Events
Mon | Tue | Wed | Thu | Fri | Sat | Sun |
---|---|---|---|---|---|---|
1 | ||||||
2 | 3 | 4 | 5 | 6 | 7 | 8 |
9 | 10 | 11 | 12 | 13 | 14 | 15 |
16 | 17 | 18 | 19 | 20 | 21 | 22 |
23 | 24 | 25 | 26 | 27 | 28 | 29 |
30 | 31 |
Legend | |||||||||||
JSG | BEG | UTIG | EPS |
DeFord Lecture: P. David Polly
Start:December 5, 2019 at 3:45 pm
End:
December 5, 2019 at 5:00 pm
Location:
JGB 2.324
View Event
UTIG Seminar: Graduate Student Talks
Start:December 6, 2019 at 10:30 am
End:
December 6, 2019 at 12:00 pm
Location:
PRC ROC Room 1.603
Contact:
Constantino Panagopulos, costa@ig.utexas.edu, 512-574-7376
View Event
In the final UTIG seminar of 2019, three UTIG grad students will present a short overview of their research and recent findings.
Host: Steve Phillips
Speakers (click name for bio):
Janaki Vamaraju
Title: Mini-batch Least-Squares Reverse Time Migration In A Deep Learning Framework
Abstract: Migration techniques have been an integral part of seismic and electromagnetic imaging workflows. They can be used to move reflection points to their correct positions beneath the surface. Least-squares reverse time migration (LSRTM) overcomes shortcomings of conventional migration algorithms by compensating illumination and removing sampling artifacts to increase spatial resolution. However, the computational cost associated with iterative LSRTM is high and can cause slow convergence in complex media. We implement LSRTM in a deep learning framework and adopt strategies from the data science domain to accelerate convergence. As a use case, we solve the problem of pres-stack seismic LSRTM. The proposed hybrid framework leverages the existing physics-based models and machine learning optimizers to achieve more accurate and cheaper solutions. Using a time-domain formulation, we show that mini-batch gradients can reduce the computation cost by using a subset of total shots for each iteration. Mini-batches not only reduce source cross-talk but are also less memory intensive. Combining mini-batch gradients with deep learning optimizers and loss functions can improve the efficiency of LSRTM. Deep learning optimizers such as the adaptive moment estimation are well suited for noisy and sparse data. We compare different optimizers and demonstrate their efficacy in mitigating migration artifacts. Regularized loss functions such as the Huber loss is used in conjunction to accelerate the inversion. We apply these techniques to 2D Marmousi and SEG/EAGE salt models and show improvements over conventional LSRTM baselines. The proposed approach achieves higher spatial resolution in lesser time according to various qualitative and quantitative evaluation metrics.
Xian Wu
Title: Predictability of El Niño Duration Based on the Onset Timing
Abstract: El Niño causes episodic sea surface temperature warming of the equatorial Pacific Ocean and affects global weather patterns through atmospheric teleconnections. El Niño events typically last one year, but about one-third of El Niño events last for a second year, which could prolong and exacerbate their climate impacts. Analysis of observational and model data shows that El Niño events developing in boreal spring-summer usually terminate after peaking in winter, while those developing after summer tend to persist into the second year. To test the predictability of El Niño duration based on the onset timing, perfect model forecasts were conducted with the Community Earth System Model version 1 (CESM1), a model that reproduces the observed dependence of El Niño duration on the onset timing. We select three El Niño events developing in April or September from the CESM1 control simulation. For each event, 30-member ensemble forecasts are initialized with the same oceanic conditions in the onset month but with slightly different atmospheric conditions and integrated for two years. The CESM1 successfully predicts the termination of El Niño after the peak in 95% of the April-initialized forecasts and the continuation of El Niño into the second year in 83% of the September-initialized forecasts. The predictability of El Niño duration arises from the initial oceanic conditions that affect the timing and magnitude of negative feedback within the equatorial Pacific, as well as from the Indian and Atlantic Oceans. The forecast spread of El Niño duration, on the other hand, originates from surface wind variability over the western equatorial Pacific in spring following the peak. These results indicate potential predictability of El Niño events beyond the current operational El Niño forecasts by one year.
