RSS 1.0Posters
Recent Submissions
-
Improving SO2 emissions over the Middle East.(Copernicus GmbH, 2023-02-26) [Poster]The Middle East is one of the most polluted regions on Earth. Besides strong natural air pollution caused by frequent dust storms, anthropogenic emissions of SO2 from power and desalination plants significantly deteriorate air quality and, as a consequence, reduce life expectancy. Additionally, sulfate aerosol formed through the chemical oxidation of SO2 has an effect on climate and cloud formation. Therefore, accurate modeling of SO2 emissions is crucial, especially in such harsh conditions as the Middle East. In this work, we attempt to improve existing SO2 emissions using inversion modeling, a high-resolution regional WRF-Chem model, and satellite observations of SO2 columns available from OMI and TOMS instruments. Obtained SO2 emission dataset is planned to be open to the community.
-
Multiple effects contributed to the intensive shaking recorded in the 6 February 2023 Kahramanmaraş (Türkiye) earthquake sequence(Copernicus GmbH, 2023-02-26) [Poster]The Kahramanmaraş earthquake sequence caused strong shaking and extensive damage in central-south Türkiye and northwestern Syria, making them the deadliest earthquakes in the region for multiple centuries. The rupture of the first mainshock (M7.8) initiated just south of the East Anatolian Fault (EAF) and then ruptured bilaterally hundreds of km of the EAF, causing major stress changes in the region and triggering the second mainshock (M7.6) about 9 hours later. We mapped the surface ruptures of the two mainshocks using pixel-offset tracking of Sentinel-1 radar images and find them to be ~300 km and 100-150 km long. The distribution of aftershocks indicates that the fault ruptures may have been even longer at depth, or about ~350 km and ~170 km, respectively. The pixel-tracking results and finite-fault modeling of the spatially variable fault slip show up to 7 and 8 m of surface fault offsets at the two faults, respectively, and that fault slip was shallow in both events, mostly above 15 km. In addition, our back-projection analysis suggests the first mainshock ruptured from the hypocenter to the northeast towards the EAF (first ~15 sec), then continued along it to the northeast (until ~55 sec), and also to the southwest towards the Hatay province, later at high rupture speeds (until ~80 sec). Furthermore, strong motion recordings show PGA values up to 2g and are particularly severe in Hatay, where multiple stations show over 0.5g PGA values. Both events are characterized by abrupt rupture cessation, generating strong stopping phases that likely contributed to the observed high shaking levels. Together the results show that directivity effects, high rupture speed, strong stopping phases, and local site effects all contributed to the intensive shaking and damage in the Hatay province.
-
Strong ground motions due to directivity and site effects inflicted by the February 6 2023 earthquake doublet, along the East Anatolian Fault(Copernicus GmbH, 2023-02-26) [Poster]Two powerful earthquakes (magnitudes 7.8 and 7.6) struck south-central Türkiye on February 6, 2023, causing significant damage across an extensive area of at least ten provinces in Türkiye as well as in multiple cities in northwestern Syria, making them one of the deadliest earthquakes in Türkiye for multiple centuries. The first mainshock started close to the well-known East Anatolian Fault (EAF) and then rupturing more than 300 km of that fault, whereas the second large earthquake occurred nine hours later around 90 km north of the first mainshock, on an east-west trending fault. In this study, we analysed recorded strong ground motions from the two events to better understand the factors contributing to the devastation caused by the earthquakes. For this, we collected 250 and 200 strong ground motion records for the first and the second event, respectively, from the Disaster and Emergency Management Authority (AFAD) in Türkiye. Maximum peak ground accelerations (PGA) of 2g were observed at a distance of 31 km northeast of the first mainshock epicenter and 0.6g for the second event 65km west to its epicenter. In addition, we find particularly high amplitude ground motions in the Hatay province for the first event, which is consistent with the extent of damage reported in that region. High shaking levels in Antakya and other parts of Hatay can be explained by a combination of strong directivity and local site effects. The results of our analysis imply that the PGA values derived from two local ground motion models (GMMs), adopted for the 2018 Turkish hazard map, are underestimated in comparison to observed strong motion recordings. In addition, we also compared observed peak and spectral ground motion characteristics with estimated seismic hazard values (10% probability to exceed in 50 years) in the East Anatolian Fault region (extracted from the 2018 Turkish seismic hazard map). Furthermore, we compare the recorded response spectra with the Turkish design code for several locations around the main faults. The results show that the observations greatly exceed the hazard values and code guidelines in the Hatay province.
