Institut für Umweltphysik / Fernerkundung (IUP)

Übergeordnete Einrichtung

Fachbereich 01: Physik/Elektrotechnik (FB 01)

Country

City

Bremen

Gerade angezeigt 1 - 10 von 24

Some of the metrics are blocked by your
consent settings
Remote sensing measurements of methane in the Arctic
(2025-12-18)
Hachmeister, Jonas
;
Buschmann, Matthias
;
Boesch, Hartmut
;
Buschmann, Matthias
Methane (CH₄) is the second-largest contributor to global warming, and plays an important role in the global carbon cycle. Atmospheric CH₄ concentrations have more than doubled since industrialization, and more than half of the global CH₄ emissions originate from human-made sources such as agriculture, waste processing, and the fossil fuel industry. The Arctic has been warming faster than the rest of the Earth, leading to increased concern about potential CH₄ emissions from the Arctic's large permafrost regions, which store enough carbon to more than double the amount currently present in the atmosphere. Monitoring of atmospheric CH₄ is possible using remote sensing systems. These systems allow indirect quantification of atmospheric CH₄ by measuring (reflected) sunlight. In this thesis, ground-based remote sensing measurements from TCCON and space-based remote sensing measurements made by the TROPOMI onboard Sentinel-5P are examined, with a focus on the Arctic region. Both datasets were investigated regarding a range of potential Arctic-specific issues. For the WFMD data, issues with the cloud filter, especially over the Arctic Ocean, were identified, and it was shown that the use of outdated or inaccurate digital elevation model data in the WFMD retrieval led to significant biases. It was furthermore shown that airmass-dependent biases are present in TCCON XCH₄ during polar vortex conditions, and that these biases can be reduced by improving the CH₄ prior profiles. Both datasets were then compared at the four Arctic TCCON sites, with overall good agreement. Following the assessment of data quality, both WFMD and TCCON data were used to calculate growth rates. To achieve this, an approach based on dynamic linear models was developed that can handle the inhomogeneous data coverage in the Arctic. First, CH₄ growth rates were calculated for global data and for zonal bands using WFMD data. Subsequently, growth rates were also derived for four Arctic TCCON sites and compared to satellite-derived growth rates. High-latitude growth rates did not differ significantly from those in the mid-latitudes.
Dissertation
41 50
Some of the metrics are blocked by your
consent settings
Plume dispersal in the Arctic Ocean - the Aurora Site at Gakkel Ridge
(2025-05-16)
Mette, Jonathan
;
Walter, Maren
This thesis assesses the dispersal of the hydrothermal plume at the Aurora Vent Site in the Arctic Ocean, based on observational data and complemented by a numerical ocean model. The theoretical background of hydrothermal plumes in the Arctic is presented and the methods used are explained. This includes the processing of data acquired from a CTD probe, water samples and an oceanographic mooring with respect to the hydrographic setting and to the tracers for identification of the hydrothermal fluid. In addition a setup for simulating the plume dispersal with the Regional Ocean Modelling System is described. The observational results reveal a plume that ascends up to a height of 1200 m and spreads laterally to at least a distance of 2500 m, although a strong core is confined to a much smaller area. The plume dispersal is highly inhomogeneous for the different investigated tracers. This, as well as the small horizontal extent is explained by the presence of slow currents that are altered by tidal or inertial oscillations. A similar vertical extent can be obtained from the model simulation. However in regards to the currents and horizontal extent, the simulation shows large discrepancies compared to the observations, therefore suggestions for improving the model setup are presented.
text::thesis::master thesis
138 75
Some of the metrics are blocked by your
consent settings
Toward physically consistent, accurate and stable machine learning-based convection parameterizations for ICON
(2026-02-06)
Heuer, Helge
;
Eyring, Veronika
;
Gentine, Pierre
Earth system models (ESMs) are important tools to project climate change, yet continue to have persistent systematic errors due to the representation of subgrid-scale processes, most notably atmospheric convection—a key driver of large-scale circulations such as the Hadley and Walker cells, as well as weather patterns such as thunderstorms. These errors contribute considerably to uncertainties in climate projections. Traditional convection parameterizations rely on physical assumptions and empirically derived relationships that fail to capture the full complexity of convective processes. This thesis contributes to showing that machine learning (ML) offers breakthroughs, leveraging high-fidelity simulations to learn data-driven parameterizations that