Kelly Olsen
Title: Strong sediment inputs at the south-central Chile margin revealed from fault patterns in 2D seismic reflection data and the relationship to megathrust earthquakes
Abstract: South-central Chile is the location of the largest (1960 Mw 9.5) and sixth-largest (2010 Mw 8.8) recorded earthquakes. 2D seismic reflection data collected in 2017 as part of CEVICHE (Crustal Experiment from Valdivia to Illapel to Characterize Huge Earthquake) show that in the region of these ruptures, thick (1.5-3 km) trench sediments are being subducting beneath the frontal wedge at the deformation front. The P-wave seismic velocity (Vp) of the incoming sediment is relatively high and thus the incoming sediment may be stronger than at other margins. We test this hypothesis by using structural analysis of conjugate fault pairs to determine the coefficient of internal friction for the incoming and slope sediments and compare these values to other subduction zones. Based on fault orientations, the average coefficient of internal friction for the incoming trench sediment is ~0.82 (conjugate normal faults dip ~62-70°), while the average frictional coefficient for the slope apron is ~0.55 (conjugate normal faults dip ~53-64°). The coefficient of internal friction for the slope sediment is consistent with values measured in the slope of other subduction zones, such as Costa Rica and Nankai. The coefficients for the incoming sediment sections showed much higher values than both the slope apron in Chile, and incoming sediment sections of other subduction zones, such as Nankai, Costa Rica, Cascadia, and Sunda. Along the Nankai drilling transect, the only samples where sediments have coefficients of friction comparable to those in the Chile trench were recovered from the off-scraped sediment at the toe of the accretionary wedge, which is likely more consolidated than trench sediment in Nankai. These results support the inference from the Vp data that the sediments entering the south-central Chile subduction zone via the trench are inherently stronger than in other subduction settings. The higher strength could allow them to develop higher basal friction and ultimately contribute to the strong interplate locking inferred along this entire segment of the Chile margin. This study highlights the importance of the physical properties of trench sediment for controlling overall development of the subduction interface and may explain why the south-central Chile margin generates such large earthquakes.
Alumni Reception during AGU in San Francisco
Start:December 11, 2019 at 12:00 pm
End:
December 11, 2019 at 2:00 pm
Location:
ThirstyBear Brewing Company, 661 Howard St, San Francisco, CA 94105
Contact:
Kristen Tucek, ktucek@jsg.utexas.edu, 512-471-2223
UTIG Seminar Series: Tanner Mills, UTIGApril, 19 2024Time: 10:30 AM - 11:30 AMLocation: PRC 196/ROC 1.603 Speaker: Tanner Mills, Postdoctoral Fellow, University of Texas Institute for Geophysics Host: Peter Flemings Title: Predicting greenhouse gas fluxes to the atmosphere from thawing permafrost Abstract: Arctic permafrost is thawing at rapid rates, which threatens to expose large stores of soil organic carbon to microbial degradation. As microbes utilize this carbon source, they produce greenhouse gasses (GHGs; CO2 and CH4) that can be emitted to the atmosphere and act as a positive feedback during future global temperature increases. While the permafrost carbon feedback has received much attention in the literature, little is known about the multiphase flow properties and the temperature dependence of microbial GHG production rates in thawing permafrost, both of which are essential for predicting GHG emissions from permafrost in the future. Flow experiments of synthetic and natural permafrost specimens under frozen conditions and incubations of permafrost samples are being performed to better understand the effective and relative permeabilities and GHG production rates of thawing permafrost soils. These data will be integral in providing new source terms for permafrost and global carbon models. |