-
The first network of Ocean Bottom Seismometers in the Red Sea to investigate the Zabargad Fracture Zone(Copernicus GmbH, 2023-02-26) [Poster]The Red Sea (RS) is an ideal natural laboratory to study the transition from continental rifting to seafloor spreading because it is one of the youngest rift basins on Earth. The RS is an ultra-slow spreading rift with an opening rate that decreases from 15 mm/yr in the south, at the Nubian-Arabian-Danakil triple junction, to 7 mm/yr in the north where it connects to the Dead Sea transform fault. While the southern RS has a well-developed seafloor spreading ridge with an axis parallel to the rift bounding faults, images of the northern RS seafloor provide limited information because of thick evaporite layers covering the main tectonic features. Nevertheless, the northern rift axis is geometrically different as it is more oblique to the bounding faults than in the south. Furthermore, the transition between the southern and northern RS is sharp with a ~100 km rift axis offset, named Zabargad Fracture Zone (ZFZ). However, the current knowledge of the seismic activity, transform fault configuration, and the crustal and upper mantle structure of the ZFZ area is too limited to assess the seismic hazard associated with this rift offset and understand the role of the ZFZ in the RS development. To fill this gap, we deployed the first broadband seismic network in the Red Sea, within the ZFZ, from November 2021 to November 2022. This network included 12 Lobster OBSs from the DEPAS pool (equipped with Güralp CMG-40T-OBS sensors), 2 additional Trillium Compact sensors deployed directly on the seafloor mounted on minimalist frames through a collaboration with the company Fugro, as well as 2 Trillium Compact Horizon and 2 posthole sensors deployed on islands and inland in Saudi Arabia, respectively. The overall data recovery rate is above 90%. Also, our preliminary data analysis confirms some of the known issues of the Lobster OBSs and their sensors (strong self-noise at periods >10 s of the Güralp sensors and high-frequency harmonic noise due to head-buoy cable strumming). Furthermore, we conducted a systematic comparison of the noise recorded by different station configurations. We find that while Lobster OBSs with a head-buoy cable set free to strum generate strong noise at about 10 Hz and its overtones, Lobster OBSs with tight cable still display harmonic noise from 10 to 40 Hz that increases during bad weather conditions, probably due to resonance of other OBS elements. The Fugro OBSs, despite their minimalist deployment setup, show noise at 40 Hz that also resonates in bad weather. All the OBSs also display a peak of noise between 0.5 and 5 Hz (separated from the secondary microseismic peak). While such a noise peak is not recorded by the inland stations, it is well exhibited by the two island stations suggesting that this is due to locally ocean-generated seismic waves, but not by the OBS frames. At periods >10 s, the Fugro OBSs perform as well as the island and inland stations. In fact, waveforms of teleseismic earthquakes recorded by the Fugro OBSs and island and inland stations have comparable signal-to-noise ratios.
-
The effect of ash, water vapor, and heterogeneous chemistry on the evolution of a Pinatubo-size volcanic cloud(Copernicus GmbH, 2023-02-26) [Poster]We employ the ECHAM5/MESSy2 atmospheric chemistry general circulation model (EMAC)that incorporates calculations of gas-phase and heterogeneous chemistry coupled with the ozone cycle and aerosol formation, transport, and microphysics to calculate the 1991 Pinatubo volcanic cloud. We considered simultaneous injections of SO2, volcanic ash, and water vapor. We conducted multiple ensemble simulations with different injection configurations to test the evolution of SO2, SO4, ash masses, stratospheric aerosol optical depth, surface area density (SAD), and 24the stratospheric temperature response against available observations. We found that the volcanic cloud evolution is sensitive to the altitude where volcanic debris is initially injected and the initial concentrations of the eruption products that affect radiative heating and lofting of the volcanic cloud. The numerical experiments with the injection of 12 Mt SO2, 75 Mt of volcanic ash, and 150 Mt of water vapor at 20 km show the best agreement with the observation of aerosol optical depth and stratospheric temperature response. Volcanic water injected by eruptive jets and/or intruding through the tropopause accelerates SO2 oxidation. But the mass of volcanic water retained in the stratosphere is controlled by the stratospheric temperature at the injection level. For example, volcanic materials are released in the cold point above the tropical tropopause, and most of the injected water freezes and sediments as ice crystals. The water vapor directly injected into the volcanic cloud increases the SO2− mass and stratospheric aerosol optical depth by about 5%. The coarse 4ash comprises 98% of the ash injected mass. It sediments within a few days, but aged submicron ash could stay in the stratosphere for a few months providing SAD for heterogeneous chemistry. The presence of ash accelerates SO2 oxidation by 10%–20% due to heterogeneous chemistry, radiative heating, lofting, and faster dispersion of volcanic debris. Ash aging affects its lifetime and optical properties, almost doubling the radiative ash heating. The 2.5-year simulations show that the stratospheric temperature anomalies forced by radiative heating of volcanic debris in our experiments with the 20 km injection height agree well with observations and reanalysis data. This indicates that the model captures the long-term evolution and climate effect of the Pinatubo volcanic cloud. The volcanic cloud’s initial lofting, facilitated by ash particles’ radiative heating, controls the oxidation rate of SO2. Ash accelerates the formation of the sulfate layer in the first 2 months after the eruption. We also found that the interactive calculations of OH and heterogeneous chemistry increase the volcanic cloud sensitivity to water vapor and ash injections. All those factors must be accounted for in modeling the impact of large-scale volcanic injections on climate and stratospheric chemistry.
