteroids.
Astronomy Don Quixote Dormant Comets Spitzer Space Telescope
Near-Earth asteroid Don Quixote has long since been considered a good candidate for an inactive comet nucleus due to its comet-like orbit. Since its discovery in 1983, comet-like has never been observed - until we observed this object with the Spitzer Space Telescope...
Astronomy Plutinos Asteroid Physical Properties Herschel Space Observatory
My first research paper. Using Herschel Space Observatory observations, we investigated the physical properties of 18 Plutinos - small bodies at the outskirts of the Solar System that have orbital properties similar to those of Pluto. What we found? That Pluto is pretty unique...
Applied Research
- Traffic Noise Estimation from Satellite Imagery with Deep Learning | 2022-06
- Estimating Power Plant Greenhouse Gas Emissions from Satellite Imagery | 2021-12
- Estimation of Surface Level NO2 from Remote Sensing Data | 2021-11
- Commercial Vehicle Traffic Detection from Satellite Imagery with Deep Learning | 2021-11
- Power Plant Classification from Remote Imaging with Deep Learning | 2021-06
- Characterization of Industrial Smoke Plumes from Remote Sensing Data | 2020-12
Applied Research Earth Observation Deep Learning Segmentation Traffic Noise Sentinel-2
Road traffic noise is a global issue that can lead to severe health effects. Despite the ubiquity of traffic noise, its quantification or estimation is complicated and detailed road traffic maps are only available for select countries or areas. We investigate whether it is possible to train a segmentation model to esimate road traffic noise from satellite imagery.
Applied Research Earth Observation Deep Learning Multi-task Learning Power Plants CO2 Sentinel-2
Tracking Greenhouse Gas emissions will be increasingly important in the future, as power generation from fossil fuels is supposed to fade out. Independent tools to monitor power plants are required for this task. We present a method that uses freely available satellite imagery to estimate power generation and CO2 emission rates on a global scale.
Applied Research Earth Observation Deep Learning Air Pollution NO2 Sentinel-2
Air pollution is a major health issue and often also contributes to climate change. Measuring air pollution is costly and therefore only available in some countries. We investigated whether freely available Earth observation data can be utilized to estimate air pollution on the surface level.
Applied Research Earth Observation Deep Learning Traffic Sentinel-2
Can we identify trucks from space and estimate truck traffic rates anywhere on the planet? Yes, we can!
Applied Research Earth Observation Deep Learning Classification Power Plants Sentinel-2
We developed a deep learning model that is able to distinguish between 10 different types of power plants in an effort to automatically identify and characterize industrial sites in satellite imagery. This work will help us to estimate greenhouse gas emission rates for individual industrial sites in the future.
Applied Research Earth Observation Deep Learning Segmentation Plumes Sentinel-2
Greenhouse gas emissions from the industrial economic sector are a major driver of the currently observed climate change. We developed a deep learning approach to identify and characterize industrial smoke plumes. In the future, we will utilize this approach to estimate greenhouse gas emissions from remote sensing data on a global scale.
Fundamental Research
- Contrastive Self-Supervised Learning for Multi-modal Earth Observation Data | 2022-06
Fundamental Research Earth Observation Deep Learning Self-supervised Learnig Data Fusion Transformers Classification Segmentation
Self-supervised learning provides a powerful means to pretrain models based on un-labeled data. Un-labeled Earth observation data are abundant: this circumstance combined with the availability of multi modal data makes Earth observation a perfect playground for self-supervised learning. Our early results are very promising...
Research topics
Plutinos
- Herschel Observations of Plutinos | 2014-10
Astronomy Plutinos Asteroid Physical Properties Herschel Space Observatory
My first research paper. Using Herschel Space Observatory observations, we investigated the physical properties of 18 Plutinos - small bodies at the outskirts of the Solar System that have orbital properties similar to those of Pluto. What we found? That Pluto is pretty unique...