better represent subgrid-scale dynamics. However, translating high offline ML performance into stable, physically consistent, and transferable online implementations in ESMs has proven challenging, often due to issues related to causality, scale separation, distributional shifts, and process separation. This dissertation addresses these challenges through two complementary studies that advance the development, interpretation, and integration of ML-based convection parameterizations into the ICOsahedral Nonhydrostatic (ICON) model. The first study develops and benchmarks a suite of ML models, including deep learning and tree-based methods, trained on filtered and coarse-grained convective fluxes derived from storm-resolving ICON simulations over the tropical Atlantic. A filtering method to isolate convective contributions from other physical processes is meant to ensure that the ML models learn to represent deep convection. Offline, a U-Net architecture outperforms other models but exhibits non-causal dependencies on precipitating tracers, as revealed by explainable artificial intelligence (AI) analysis using SHapley Additive exPlanations (SHAP). Ablating these inputs yields a more physically interpretable and causally sound parameterization that demonstrates improvements in online stability, maintaining 180-day integrations in ICON while reducing biases in precipitation extremes compared to conventional schemes. However, a significant smoothing bias in the column water vapor distribution as well as biases in the mean temperature persist. Building on these insights, the second study presents a proof-of-concept for cross-model transferability and long-term integrability. A bidirectional long short-term memory model trained on the global ClimSim dataset, derived from superparameterized Energy Exascale Earth System Model-Multiscale Modeling Framework (E3SM-MMF) simulations, is transferred to the ICON-A atmosphere model. Several innovations ensure physical consistency, accuracy, and robustness: removal of radiative tendencies to isolate the convective signal, physics-informed and vertical consistency losses, confidence-guided mixing with a conventional scheme, and additive input noise during training to enhance extrapolation and stability. This hybrid AI-physics approach enables the stable multi-decadal (20-year) integration of an ML-based convection parameterization in ICON. Evaluating the resulting simulations against observations indicates improved precipitation statistics, including reduced root mean square error in the zonal mean, better spatial distribution, and improved precipitation extremes, as well as a more accurate spatial distribution of the near-surface temperature, relative to the reference ICON configuration. Furthermore, the developed scheme exhibits a physically interpretable regime behavior across column water vapor and stability metrics. However, the used training dataset has biases as well, e.g., the zonal mean precipitation does not match the observed climatology and conservation laws are not strictly enforced by the developed framework and would require a refined training dataset to do so. Moreover, the scheme is trained and evaluated at a relatively coarse horizontal resolution of ∼160 km; implementing it in the currently developed version of the hybrid ICON model, which has a horizontal resolution of ∼80 km, may require vertical interpolation and tuning. Together, these studies demonstrate that ML-based parameterizations can become stable, interpretable, and transferable components of next-generation climate models. They highlight the critical importance of physical consistency, learning causal relationships, and robust training practices in achieving reliable long-term simulations, thereby demonstrating that multi-decadal stable hybrid simulations are achievable, paving the way toward more accurate and trustworthy climate projections.
Dissertation
37 26
Some of the metrics are blocked by your
consent settings
Detecting and understanding extreme temperature events and heatwaves using machine-learning
(2025-09-26)
Paçal, Aytaç
;
Eyring, Veronika
;
Camps-valls Gustau
;
Eyring, Veronika
The intensification of extreme temperature events, particularly heatwaves, is among the most severe impacts of human-induced climate change. These events have profound implications for public health, agriculture, infrastructure, and natural systems. Accurate detection and understanding of such extremes are essential for adaptation and mitigation strategies. However, traditional methods for detecting extreme temperature events often rely on assumptions of unimodal distributions, which may inadequately capture the complex and evolving nature of extreme heat events in a warming world. This dissertation addresses these limitations by developing and applying unsupervised machine learning approaches to detect and understand extreme temperature events, with a focus on both statistical frequency and associated atmospheric dynamics. In the