Geology/Geophysics Seminar

Venus and Earth are often described as planetary “twins”; however, the large-scale volcano-tectonic surface features on each body indicate a different interior state and suggest that somewhere in their evolutions, they diverged. Certain features that represent a surface manifestation of the mantle convection within Venus show a “plume-thermal dichotomy”: large-scale mantle upwellings (plumes) appear to coexist in both space and time with many more small-scale mantle upwellings (i.e., thermals). How can we explain this coexistence of multiple scales (~100 km vs ~1000 km) of mantle convection on Venus, and why do these smaller-scale convective features, called coronae, not appear on the surface of the Earth? This talk will discuss the role that mineral phase transitions may play in the mantle convective dynamics of present-day Venus. The talk will also discuss the role that changes in mineral phase transitions may play in the long-term thermal evolution of terrestrial planetary bodies like Earth and Venus.

The surface of Mars once hosted flowing liquid water and a warmer climate than today, with past water-rock interactions recorded by the carbonate deposits found on its surface. This work explores the depositional setting of the Margin unit, a major Mg-carbonate deposit near the fluvial inlet to Jezero crater, using ground-penetrating radar data collected by the NASA Mars 2020 Perseverance rover’s Radar Imager for Mars Subsurface Experiment (RIMFAX) instrument. Soundings are reported from more than 35 m below ground, ~1.75 times deeper than other Jezero geologic units explored to date. We identify numerous subsurface features and submeter to hundred-meter scale layering across a ~6.1-km rover traverse. We infer that subsurface reflectors are consistent with buried fluvial features that have undergone multiple erosional-depositional episodes. This work extends the known history of aqueous activity within Jezero crater. It illuminates a well-preserved paleo landscape wherein a deltaic environment developed before the formation of the Jezero Western Delta, as early as the Noachian (~4.2 to 3.7 billion years ago).

Earth’s only natural satellite, the Moon, orbits our planet at about 239,000 miles (385,000 kilometers, or ~60 Earth radii). Each month, it spends about 3–4 days passing through the nightside of Earth’s magnetosphere — the magnetotail. Far from Earth, the magnetotail has a weak, highly variable magnetic field and strong spatial gradients. This makes it a natural laboratory for studying magnetic fields and charged particle interactions under conditions opposite to those in the dipole-dominated inner magnetosphere. Using observations from two ARTEMIS (Acceleration, Reconnection, Turbulence, and Electrodynamics of the Moon’s Interaction with the Sun) spacecraft, orbiting the Moon on elongated, near-equatorial (~100 km × 19,000 km) trajectories, we examine the near-Moon magnetic field and plasma properties under varying space weather conditions. We find that during quiet times, the plasma properties in the distant lunar magnetotail are largely controlled by the energy of the solar wind, and high-energy particles are rare. During active periods, however, we observe bursts of very energetic – sometimes even relativistic – ‘killer’ electrons near the Moon. Our results suggest that these electrons were energized within the magnetotail itself, although the exact mechanisms remain unclear.

This work integrates spatiotemporal datasets in California over a twenty-year period (2003-2023) to identify links between anomalies in fecal indicator bacteria (FIB) — Escherichia coli, total coliforms, fecal coliforms, and Enterococcus — at coastal areas and the occurrence of recent wildfires. Average FIB concentrations, land cover, precipitation, and burn histories are quantified with available data for every month across all watersheds flowing into the California Coast. Monthly non-burn background FIB concentrations are calibrated in coastal areas. FIB anomalies that are statistically distinct from these background levels are identified for both burn periods (>10% of the watershed burned within a two-year window) and non-burn periods and are reported as the anomalous FIB concentrations normalized by the background concentrations. Correlations between FIB anomalies and monthly precipitation and land cover fractions are identified by comparing data from all watersheds impacted by large fires. During burn periods, anomalous total coliform concentrations relative to baseline values (anomalies) exhibit a negative correlation with urban and positive correlations with coastal oak woodland, mixed chaparral, and redwood land cover fraction and monthly precipitation. Fecal coliform anomalies exhibited a positive correlation with urban land cover fraction after fires. Results indicate prominent post-fire export of total coliform bacteria (perhaps from decaying plant material and soils) in watersheds with less urbanization, and prominent export of fecal coliform bacteria from humans and animals in urbanized watersheds to coastal waters. This is the first effort to generalize post-fire FIB responses in coastal waters. The results may inform future risk assessments and proactive measures.