Asteroid Physical Properties
- SpitzerNEOs - Diameters and Albedos for 2132 Near-Earth Objects | 2020-02
- A plethora of asteroids in TESS data | 2019-03
- Asteroid Shape Information from Gaia DR2 | 2019-01
- Physical Properties of two tiny Asteroids from Spitzer Observations | 2014-10
- Herschel Observations of Plutinos | 2014-10
Astronomy Near-Earth Asteroids Asteroid Physical Properties Asteroid Thermal Modeling Spitzer Space Telescope
Based on Spitzer Space Telescope observations, we derived diameter and albedo estimates for 2132 asteroids in near-Earth space - the largest dataset of its kind. Our results and data are available online.
Astronomy Asteroids Asteroid Physical Properties TESS
TESS is a satellite that observes bright stars to find brightness modulations revealing exoplanet transits - but it also observes a lot of asteroids over a long period of time, which makes it a unique asset for deriving long asteroid rotation periods.
Astronomy Asteroids Asteroid Physical Properties Gaia
The Gaia mission builds a unique survey of the stars in our Milky Way - but it also observes asteroids that are crossing its field of view. We derive ensemble shape information for different asteroid populations from the first batch of asteroid data from Gaia.
Astronomy Near-Earth Asteroids Asteroid Physical Properties Spitzer Space Telescope
We observed two very small asteroids with the Spitzer Space Telescope to actually measure their physical properties for the first time. The constraints that we were able to place on their properties do not agree with the standard picture of small asteroids.
Astronomy Plutinos Asteroid Physical Properties Herschel Space Observatory
My first research paper. Using Herschel Space Observatory observations, we investigated the physical properties of 18 Plutinos - small bodies at the outskirts of the Solar System that have orbital properties similar to those of Pluto. What we found? That Pluto is pretty unique...
Herschel Space Observatory
- Herschel Observations of Plutinos | 2014-10
Astronomy Plutinos Asteroid Physical Properties Herschel Space Observatory
My first research paper. Using Herschel Space Observatory observations, we investigated the physical properties of 18 Plutinos - small bodies at the outskirts of the Solar System that have orbital properties similar to those of Pluto. What we found? That Pluto is pretty unique...
Don Quixote
- Don Quixote and the Dormant Comets - My Astronomical Legacy | 2020-05
- Detection of Cometary Activity in NEO Don Quixote | 2014-10
Astronomy Don Quixote Dormant Comets Asteroids
Don Quixote has been considered asteroidal since its discovery in 1983, despite its comet-like orbit. We found activity in this object during its 2009 and 2018 apparitions, leading us to believe that it really is an active comet. But are there other objects like Don Quixote? I monitored a sample of ~100 asteroids that are somewhat likely to turn active, too, over a period of more than 4 years. Bottom line: Don Quixote is rather unique.
Astronomy Don Quixote Dormant Comets Spitzer Space Telescope
Near-Earth asteroid Don Quixote has long since been considered a good candidate for an inactive comet nucleus due to its comet-like orbit. Since its discovery in 1983, comet-like has never been observed - until we observed this object with the Spitzer Space Telescope...
Dormant Comets
- Don Quixote and the Dormant Comets - My Astronomical Legacy | 2020-05
- How many Dead Comets are there? | 2015-08
- Detection of Cometary Activity in NEO Don Quixote | 2014-10
Astronomy Don Quixote Dormant Comets Asteroids
Don Quixote has been considered asteroidal since its discovery in 1983, despite its comet-like orbit. We found activity in this object during its 2009 and 2018 apparitions, leading us to believe that it really is an active comet. But are there other objects like Don Quixote? I monitored a sample of ~100 asteroids that are somewhat likely to turn active, too, over a period of more than 4 years. Bottom line: Don Quixote is rather unique.
Astronomy Near-Earth Asteroids Dormant Comets
Dead comets are small bodies that appear as inactive asteroids, but have a cometary origin. Some of these objects might still harbor ices and can still activate and appear comet-like. Since these objects might signifcantly contribute to the volatile reservoir of the asteroid population, it is important to understand how many of these objects there are.
Astronomy Don Quixote Dormant Comets Spitzer Space Telescope
Near-Earth asteroid Don Quixote has long since been considered a good candidate for an inactive comet nucleus due to its comet-like orbit. Since its discovery in 1983, comet-like has never been observed - until we observed this object with the Spitzer Space Telescope...