first part of the dissertation, I present a novel approach for detecting extreme temperature events using Gaussian Mixture Model (GMM), applied to global, daily near-surface maximum temperature data from both European Centre for Medium-Range Weather Forecast (ECMWF) reanalysis v5 (ERA5) and Coupled Model Intercomparison Project Phase 6 (CMIP6) simulations. By modeling temperature distributions as multimodal rather than unimodal, GMM provides a better fit for daily maximum temperature data, particularly in mid-latitude regions where cold and warm seasons create distinct modes in the temperature distribution. This method captures the differences in geographical regions and shows that the frequency of extreme events will be even higher than reported in previous studies. Globally, a 10-year extreme temperature event relative to 1985–2014 conditions will occur 13.6 times more often in the future under a Global Warming Levels (GWL) of 3.0 ◦C. The frequency increase can be even higher in tropical regions, such that 10-year extreme temperature events will occur almost twice a week. Additionally, the hot temperatures are increasing faster than cold temperatures in low latitudes, while the cold temperatures are increasing faster than the hot temperatures in high latitudes under different GWL. The smallest changes in temperature distribution can be found in tropical regions, where the annual temperature range is small. The second part of the dissertation shifts from univariate statistical modeling to the spatiotemporal analysis of multivariate heatwave dynamics using deep learning. A spatiotemporal Variational Autoencoder (VAE) was trained on year-round eleven-day heatwave samples from the ERA5 reanalysis dataset from 1941-1990 over the North Atlantic region, incorporating nine atmospheric variables including temperature, wind, humidity, cloud cover, radiation, and geopotential height. The VAE encodes each multivariate event into a compact latent space, which is then clustered using GMM to identify characteristic heatwave regimes. Then, the VAE was tested with heatwave samples from 2001-2022 to analyze atmospheric patterns before and during the Western European heatwaves. Notably, recent summer heatwaves form a distinct and previously unseen cluster, suggesting a shift in atmospheric circulation patterns in response to climate change. Composite anomaly maps reveal coherent pre-onset signatures across variables, indicating that the VAE can extract physically interpretable features from complex, high-dimensional data. Together, the two studies presented in this dissertation demonstrate the value of unsupervised learning in climate science. While GMM provide a flexible and interpretable statistical framework for quantifying changes in the frequency and distribution of extreme temperature events, the VAE offers a novel approach for understanding the underlying physical drivers of heatwaves. By applying these methods to observational reanalysis data and climate model outputs, this dissertation contributes to advancing our understanding of how extreme temperature events are evolving in a warming world. The findings not only reveal the increasing risk of high-impact extreme temperature events but also highlight the importance of combining statistical and physical perspectives to better characterize, monitor, and ultimately predict these events.
Dissertation
100 71
Some of the metrics are blocked by your
consent settings
Evaluation and application of TROPOMI data: cloud retrieval algorithms and global shipping NO2 detection
(2025-12-11)
Latsch, Miriam
;
Burrows, John P.
;
Wagner, Thomas
;
Richter, Andreas
Satellite-based atmospheric measurements provide an important data basis for monitoring air pollution over long periods. Using the DOAS technique, tropospheric columns of nitrogen dioxide (NO2), a key air pollutant, can be retrieved from these observations. The TROPOspheric Monitoring Instrument (TROPOMI), launched on Sentinel-5P in October 2017, provides daily global measurements, significantly enhancing the ability to detect small-scale emissions like shipping plumes due to its high signal-to-noise ratio and spatial resolution. Clouds have a substantial impact on satellite measurements of tropospheric trace gases in the UV, VIS, and NIR spectral ranges. Therefore, NO2 retrievals rely on information on cloud fraction and cloud height from satellite cloud products. In the first part of this thesis, the cloud parameters from different cloud retrieval algorithms for TROPOMI are compared. Overall, the cloud products show qualitative consistency in processor version 1.x and reasonable agreement in processor version 2.x, except for the VIIRS cloud fraction. Differences between the cloud retrievals are found especially for low cloud heights and small cloud fractions, i.e., clouds that are particularly relevant for tropospheric trace gas retrievals. Cloud fractions primarily differ over snow- and ice-covered pixels and scenes with sun glint, for which only