-
Future projection of the African easterly waves in a high-resolution AGCM(Copernicus GmbH, 2023-02-22) [Poster]African Easterly Waves (AEWs) are a significant control of West African rainfall and the associated Mesoscale Convective Systems (MCSs) and squall lines embedded within them. More than 40% of the total MCSs over the region are associated with AEWs and these MCSs account for approximately 80% of the total annual rainfall over the Sahel. Approximately 60% of all Atlantic hurricanes and 80% of major hurricanes have their genesis associated with AEWs. Simulating the features of AEWs, such as their westward propagation off the east Atlantic coast, is challenging for coarse-resolution climate models. In this study, we use High-Resolution Atmospheric Model (HiRAM) to simulate AEWs and analyze their future projections by the end of the 21st century. The simulations are performed globally at a horizontal resolution of 25km. The model uses shallow convective parameterization for moist convection and stratiform cloudiness. Future projections are conducted using representative concentration pathway 8.5. AEWs are separated with respect to their periods as 3–5 and 6–9-day period AEWs, and bandpass filtering is used to filter the waves from the mean flow. HiRAM simulates the structure and propagation of the waves well; however, it tends to overestimate the associated precipitation. In the future, the AEW precipitation and intensity of the circulation will considerably increase. The northward extent of the AEW track also shows a significant increase in the future. Enhanced baroclinic overturning and eddy available potential energy generated due to diabatic heating is also observed in the future.
-
Impact of Forestation and Land-use Changes on Desert Climate(Copernicus GmbH, 2023-02-22) [Poster]Growing forests is an effective way of removing CO2 from the atmosphere. Forestation projects were started in China, Germany, and the Middle East. Saudi Arabia announced its ambitious “Saudi Green Initiative,” intending to plant ten billion trees. Given the insufficient rainfall to support the initiative, vegetated areas will require irrigation, effectively increasing evaporation. In addition, those areas have a lower albedo than bare land, absorbing more solar radiation. Enhancing precipitation due to the recycling of evaporated water is important as it reduces the amount of freshwater required for irrigation. In this study, we focus on the regional climate impact of irrigated forested or vegetated areas on temperature and precipitation over the Arabian Peninsula to quantify their effect on livability and evaluate the water recycling potential. First, we studied the climate effect of irrigated farming developing over vast areas in Saudi Arabia since the 1980s. The agricultural areas were mapped using available satellite-based observations from the Landsat platforms, which capture optical and thermal data every 16 days at a resolution of 30 m to 100 m. Second, we projected the climate impact of widespread forestation over the Arabian Peninsula. The analysis of the long-term precipitation changes caused by irrigated farming is hindered by the lack of in situ observations and the limitations of global-scale observation data sets. Most reanalysis products have contradictory evaporation trends and indicate an overall reduction in rainfall since the 1980s. The recycled precipitation cannot be estimated reliably because of reanalysis increments and background rainfall variability. Presumably, the local increase in rains occurs downstream of the irrigated areas rather than over them. Along with the analysis of observations, we conducted numerical experiments mimicking the effect of irrigated agricultural fields using a non-hydrostatic regional meteorological model (WRF), covering the whole Arabian Peninsula by a 9x9 km2 grid, with 3x3 km2 nesting over the irrigated areas. Irrigation water is accounted for by tagging moisture evaporated from agricultural regions. The amount of tagged water vapor falling as rain represents recycled precipitation. The simulated evaporation and local temperature response strongly depends on the level of irrigation. Large-scale subsidence suppresses the local deep convection over most parts of the Arabian Peninsula. Strong turbulence quickly mixes evaporated water vapor within a six km thick atmospheric boundary layer, preventing precipitation in shallow convection so that the fraction of recycled rainfall appears to be low.