Spitzer Space Telescope
- SpitzerNEOs - Diameters and Albedos for 2132 Near-Earth Objects | 2020-02
- Physical Properties of two tiny Asteroids from Spitzer Observations | 2014-10
- Detection of Cometary Activity in NEO Don Quixote | 2014-10
Astronomy Near-Earth Asteroids Asteroid Physical Properties Asteroid Thermal Modeling Spitzer Space Telescope
Based on Spitzer Space Telescope observations, we derived diameter and albedo estimates for 2132 asteroids in near-Earth space - the largest dataset of its kind. Our results and data are available online.
Astronomy Near-Earth Asteroids Asteroid Physical Properties Spitzer Space Telescope
We observed two very small asteroids with the Spitzer Space Telescope to actually measure their physical properties for the first time. The constraints that we were able to place on their properties do not agree with the standard picture of small asteroids.
Astronomy Don Quixote Dormant Comets Spitzer Space Telescope
Near-Earth asteroid Don Quixote has long since been considered a good candidate for an inactive comet nucleus due to its comet-like orbit. Since its discovery in 1983, comet-like has never been observed - until we observed this object with the Spitzer Space Telescope...
Near-Earth Asteroids
- SpitzerNEOs - Diameters and Albedos for 2132 Near-Earth Objects | 2020-02
- How many Dead Comets are there? | 2015-08
- Rapid-Response Spectrophotometric Observations of NEOs with UKIRT and RATIR | 2014-11
- Physical Properties of two tiny Asteroids from Spitzer Observations | 2014-10
Astronomy Near-Earth Asteroids Asteroid Physical Properties Asteroid Thermal Modeling Spitzer Space Telescope
Based on Spitzer Space Telescope observations, we derived diameter and albedo estimates for 2132 asteroids in near-Earth space - the largest dataset of its kind. Our results and data are available online.
Astronomy Near-Earth Asteroids Dormant Comets
Dead comets are small bodies that appear as inactive asteroids, but have a cometary origin. Some of these objects might still harbor ices and can still activate and appear comet-like. Since these objects might signifcantly contribute to the volatile reservoir of the asteroid population, it is important to understand how many of these objects there are.
Astronomy Near-Earth Asteroids Rapid-Response Observations Spectrophotometry UKIRT
With rapid-response observations using UKIRT, we are able to observe newly discovered near-Earth asteroids when they are still close to Earth and thus bright. Our observations enable a probabilistic taxonomic classification of asteroids with typically small sizes and the determination of their compositional distribution.
Astronomy Near-Earth Asteroids Asteroid Physical Properties Spitzer Space Telescope
We observed two very small asteroids with the Spitzer Space Telescope to actually measure their physical properties for the first time. The constraints that we were able to place on their properties do not agree with the standard picture of small asteroids.
Rapid-Response Observations
- Rapid-Response Spectrophotometric Observations of NEOs with UKIRT and RATIR | 2014-11
Astronomy Near-Earth Asteroids Rapid-Response Observations Spectrophotometry UKIRT
With rapid-response observations using UKIRT, we are able to observe newly discovered near-Earth asteroids when they are still close to Earth and thus bright. Our observations enable a probabilistic taxonomic classification of asteroids with typically small sizes and the determination of their compositional distribution.
Spectrophotometry
- Rapid-Response Spectrophotometric Observations of NEOs with UKIRT and RATIR | 2014-11
Astronomy Near-Earth Asteroids Rapid-Response Observations Spectrophotometry UKIRT
With rapid-response observations using UKIRT, we are able to observe newly discovered near-Earth asteroids when they are still close to Earth and thus bright. Our observations enable a probabilistic taxonomic classification of asteroids with typically small sizes and the determination of their compositional distribution.
UKIRT
- Rapid-Response Spectrophotometric Observations of NEOs with UKIRT and RATIR | 2014-11
Astronomy Near-Earth Asteroids Rapid-Response Observations Spectrophotometry UKIRT
With rapid-response observations using UKIRT, we are able to observe newly discovered near-Earth asteroids when they are still close to Earth and thus bright. Our observations enable a probabilistic taxonomic classification of asteroids with typically small sizes and the determination of their compositional distribution.