MICRU includes an explicit treatment. All cloud parameters exhibit systematic issues related to across-track dependence, and the consistency among these parameters depends strongly on how the data is filtered. In summary, clear differences were found between the results of various algorithms, but these differences are reduced in the most recent versions of the cloud data. Shipping is an important source of NOx emissions worldwide, negatively affecting marine environments and human health. The second part of the thesis uses the TROPOMI tropospheric NO2 slant columns (tSCDs) to qualitatively identify global shipping routes. Preprocessing techniques, including iterative high-pass and Fourier filtering, markedly improve the detection of shipping lanes, revealing many previously undetectable routes. The analysis examines the impact of high-pass filter box sizes, demonstrating that smaller sizes enhance the visibility of narrow shipping features, whereas larger box sizes increase overall NO2 signals. Additionally, various flagging criteria are investigated that affect NO2 signal distribution, highlighting the critical importance of careful selection for accurate emission monitoring. Filtered TROPOMI NO2 tSCDs over oceans show a strong correlation with shipping activities, as confirmed by comparison with the CAMS-GLOB-SHIP inventory, and reveal unknown shipping routes. TROPOMI also effectively captures NO2 emissions from oil and gas platforms. Finally, the filtered TROPOMI tropospheric NO2 vertical columns (tVCDs) are compared with those from the CAMS global model. While both datasets show NO2 enhancements over global shipping lanes, the CAMS NO2 values are significantly larger than the TROPOMI measurements in the North Atlantic and strongly depend on the masking threshold in the high-pass filtering method in the South Atlantic.
Dissertation
43 47
Some of the metrics are blocked by your
consent settings
WF-DOAS V1.0 Validation Report
(2004-01)
Weber, Mark
;
Lamsal, L. N.
;
Coldewey-Egbers, Melanie
;
Bramstedt, Klaus
;
Vanicek, K.
This report summarizes the validation of the GOME total ozone derived from applying the weighting function DOAS (WF-DOAS) algorithm Version 1.0. This algorithm was developed as part of the GOTOCORD ESA project (see accompanying ATBD Report). The WF-DOAS results have been compared with selected ground-based measurements from the WOUDC (World Ozone and UV Radiation Data Centre) which collects total zone measurements from a global network of stations. Very few of these stations carry out simultaneous measurements by Brewer and Dobson spectrometers over an extended period (three years or more). Simultaneous Brewer and Dobson measurements from Hradec Kralove, Czech Republic (50.2◦N, 15.8◦E) and Hohenpeissenberg, Germany (47.8◦N, 11.0◦E) covering the period 1996-1999 have been compared with our GOME results and particular attention is paid to the differences between the two ground spectrometer types and GOME.
Bericht
Heft:Issue 1.2
245 111
Some of the metrics are blocked by your
consent settings
Relevance of warm air intrusions for Arctic satellite sea ice concentration time series
(2023)
Rostosky, Philip
;
Spreen, Gunnar
Winter warm air intrusions entering the Arctic region can strongly modify the microwave emission of the snow-covered sea ice system due to temperature-induced snow metamorphism and ice crust formations, e.g., after melt–refreeze events. Common microwave radiometer satellite sea ice concentration retrievals are based on empirical models using the snow-covered sea ice emissivity and thus can be influenced by strong warm air intrusions. Here, we carry out a long-term study analyzing 41 years of winter sea ice concentration observations from different algorithms to investigate the impact of warm air intrusions on the retrieved ice concentration. Our results show that three out of four algorithms underestimate the sea ice concentration during (and up to 10 d after) warm air intrusions which increase the 2 m air temperature (daily maximum) above − 5 ∘C. This can lead to sea ice area underestimations in the order of 104 to 105 km2. If the 2 m temperature during the warm air intrusions crosses − 2 ∘C, all algorithms are impacted. Our analysis shows that the strength of these strong warm air intrusions increased in recent years, especially in April. With a further climate change, such warm air intrusions are expected to occur more frequently and earlier in the season, and their influence on sea ice climate data records will become more important.
Wissenschaftlicher Artikel
Band:17
46 45
Some of the metrics are blocked by your
consent settings
Seasonal evolution and spatial variability of sea ice properties from multi-frequency electromagnetic induction sounding
(2026-01-29)
Neudert, Mara
;
Haas, Christian
;
Uhlemann, Sebastian
;
Haas, Christian