-
Intracellular “in silico microscopes” – fully 3D spatial Hepatitis C virus replication model simulations(Georg Thieme Verlag, 2023-01-18) [Poster]Virus pandemics and endemics cause enormous pain and economic, political, and social costs and turmoil. While the Covid19 pandemics induced obvious damages, the "silent" Hepatitis C virus (HCV) infection induced liver damages are the main reason for liver transplantations. HCV-generated virus genome replication factories are housed within virus-induced intracellular structures termed membranous webs (MW) which are derived from the Endoplasmatic Reticulum (ER). Up to now, very advanced experimental data such as highly spatially resolved fluorescence and electron-tomography data often do not enter computational HCV viral RNA (vRNA) cycle models. Based upon diffusion-reaction partial differential equation (PDE) models, we are developing fully 3D resolved “in silico microscopes” to mirror in vitro / in vivo experiments of the intracellular vRNA cycle dynamics. Our first models described the major components (vRNA, non-structural viral proteins - NSPs - and a host factor). The next steps incorporated additional parameters: Different aggregate states of vRNA and NSPs, and population dynamics inspired diffusion and reaction coefficients instead of multilinear ones. Our work in progress framework presently is merging effects restricted to 2D manifold surface grids (e.g. ER surface, NSP diffusion) with others occurring in 3D volume meshes (e.g. cytosol, host factor supply). We estimate and incorporate realistic parameters such as NSP diffusion constants. The simulations are performed upon experimental data based reconstructed cell geometries and help understanding the relation of form and function of virus replication. In the long run, our framework might help to facilitate the systematic development of efficient direct antiviral agents and vaccines.
-
MoStGAN: Video Generation with Temporal Motion Styles(2022-12-06) [Poster]Video generation remains a challenging task due to spatiotemporal complexity and the requirement of synthesizing diverse motions with temporal consistency. Previous works attempt to generate videos in arbitrary lengths either in an autoregressive manner or regarding time as a continuous signal. However, they struggle to synthesize detailed and diverse motions with temporal coherence and tend to generate repetitive scenes after a few time steps. In this work, we argue that a single time-agnostic latent vector of style-based generator is insufficient to model various and temporally-consistent motions. Hence, we introduce additional time-dependent motion styles to model diverse motion patterns. In addition, a extbf{Mo}tion extbf{St}yle extbf{Att}ention modulation mechanism, dubbed as MoStAtt, is proposed to augment frames with vivid dynamics for each specific scale (i.e., layer), which assigns attention score for each motion style w.r.t deconvolution filter weights in the target synthesis layer and softly attends different motion styles for weight modulation. Experimental results show our model achieves state-of-the-art performance on four unconditional $256^2$ video synthesis benchmarks trained with only 3 frames per clip and produces better qualitative results with respect to dynamic motions.
-
Learning The Rules of Minihack Environment Using DreamerV2(2022-12-06) [Poster]DreamerV2 is proven to be able to simulate a visually-dominated environments such as Atari games. However, is it able to simulate environments with hidden rules that are not visual? In this paper I will answer if the world-model of dreamerv2 is capable of learning non-visual rules of the environment minihack.