Early atmospheric chemistry and climate play an important part in what worlds have liquid water and how life could emerge, but preserving evidence for this over geologic time often presents a challenge. Compared to Earth, which undergoes continuous erosion by plate tectonics, Mars offers a superb window into its earliest history, preserved right on its surface, and is therefore an excellent case study for planet evolution. What we learn here can extend to other planets in the solar system, the early Earth, Venus, and even outside the solar system at exoplanets.

Extensive surface evidence suggests Mars once hosted large volumes of liquid water, implying a warmer climate sustained by a thicker atmosphere than present day. Previous studies propose that CO2-dominated atmospheres enriched with H2 or CH4 could have driven warming, though such gases exhibit short atmospheric lifetimes. This talk will first present a novel explanation for sustaining warm climates up to 40 million years: hydrogen release through crustal hydration. Evidence of atmospheric climate and redox transformations may remain preserved on the surface and in meteorites. To test our hypothesis, we show that deposits measured by Mars Science Laboratory and isotopic fractionation measured in meteorites can both be explained photochemically under specific climates and atmospheric redox states. We also present predictions of isotopic signals under different climate and redox states that Mars Sample Return missions should search for.

Like Mars, Venus may have also experienced large climate changes during its evolution too. Modern Venus is a hot, arid environment, yet the possibility of an early cool, habitable phase remains unresolved. The talk will explore when sulfuric acid hazes in Venus’ history could have sufficiently cooled the planet to allow surface rainfall. This scenario may be testable using NASA’s upcoming Habitable Worlds Observatory (HWO), which will observe exoplanets with Venus-like characteristics.

While HWO’s launch lies years ahead, exciting exoplanet data have already been measured with JWST, Hubble, and Spitzer. When we observe planets also influences their chemistry, and the presentation will explain a new hypothesis for missing methane at warm Neptunes. The talk also highlights ongoing efforts to develop modeling tools to interpret future data, including microphysics and 3D dynamics. These analyses aim to refine atmospheric interpretations and advance our understanding of planetary evolution across diverse environments.

Large-scale flows in gas giant atmospheres, icy moon subsurface oceans, and liquid planetary cores control the transport of heat and chemical species, as well as the generation of magnetic fields in electrically conducting fluid regions. Understanding these fluid systems fundamentally depends on our ability to predict the variety of dynamical length scales that arise from the interplay of rotation, buoyancy, and planetary curvature. These ingredients are simulated in UCLA’s Coreaboloid, a rotating convection device equipped with an infrared (IR) camera aimed axially downwards at the free surface of the fluid. IR-derived surface temperature fields function much like the clouds in Jupiter’s weather layer, revealing the multi-scale dynamics of the rapidly-rotating flow. The small scales in the temperature field are dominated by eddies generated through rotating convection and baroclinic instabilities. Kinetic energy injected through these processes feed into east-west Jovian-like jets, which saturate at large scales set by friction. This alternating pattern of zonal flows influences the transport of heat, producing bands of matching width in the temperature field, reminiscent of Jupiter’s stripes. Complemented by particle image velocimetry, the IR-derived temperature data provide the first match between observed and predicted convective, baroclinic, and jet length-scales coexisting in laboratory experiments of rotating convection with multiple coherent jets.

Laboratory tank experiments have long served as simplified analogues for planetary fluid dynamics, offering a controlled way to isolate the effects of rotation and external forcings. Although no laboratory setup can reproduce the full complexity of a planet, such experiments can successfully capture the dominant dynamical balances that govern large scale geophysical flows. In this talk, I discuss how tank scale experiments reflect key aspects of planetary fluid dynamics at a conceptual level, with emphasis on what is gained—and what is lost—through this reduction in scale and geometry.

GPS measurements of the vertical motion of Earth’s surface along with gravity measurements from the Gravity Recovery and Climate Experiment (GRACE) and GRACE-Follow On (GRACE-FO) missions provide valuable data to analyze changes in terrestrial water storage (TWS) as a function of position and time. These integrated datasets effectively quantify TWS variations across global and regional scales. Here we expand upon previous studies that have incorporated GPS and GRACE(-FO) to determine long-term water cycle changes in the United States, by applying similar methodologies to provide one of the first looks into water cycle changes throughout Europe. Through applying a standard list of processing steps, we filter out unwanted signals in GPS measurements of vertical ground motion and identify GPS stations in Europe measuring only elastic displacements. By analyzing these elastic GPS displacements alongside GRACE(-FO) inferred data, we identify local and regional water cycle trends over the past 23 years, offering crucial insights into Europe’s hydrological evolution.