Master\'s Thesis PresentationsApril, 19 2024Time: 2:00 PM - 5:00 PMLocation: JGB The Master of Science (MS) degree at the Jackson School of Geosciences is considered to be the professional degree for a career in the Geosciences. This degree is the foundation for students pursuing employment in the petroleum industry, environmental and hydrogeological fields, state and federal government agencies, and other related geoscience fields. Some students also use the MS degree as preparation for pursuing a Ph.D. The Energy & Earth Resources Interdisciplinary program provides the opportunity for students to prepare themselves in management, finance, economics, law and policy leading to analytical and leadership positions in resource-related fields. The private sector and government organizations face a growing need for professionals that can plan, evaluate, and manage complex resource projects, commonly international in scope, which often include partners with a variety of professional backgrounds. As requirements for these degrees, students must present a professional talk on Master\'s Thesis Presentations. |
Hot Science - Cool Talks: \"Humans vs AI\"April, 19 2024Time: 5:30 PM - 8:15 PMLocation: Burdine 106 Advancements in AI have unleashed astonishing capabilities, but it is not magic. Peter Stone reveals his insights into cutting-edge AI and robotics and explores how they may reshape our world. Someday these technologies could win the World Cup, and they are already outperforming the best humans at complex tasks like high-speed racing. |
Planetary Habitability Seminar SeriesApril, 22 2024Time: 1:00 PM - 2:00 PMLocation: PMA 15.216B UT Center for Planetary Systems Habitability Seminar Series. See website for speaker schedule and more details: View Events Join remotely: https://utexas.zoom.us/j/94052130734 In person: Classroom 15.216B, Physics, Math and Astronomy Bldg. UT Austin, Department of Astronomy 2515 Speedway, Stop C1400 Austin, Texas 78712-1205 |
UTIG Discussion Hour: Nicholas Montiel - PhD Talk (UTIG)April, 23 2024Time: 2:00 PM - 3:00 PMLocation: ROC 2.201 |
UTIG Seminar Series: Cornelia Rasmussen, UTIGApril, 26 2024Time: 10:30 AM - 11:30 AMLocation: PRC 196/ROC 1.603 Speaker: Cornelia Rasmussen, Research Associate, University of Texas Institute for Geophysics Host: Krista Soderlund Title: The Emerging Field Of Position-Specific Isotope Analysis: Applications in chemical forensics, exobiology, geo- and environmental sciences Abstract: Complex organics can be found all over our solar system and within each living thing on our planet, be it as part of its physiology or as a contaminant. However, different processes can lead to the formation of chemical identical molecules. This makes answering a number of scientific questions challenging. One example is distinguishing between biotic and abiotic molecules, hence hindering life detection on early Earth but especially on other planetary bodies, such as on Mars, Titan, Enceladus and on meteorites where organics have been detected. Moreover, tracing molecules as they move through the environment can be demanding, yet is essential in studying the flow of organic molecules as well as correlating pollutants with their source. Novel tools to address these challenges are currently being developed. Especially, the emerging field of position-specific isotope analysis is beginning to grant access to the unique intramolecular carbon (13C/12C) isotope fingerprint preserved in complex molecules. This fingerprint can be applied in various scientific disciplines, ranging from forensics to exobiology, geo- and environmental sciences, including geo health. Nuclear magnetic resonance spectroscopy (NMR) has the potential to become a key player in this research area, as it allows the analysis of organics within complex mixtures, all without the need to fragment the molecule into single carbon units or the combustion of the molecule of interest. We have been developing several NMR tools that allow us to investigate the intramolecular carbon isotope distribution within various molecule classes and to test the central hypothesis that the position-specific carbon isotope distribution within complex organics depends on a molecule’s source and formation history. |
Planetary Habitability Seminar SeriesApril, 29 2024Time: 1:00 PM - 2:00 PMLocation: PMA 15.216B UT Center for Planetary Systems Habitability Seminar Series. See website for speaker schedule and more details: View Events Join remotely: https://utexas.zoom.us/j/94052130734 In person: Classroom 15.216B, Physics, Math and Astronomy Bldg. UT Austin, Department of Astronomy 2515 Speedway, Stop C1400 Austin, Texas 78712-1205 |
UTIG Discussion Hour: Kristian Chan - PhD Talk (UTIG)April, 30 2024Time: 2:00 PM - 3:00 PMLocation: ROC 2.201 |