Open-Source Software
- sbpy - A Python module for small-body planetary astronomy | 2018-07
- Photometrypipeline | 2017-02
Astronomy Open-Source Software
sbpy is an astropy affiliated package for small-body planetary astronomers. We proposed this idea to NASA and were funded for developing this Python module, which is outlined here.
Astronomy Open-Source Software Photometry Pipeline
photometrypipeline is a Python software package for automated image registration, calibration, and extraction of photometry tailored to the needs of asteroid observer. However, pp can also be applied to other imaging observations.
Photometry
- Photometrypipeline | 2017-02
Astronomy Open-Source Software Photometry Pipeline
photometrypipeline is a Python software package for automated image registration, calibration, and extraction of photometry tailored to the needs of asteroid observer. However, pp can also be applied to other imaging observations.
Pipeline
- Photometrypipeline | 2017-02
Astronomy Open-Source Software Photometry Pipeline
photometrypipeline is a Python software package for automated image registration, calibration, and extraction of photometry tailored to the needs of asteroid observer. However, pp can also be applied to other imaging observations.
Asteroid Thermal Modeling
- SpitzerNEOs - Diameters and Albedos for 2132 Near-Earth Objects | 2020-02
- Are there Limits on the Applicability of Asteroid Thermal Models for Near-Earth Asteroids? | 2019-01
Astronomy Near-Earth Asteroids Asteroid Physical Properties Asteroid Thermal Modeling Spitzer Space Telescope
Based on Spitzer Space Telescope observations, we derived diameter and albedo estimates for 2132 asteroids in near-Earth space - the largest dataset of its kind. Our results and data are available online.
Astronomy Asteroid Thermal Modeling Asteroid Thermophysical Modeling
There is a range of asteroid thermal models that are based on different physical assumptions. In this analysis, I investigate how these assumptions affect their results and applicability to different situations.
Asteroid Thermophysical Modeling
- Are there Limits on the Applicability of Asteroid Thermal Models for Near-Earth Asteroids? | 2019-01
Astronomy Asteroid Thermal Modeling Asteroid Thermophysical Modeling
There is a range of asteroid thermal models that are based on different physical assumptions. In this analysis, I investigate how these assumptions affect their results and applicability to different situations.
Asteroids
- Don Quixote and the Dormant Comets - My Astronomical Legacy | 2020-05
- A plethora of asteroids in TESS data | 2019-03
- Asteroid Shape Information from Gaia DR2 | 2019-01
Astronomy Don Quixote Dormant Comets Asteroids
Don Quixote has been considered asteroidal since its discovery in 1983, despite its comet-like orbit. We found activity in this object during its 2009 and 2018 apparitions, leading us to believe that it really is an active comet. But are there other objects like Don Quixote? I monitored a sample of ~100 asteroids that are somewhat likely to turn active, too, over a period of more than 4 years. Bottom line: Don Quixote is rather unique.
Astronomy Asteroids Asteroid Physical Properties TESS
TESS is a satellite that observes bright stars to find brightness modulations revealing exoplanet transits - but it also observes a lot of asteroids over a long period of time, which makes it a unique asset for deriving long asteroid rotation periods.
Astronomy Asteroids Asteroid Physical Properties Gaia
The Gaia mission builds a unique survey of the stars in our Milky Way - but it also observes asteroids that are crossing its field of view. We derive ensemble shape information for different asteroid populations from the first batch of asteroid data from Gaia.
Gaia
- Asteroid Shape Information from Gaia DR2 | 2019-01
Astronomy Asteroids Asteroid Physical Properties Gaia
The Gaia mission builds a unique survey of the stars in our Milky Way - but it also observes asteroids that are crossing its field of view. We derive ensemble shape information for different asteroid populations from the first batch of asteroid data from Gaia.
TESS
- A plethora of asteroids in TESS data | 2019-03
Astronomy Asteroids Asteroid Physical Properties TESS
TESS is a satellite that observes bright stars to find brightness modulations revealing exoplanet transits - but it also observes a lot of asteroids over a long period of time, which makes it a unique asset for deriving long asteroid rotation periods.