This thesis advances electromagnetic (EM) induction sounding methods for measuring sea ice thickness and internal layers, focusing on the sub-ice platelet layer (SIPL) and slush. The SIPL is a layer of loosely aggregated ice crystals beneath Antarctic sea ice, characterized by exceptionally high biological activity, substantial regional contributions to sea ice mass, and its role as an indicator of ice shelf--ocean interactions. In the first study, we analyzed approximately 1000 km of multi-frequency EM survey data collected within a single season on fast ice in Atka Bay, eastern Weddell Sea, with a GEM-2 towed by a snowmobile. This unprecedented dataset allowed us to resolve both SIPL thickness and ice plus snow thicknesses. Using an open-source inversion framework, detailed spatial maps of SIPL thickness and conductivity were produced. SIPL solid fractions were estimated from SIPL conductivities and ranged from 0.11 to 0.28. Calibration in a zero-conductivity environment minimized sensor drift and offsets. Validation against drill hole measurements showed that both EM-derived SIPL thickness and ice plus snow thickness were accurate within a few decimeters. Results revealed regional patterns of platelet ice accumulation and provided the first bay-wide SIPL map including ice shelf fringes, with modal thicknesses around 5 m, local minima around 2 m and previously uninvestigated local maxima reaching up to 9 m in the south-eastern bay. In the second study, we use the GEM-2 to investigate slush from sea water flooding snow-covered sea ice. In Antarctica, this promotes snow-ice formation, contributing up to 25 % of ice thickness in young first-year ice in the Weddell Sea, while in the Arctic, it is rarer but may become more common under climate change, additionally posing travel safety concerns for Arctic communities. We showed that multi-frequency EM sounding can resolve slush and ice plus snow thickness jointly, with the optimal frequency combination achieving mean absolute errors below 5 cm for slush up to 60 cm in inversions of modelled data. Field surveys in Qikiqtarjuaq, Nunavut, Canada, confirmed robustness under variable conditions for slush up to 20 cm. A reliable height-step calibration routine using a wooden ladder was established and calibration parameters remained stable over two weeks for a GEM-2 sensor of the latest generation. In Antarctica, the combination of platelet ice and surface slush measurements could enable more accurate assessments of sea ice mass balance changes. In the Arctic, this work represents an initial step towards developing an operational system that can support local communities in mapping hazardous areas. Because access to potentially hazardous ice, particularly thin or slush-covered areas, is restricted when using an EM sensor mounted on a sled pulled by a snowmobile, we tested a drone-based system to extend its operational reach. A GEM-2 combined with custom altitude and attitude monitoring was flown as suspended load or sled-towed by a drone over landfast sea ice near Qikiqtarjuaq, Nunavut, Canada. Drone hover tests caused high EM noise at a 4 m distance, decreasing below thresholds required for 5 cm thickness resolution at 7 m. An in-flight calibration was tested, and the resulting calibrated EM signals differed only slightly from those obtained with ladder-based calibration, confirming effective calibration without the use of a ladder. When the GEM-2 was used with the drone, the thickness profiles agreed well with drill hole measurements, with mean thickness differences below 10 cm and closely matching thickness distributions. Flight stability was high, with sensor roll and pitch standard deviations below 3° and maxima around ±10°, ensuring reliable sensor orientation. These results confirm that the drone-borne EM sensor can provide accurate ice thickness measurements while minimizing operator risk and enabling surveys over thin or hazardous ice. Together, these studies advance EM methods for mapping sub-ice platelet and slush layers and introduce innovative approaches for airborne thickness surveys. By providing high-resolution, field-validated tools, this work enhances our ability to monitor and understand the mass balance and structural complexity of polar sea ice and improve safety on sea ice.
Dissertation
47 42
Some of the metrics are blocked by your
consent settings
Turbulence and its Drivers in the Weddell Sea Bottom Water Gravity Current
(2026-06-30)
Pinner, Ole
;
Kanzow, Torsten
;
Walter, Maren
;
Kanzow, Torsten
;
Janout, Markus