-
The Anatomy of Rainy Sound(2022-11-22) [Poster]
-
2D Drop Splash(2022-11-22) [Poster]
-
Machine learning-assisted CO2 Storage Capacity Prediction in Deep Saline Aquifers: Uncertainty and Global Sensitivity Analysis(2022-11-15) [Poster]Geological CO2 sequestration (GCS) has been a practical approach used to mitigate global climate change. Uncertainty and sensitivity analysis of CO2 storage capacity prediction are essential aspects for large-scale CO2 sequestration. This work presents a rigorous machine learning-assisted (ML) workflow for the uncertainty and global sensitivity analysis of CO2 storage capacity prediction in deep saline aquifers. The proposed workflow comprises three main steps: 1) dataset generation we first identify the uncertainty parameters that impact CO2 storage in deep saline aquifers and then determine their corresponding ranges and distributions. We generate the required data samples by combining the Latin Hypercube Sampling (LHS) technique with high-resolution simulations. 2) ML model development a data-driven ML model is developed to map the nonlinear relationship between the input parameters and corresponding output interests from the previous step. The implementation of Bayesian optimization accelerates the tunning process of hyper-parameters instead of traditional trial-error analysis. 3) uncertainty and global sensitivity analysis Monte Carlo simulations based on the optimized surrogate are performed to explore the time-dependent uncertainty propagation of model outputs. Then the key contributors are identified by calculating the Sobol indices based on the global sensitivity analysis. The proposed workflow is accurate and efficient and could be readily implemented in field-scale CO2 sequestration in deep saline aquifers.
-
Generative Adversarial Zero-Shot Learning For Cold-start News Recommendation(2022-11-15) [Poster]News recommendation models extremely rely on the interactive information between users and news articles to personalize the recommendation. Therefore, one of their most serious challenges is the cold-start problem (CSP). Their performance is dropped intensely for new users or new news. Zero-shot learning helps in synthesizing a virtual representation of the missing data in a variety of application tasks. Therefore, it can be a promising solution for CSP to generate virtual interaction behaviors for new users or new news articles. In this work, we utilize the generative adversarial zero-shot learning in building a framework, namely, GAZRec, which is able to address the CSP caused by purely new users or new news. GAZRec can be flexibly applied to any neural news recommendation model. According to the experimental evaluations, applying the proposed framework to various news recommendation baselines attains a significant AUC improvement of 1% - 21% in different cold start scenarios and 1.2% - 6.6% in the regular situation when both users and news have a few interactions.
-
Deep learning-based regularization of post-stack seismic inversion(2022-11-15) [Poster]Seismic inversion is the prime method to estimate subsurface properties from seismic data. However, such inversion is a notoriously ill-posed inverse problem due to the band-limited and noisy nature of the data. Consequently, the data misfit term must be augmented with appropriate regularization that incorporates prior information about the sought-after solution. Conventionally, model-based regularization terms are problem-dependent and hand-crafted; this can limit the modeling capability of the inverse problem. Recently, a new framework has emerged under the name of Plug-and-Play (PnP) regularization, which suggests reinterpreting the effect of the regularizer as a denoising problem. Convolutional neural networks-based denoisers are state-of-the-art methods for image denoising: their adoption in the PnP framework has led to algorithms with improved capabilities over classical regularization in computer vision and medical imaging applications. In this work, we present a comparison between standard model-based and data-driven regularization techniques in post-stack seismic inversion and give some insights into the optimization and denoiser-related parameters tuning. The results on synthetic seismic data indicate that PnP regularization using a bias-free CNN-based denoiser with an additional noise map as input can outperform standard model-based methods.
-
GANs for 3D Porous media generation(2022-11-15) [Poster]Linking the fluid flow at the pore scale and reservoir scale is an active area of research in projects related to CO2 storage and oil and gas recovery. A key obstacle to understanding such a process is the lack of physical samples from relevant geological areas. This issue can be addressed by generating accurate, digital representations of the rock samples available for numerical fluid flow simulations. A new promising avenue for generating realistic digital rock samples is opening up because of recent advancements in Machine Learning and Deep Generative Modeling. In particular, Generative Adversarial Networks (GANs) can learn complex distributions with high dimensions and produce high-quality samples. This study presents a Wasserstein GAN with gradient penalty (WGAN-GP) to generate high-quality porous media samples in 3D. Additionally, an evaluation metric set inspired by geometry, topology, and fluid flow properties is established to assess the generative quality.
-
Tensor PDE model for biological network(2022-11-15) [Poster]
-
A Sequential Discontinuous Galerkin Scheme for Two-phase Poroelasticity(2022-11-15) [Poster]We formulate a numerical method for solving the two-phase flow poroelasticity equations. The scheme employs the interior penalty discontinuous Galerkin method and a sequential time-stepping method. The unknowns are the phase pressures and the displacement. Existence of the solution is proved. Three-dimensional numerical results show the accuracy and robustness of the proposed method.
-
Privacy in Bioinformatics and Intelligent Healthcare(2022-11-10) [Poster]