Machine Learning
- Automated Cloud Detection with Machine Learning | 2020-04
Astronomy Machine Learning Deep Learning Allsky-Camera
This is a toy project that turned into a real research project and a preparation for my new job as research scientist in computer vision: using machine learning techniques to identify clouds in all-sky camera images.
Deep Learning
- Traffic Noise Estimation from Satellite Imagery with Deep Learning | 2022-06
- Contrastive Self-Supervised Learning for Multi-modal Earth Observation Data | 2022-06
- Estimating Power Plant Greenhouse Gas Emissions from Satellite Imagery | 2021-12
- Estimation of Surface Level NO2 from Remote Sensing Data | 2021-11
- Commercial Vehicle Traffic Detection from Satellite Imagery with Deep Learning | 2021-11
- Power Plant Classification from Remote Imaging with Deep Learning | 2021-06
- Characterization of Industrial Smoke Plumes from Remote Sensing Data | 2020-12
- Automated Cloud Detection with Machine Learning | 2020-04
Applied Research Earth Observation Deep Learning Segmentation Traffic Noise Sentinel-2
Road traffic noise is a global issue that can lead to severe health effects. Despite the ubiquity of traffic noise, its quantification or estimation is complicated and detailed road traffic maps are only available for select countries or areas. We investigate whether it is possible to train a segmentation model to esimate road traffic noise from satellite imagery.
Fundamental Research Earth Observation Deep Learning Self-supervised Learnig Data Fusion Transformers Classification Segmentation
Self-supervised learning provides a powerful means to pretrain models based on un-labeled data. Un-labeled Earth observation data are abundant: this circumstance combined with the availability of multi modal data makes Earth observation a perfect playground for self-supervised learning. Our early results are very promising...
Applied Research Earth Observation Deep Learning Multi-task Learning Power Plants CO2 Sentinel-2
Tracking Greenhouse Gas emissions will be increasingly important in the future, as power generation from fossil fuels is supposed to fade out. Independent tools to monitor power plants are required for this task. We present a method that uses freely available satellite imagery to estimate power generation and CO2 emission rates on a global scale.
Applied Research Earth Observation Deep Learning Air Pollution NO2 Sentinel-2
Air pollution is a major health issue and often also contributes to climate change. Measuring air pollution is costly and therefore only available in some countries. We investigated whether freely available Earth observation data can be utilized to estimate air pollution on the surface level.
Applied Research Earth Observation Deep Learning Traffic Sentinel-2
Can we identify trucks from space and estimate truck traffic rates anywhere on the planet? Yes, we can!
Applied Research Earth Observation Deep Learning Classification Power Plants Sentinel-2
We developed a deep learning model that is able to distinguish between 10 different types of power plants in an effort to automatically identify and characterize industrial sites in satellite imagery. This work will help us to estimate greenhouse gas emission rates for individual industrial sites in the future.
Applied Research Earth Observation Deep Learning Segmentation Plumes Sentinel-2
Greenhouse gas emissions from the industrial economic sector are a major driver of the currently observed climate change. We developed a deep learning approach to identify and characterize industrial smoke plumes. In the future, we will utilize this approach to estimate greenhouse gas emissions from remote sensing data on a global scale.
Astronomy Machine Learning Deep Learning Allsky-Camera
This is a toy project that turned into a real research project and a preparation for my new job as research scientist in computer vision: using machine learning techniques to identify clouds in all-sky camera images.
Allsky-Camera
- Automated Cloud Detection with Machine Learning | 2020-04
Astronomy Machine Learning Deep Learning Allsky-Camera
This is a toy project that turned into a real research project and a preparation for my new job as research scientist in computer vision: using machine learning techniques to identify clouds in all-sky camera images.