This study investigates the role of wave-induced turbulence in the dynamics of the Weddell Sea Bottom Water gravity current. The current transports dense water from its formation sites on the shelf to the deep sea and is a crucial component of the Southern Ocean overturning circulation. The analysis is based on data from a mooring array deployed across the continental slope between January 2017 and January 2019, and vertical profiles of temperature and salinity measured on various ship expeditions on a transect along the array. Previous studies suggest that internal waves may play a crucial role in driving turbulence within gravity currents. However, this influence has until now not been quantitatively assessed. To quantify the contribution of internal waves to turbulence in this particular gravity current along the continental slope, I employ three independent methods for estimating dissipation rates. First, I use a Thorpe scale approach to compute total, process-independent dissipation rates from density inversions in density profiles. Second, I apply the finestructure parameterization to estimate wave-induced dissipation rates from vertical profiles of strain, calculated from temperature and salinity profiles. Third, I develop a new method to estimate wave-induced dissipation rates from moored velocity time series. For that, I estimate wave energy levels from kinetic energy spectra and deduce dissipation rates by applying a formulation that is at the heart of the finestructure parameterization. A direct comparison of the time-averaged results from both the wave energy level method and the established finestructure parameterization differ in most cases by less than a factor of 3, below the associated methodological uncertainty. Turbulence is highest at the shelf break and decreases towards the deep sea, in line with decreasing strength of wave-induced turbulence. I observe a two-layer structure of the gravity current, a strongly turbulent, about 60-80m thick bottom layer and an upper, more quiescent interfacial layer. In the interfacial layer, internal waves induce an important part of the dissipation rate and therefore drive entrainment of warmer upper water into the gravity current. A precise quantification of the contribution is complicated by large method uncertainties. A comparison with turbulence measurements up- and downstream of our study site indicates that the processes dominating turbulence generation may depend on the location along the Weddell Sea Bottom Water gravity current: on the shelf, trapped waves are most important; on the continental slope, breaking internal waves dominate; and in the basin, symmetric instability is likely the main driver of turbulence. In addition to the horizontal and vertical spatial variability of turbulence along the continental slope, I investigate its seasonal variability. Variations of the turbulence in the Southern Ocean over time are generally ill-described, as there are almost no publications of dissipation rate time series on scales longer than days. Moored measurements are sliced to yield time series of dissipation rates, utilizing the here newly developed method. The near-bottom turbulence is observed to have considerable temporal variability. The seasonal variability correlates thereby with surface stress, with a time lag in the order of days, even at the abyssal mooring sites. The short time period between a signal on the surface and in the deep makes it highly unlikely that barotropic mechanisms are responsible but rather that the surface and the seafloor are linked by wind-generated near-inertial waves. From this, the drivers of wave-induced turbulence can be distinguished. The spatial variability of wave-induced turbulence is determined by internal tides, while the temporal variability is determined by intermittent near-inertial waves, even at the abyssal base of the continental slope. However, the exact mechanisms are still unknown, especially regarding the deep propagation of near-inertial waves and their dissipation mechanisms above the sea floor. Notwithstanding, I hypothesize that near-inertial waves driving near-bottom turbulence are generally more important than their current representation in literature suggests.
Dissertation
31 10
Some of the metrics are blocked by your
consent settings
Solar activity during solar cycle 23 monitored by GOME
(ESTEC Conference Bureau, 1999)
Weber, Mark
The daily solar spectral measurements by the Global Ozone Monitoring Experiment (GOME) aboard ERS2 are well suited to track solar activity by looking at irra- diance changes in the center of selected Fraunhofer lines. The major modulation of the irradiance changes con- tain periodicity of about 27-28 days, corresponding to the mean solar rotation period, and eleven years due to solar activity associated with the 22-year magnetic cycle of the sun. The MgII doublet at 280nm and the CaII K line at around 395nm have been analysed to derive a proxy so- lar activity indicator time series from GOME starting in July 1995 near the end of the declining phase of solar cycle 22. GOME is currently observing the onset of solar cycle 23. An excellent correlation with solar observations provided by the UARS instrument SUSIM (Solar Ultravi- olet Spectral Irradiance Monitor) has been achieved after removing instrumental artefacts such as UV degradation and etaloning from the GOME level-1 spectra. An esti- mate of solar UV irradiance variability with a change in the GOME MgII index is given.
Konferenzbeitrag
Band:161
198 91

Institut für Umweltphysik / Fernerkundung (IUP)

Filter

Einstellungen

Sortieren nach

Ergebnisse pro Seite