Earth Observation
- Traffic Noise Estimation from Satellite Imagery with Deep Learning | 2022-06
- Contrastive Self-Supervised Learning for Multi-modal Earth Observation Data | 2022-06
- Estimating Power Plant Greenhouse Gas Emissions from Satellite Imagery | 2021-12
- Estimation of Surface Level NO2 from Remote Sensing Data | 2021-11
- Commercial Vehicle Traffic Detection from Satellite Imagery with Deep Learning | 2021-11
- Power Plant Classification from Remote Imaging with Deep Learning | 2021-06
- Characterization of Industrial Smoke Plumes from Remote Sensing Data | 2020-12
Applied Research Earth Observation Deep Learning Segmentation Traffic Noise Sentinel-2
Road traffic noise is a global issue that can lead to severe health effects. Despite the ubiquity of traffic noise, its quantification or estimation is complicated and detailed road traffic maps are only available for select countries or areas. We investigate whether it is possible to train a segmentation model to esimate road traffic noise from satellite imagery.
Fundamental Research Earth Observation Deep Learning Self-supervised Learnig Data Fusion Transformers Classification Segmentation
Self-supervised learning provides a powerful means to pretrain models based on un-labeled data. Un-labeled Earth observation data are abundant: this circumstance combined with the availability of multi modal data makes Earth observation a perfect playground for self-supervised learning. Our early results are very promising...
Applied Research Earth Observation Deep Learning Multi-task Learning Power Plants CO2 Sentinel-2
Tracking Greenhouse Gas emissions will be increasingly important in the future, as power generation from fossil fuels is supposed to fade out. Independent tools to monitor power plants are required for this task. We present a method that uses freely available satellite imagery to estimate power generation and CO2 emission rates on a global scale.
Applied Research Earth Observation Deep Learning Air Pollution NO2 Sentinel-2
Air pollution is a major health issue and often also contributes to climate change. Measuring air pollution is costly and therefore only available in some countries. We investigated whether freely available Earth observation data can be utilized to estimate air pollution on the surface level.
Applied Research Earth Observation Deep Learning Traffic Sentinel-2
Can we identify trucks from space and estimate truck traffic rates anywhere on the planet? Yes, we can!
Applied Research Earth Observation Deep Learning Classification Power Plants Sentinel-2
We developed a deep learning model that is able to distinguish between 10 different types of power plants in an effort to automatically identify and characterize industrial sites in satellite imagery. This work will help us to estimate greenhouse gas emission rates for individual industrial sites in the future.
Applied Research Earth Observation Deep Learning Segmentation Plumes Sentinel-2
Greenhouse gas emissions from the industrial economic sector are a major driver of the currently observed climate change. We developed a deep learning approach to identify and characterize industrial smoke plumes. In the future, we will utilize this approach to estimate greenhouse gas emissions from remote sensing data on a global scale.
Segmentation
- Traffic Noise Estimation from Satellite Imagery with Deep Learning | 2022-06
- Contrastive Self-Supervised Learning for Multi-modal Earth Observation Data | 2022-06
- Characterization of Industrial Smoke Plumes from Remote Sensing Data | 2020-12
Applied Research Earth Observation Deep Learning Segmentation Traffic Noise Sentinel-2
Road traffic noise is a global issue that can lead to severe health effects. Despite the ubiquity of traffic noise, its quantification or estimation is complicated and detailed road traffic maps are only available for select countries or areas. We investigate whether it is possible to train a segmentation model to esimate road traffic noise from satellite imagery.
Fundamental Research Earth Observation Deep Learning Self-supervised Learnig Data Fusion Transformers Classification Segmentation
Self-supervised learning provides a powerful means to pretrain models based on un-labeled data. Un-labeled Earth observation data are abundant: this circumstance combined with the availability of multi modal data makes Earth observation a perfect playground for self-supervised learning. Our early results are very promising...
Applied Research Earth Observation Deep Learning Segmentation Plumes Sentinel-2
Greenhouse gas emissions from the industrial economic sector are a major driver of the currently observed climate change. We developed a deep learning approach to identify and characterize industrial smoke plumes. In the future, we will utilize this approach to estimate greenhouse gas emissions from remote sensing data on a global scale.
Plumes
- Characterization of Industrial Smoke Plumes from Remote Sensing Data | 2020-12
Applied Research Earth Observation Deep Learning Segmentation Plumes Sentinel-2
Greenhouse gas emissions from the industrial economic sector are a major driver of the currently observed climate change. We developed a deep learning approach to identify and characterize industrial smoke plumes. In the future, we will utilize this approach to estimate greenhouse gas emissions from remote sensing data on a global scale.
Sentinel-2
- Traffic Noise Estimation from Satellite Imagery with Deep Learning | 2022-06
- Estimating Power Plant Greenhouse Gas Emissions from Satellite Imagery | 2021-12
- Estimation of Surface Level NO2 from Remote Sensing Data | 2021-11
- Commercial Vehicle Traffic Detection from Satellite Imagery with Deep Learning | 2021-11
- Power Plant Classification from Remote Imaging with Deep Learning | 2021-06
- Characterization of Industrial Smoke Plumes from Remote Sensing Data | 2020-12
Applied Research Earth Observation Deep Learning Segmentation Traffic Noise Sentinel-2
Road traffic noise is a global issue that can lead to severe health effects. Despite the ubiquity of traffic noise, its quantification or estimation is complicated and detailed road traffic maps are only available for select countries or areas. We investigate whether it is possible to train a segmentation model to esimate road traffic noise from satellite imagery.
Applied Research Earth Observation Deep Learning Multi-task Learning Power Plants CO2 Sentinel-2
Tracking Greenhouse Gas emissions will be increasingly important in the future, as power generation from fossil fuels is supposed to fade out. Independent tools to monitor power plants are required for this task. We present a method that uses freely available satellite imagery to estimate power generation and CO2 emission rates on a global scale.
Applied Research Earth Observation Deep Learning Air Pollution NO2 Sentinel-2
Air pollution is a major health issue and often also contributes to climate change. Measuring air pollution is costly and therefore only available in some countries. We investigated whether freely available Earth observation data can be utilized to estimate air pollution on the surface level.
Applied Research Earth Observation Deep Learning Traffic Sentinel-2
Can we identify trucks from space and estimate truck traffic rates anywhere on the planet? Yes, we can!
Applied Research Earth Observation Deep Learning Classification Power Plants Sentinel-2
We developed a deep learning model that is able to distinguish between 10 different types of power plants in an effort to automatically identify and characterize industrial sites in satellite imagery. This work will help us to estimate greenhouse gas emission rates for individual industrial sites in the future.
Applied Research Earth Observation Deep Learning Segmentation Plumes Sentinel-2
Greenhouse gas emissions from the industrial economic sector are a major driver of the currently observed climate change. We developed a deep learning approach to identify and characterize industrial smoke plumes. In the future, we will utilize this approach to estimate greenhouse gas emissions from remote sensing data on a global scale.
Classification
- Contrastive Self-Supervised Learning for Multi-modal Earth Observation Data | 2022-06
- Power Plant Classification from Remote Imaging with Deep Learning | 2021-06
Fundamental Research Earth Observation Deep Learning Self-supervised Learnig Data Fusion Transformers Classification Segmentation
Self-supervised learning provides a powerful means to pretrain models based on un-labeled data. Un-labeled Earth observation data are abundant: this circumstance combined with the availability of multi modal data makes Earth observation a perfect playground for self-supervised learning. Our early results are very promising...
Applied Research Earth Observation Deep Learning Classification Power Plants Sentinel-2
We developed a deep learning model that is able to distinguish between 10 different types of power plants in an effort to automatically identify and characterize industrial sites in satellite imagery. This work will help us to estimate greenhouse gas emission rates for individual industrial sites in the future.
Power Plants
- Estimating Power Plant Greenhouse Gas Emissions from Satellite Imagery | 2021-12
- Power Plant Classification from Remote Imaging with Deep Learning | 2021-06
Applied Research Earth Observation Deep Learning Multi-task Learning Power Plants CO2 Sentinel-2
Tracking Greenhouse Gas emissions will be increasingly important in the future, as power generation from fossil fuels is supposed to fade out. Independent tools to monitor power plants are required for this task. We present a method that uses freely available satellite imagery to estimate power generation and CO2 emission rates on a global scale.
Applied Research Earth Observation Deep Learning Classification Power Plants Sentinel-2
We developed a deep learning model that is able to distinguish between 10 different types of power plants in an effort to automatically identify and characterize industrial sites in satellite imagery. This work will help us to estimate greenhouse gas emission rates for individual industrial sites in the future.
Traffic
- Commercial Vehicle Traffic Detection from Satellite Imagery with Deep Learning | 2021-11
Applied Research Earth Observation Deep Learning Traffic Sentinel-2
Can we identify trucks from space and estimate truck traffic rates anywhere on the planet? Yes, we can!
Air Pollution
- Estimation of Surface Level NO2 from Remote Sensing Data | 2021-11
Applied Research Earth Observation Deep Learning Air Pollution NO2 Sentinel-2
Air pollution is a major health issue and often also contributes to climate change. Measuring air pollution is costly and therefore only available in some countries. We investigated whether freely available Earth observation data can be utilized to estimate air pollution on the surface level.
NO2
- Estimation of Surface Level NO2 from Remote Sensing Data | 2021-11
Applied Research Earth Observation Deep Learning Air Pollution NO2 Sentinel-2
Air pollution is a major health issue and often also contributes to climate change. Measuring air pollution is costly and therefore only available in some countries. We investigated whether freely available Earth observation data can be utilized to estimate air pollution on the surface level.
Multi-task Learning
- Estimating Power Plant Greenhouse Gas Emissions from Satellite Imagery | 2021-12
Applied Research Earth Observation Deep Learning Multi-task Learning Power Plants CO2 Sentinel-2
Tracking Greenhouse Gas emissions will be increasingly important in the future, as power generation from fossil fuels is supposed to fade out. Independent tools to monitor power plants are required for this task. We present a method that uses freely available satellite imagery to estimate power generation and CO2 emission rates on a global scale.
CO2
- Estimating Power Plant Greenhouse Gas Emissions from Satellite Imagery | 2021-12
Applied Research Earth Observation Deep Learning Multi-task Learning Power Plants CO2 Sentinel-2
Tracking Greenhouse Gas emissions will be increasingly important in the future, as power generation from fossil fuels is supposed to fade out. Independent tools to monitor power plants are required for this task. We present a method that uses freely available satellite imagery to estimate power generation and CO2 emission rates on a global scale.
Self-supervised Learnig
- Contrastive Self-Supervised Learning for Multi-modal Earth Observation Data | 2022-06
Fundamental Research Earth Observation Deep Learning Self-supervised Learnig Data Fusion Transformers Classification Segmentation
Self-supervised learning provides a powerful means to pretrain models based on un-labeled data. Un-labeled Earth observation data are abundant: this circumstance combined with the availability of multi modal data makes Earth observation a perfect playground for self-supervised learning. Our early results are very promising...
Data Fusion
- Contrastive Self-Supervised Learning for Multi-modal Earth Observation Data | 2022-06
Fundamental Research Earth Observation Deep Learning Self-supervised Learnig Data Fusion Transformers Classification Segmentation
Self-supervised learning provides a powerful means to pretrain models based on un-labeled data. Un-labeled Earth observation data are abundant: this circumstance combined with the availability of multi modal data makes Earth observation a perfect playground for self-supervised learning. Our early results are very promising...
Transformers
- Contrastive Self-Supervised Learning for Multi-modal Earth Observation Data | 2022-06
Fundamental Research Earth Observation Deep Learning Self-supervised Learnig Data Fusion Transformers Classification Segmentation
Self-supervised learning provides a powerful means to pretrain models based on un-labeled data. Un-labeled Earth observation data are abundant: this circumstance combined with the availability of multi modal data makes Earth observation a perfect playground for self-supervised learning. Our early results are very promising...
Traffic Noise
- Traffic Noise Estimation from Satellite Imagery with Deep Learning | 2022-06
Applied Research Earth Observation Deep Learning Segmentation Traffic Noise Sentinel-2
Road traffic noise is a global issue that can lead to severe health effects. Despite the ubiquity of traffic noise, its quantification or estimation is complicated and detailed road traffic maps are only available for select countries or areas. We investigate whether it is possible to train a segmentation model to esimate road traffic noise from satellite imagery.