Between 2010 and January 2025, California’s western Transverse Ranges experienced 18 fires with burn areas larger than 20 km2, the largest of which was the ~1141 km2 Thomas Fire, which burned across Ventura and Santa Barbara Counties in December 2017. During the same period, this area occasionally experienced unusually wet seasons with several landfalling atmospheric rivers. The most notable was the 2022-2023 atmospheric river season, which saw significantly elevated precipitation totals relative to the 30-year average and which triggered widespread landslides and debris flows in burned hillslopes across Southern California. Terrain, cloud cover, and vegetation cover complicate comprehensive range-scale post-atmospheric river surface disturbance mapping and monitoring. We created a supervised threshold-based NDVI (normalized difference vegetation index) change detection to map surface disturbances. We then run a frequency ratio-based susceptibility model using confirmed landslide heads. We applied these models to the eastern Santa Ynez Mountains northwest of Ojai, CA, on terrain burned by the December 2017 Thomas Fire and affected by the January 9, 2023 atmospheric river, to better understand the relationship between local topographic factors, landslide triggering, and extent of downstream sedimentation. We considered eleven different parameters at 250 landslide heads in (distance to channel, eastness, elevation, northness, plan curvature, profile curvature, roughness, sediment transport index, slope, and stream power index, and topographic wetness index). Of these, we found the three most statistically relevant metrics in our study area were plan curvature, topographic wetness index, and distance to channel.

Stars of all sizes and ages are plagued by “missing mixing problems,” or discrepancies between stellar models and observations that indicate some unknown turbulent mixing occurs in their interiors. The precise nature and degree of this anomalous mixing can hold major consequences for the fates of such stars, their spins, and their chemical makeup over time. I will discuss a particularly ubiquitous mixing problem in red giant branch stars—a widely observed class of star that our own sun will eventually become. The leading candidate to explain the missing mixing in these stars is salt-finger (or “thermohaline”) convection, a form of turbulence also found in tropical oceans. I will show that the dynamics of salt fingers in stars are often surprisingly similar to oceans (despite the enormous difference in fluid properties between stellar plasmas and salt water) except when magnetic fields are present. In such cases, magnetic fields dramatically enhance the turbulent mixing rate, potentially solving this long-standing missing mixing problem.

Atmospheric compositions of sub-Neptunes and super-Earths are often interpreted as tracers of formation history and volatile delivery. However, many of these planets likely experience long-lived magma-ocean phases during which their atmospheres can chemically equilibrate with a molten silicate interior. In this talk, I will show how magma-ocean–atmosphere interactions reshape atmospheric compositions and alter observable signatures. Using a global chemical equilibrium framework, I demonstrate that key abundance ratios such as C/O and H2O are not simply inherited from disk accretion, but can be substantially modified by interior processing. This has important implications for the interpretation of apparently water-rich worlds, which may in fact be significantly drier than predicted by previous models. I will also present new extensions of the model including sulfur and nitrogen chemistry, which provide additional diagnostics of volatile redistribution between atmosphere and interior and further strengthen the connection between interior evolution, atmospheric composition, and spectroscopic observations.

Giant impacts are the last major stage in the formation of terrestrial planets. They also are believed to play a major role in the assembly, long-term orbital stability, and resonant structure of exoplanetary systems. Both the mechanical shock and thermal heating of a giant impact can cause atmospheric mass loss. Previous works have shown that for planets with hydrogen-dominated atmospheres, the post-impact thermal heating acts as the dominant mass loss mechanism by inflating the planetary envelope and driving a Parker wind. However, past studies have considered the atmosphere and interior of these planets as discrete, non-chemically-interacting layers. We revisit this idea through the lens of a coupled atmosphere-interior system, allowing for chemical interactions between the core and envelope as the planet experiences heating and subsequent cooling after a giant impact. Considering a miscible interior affects the mean molecular weight of the envelope, allows hydrogen to mix into the core, and modifies the compositional gradients in the planet, thus altering the atmospheric mass loss. We present new results for the atmospheric loss due to giant impacts and discuss their implications for the formation and composition of super-Earths, sub-Neptunes and the radius valley.

The Main Belt is a region lying between Jupiter and Mars that contains a vast number of small bodies ranging in both size and shape. Until the 1990s, it was believed that the Main Belt consisted primarily of asteroids. However, the observation of 7968 Elst–Pizarro (133P/Elst–Pizarro) in 1996 by Eric Elst and Guido Pizarro, along with the work of Jewitt and Hsieh in the early 2000s, led to a new classification of small bodies: Main Belt Comets (MBCs). These comets – later deemed a subset of Active Asteroids – exhibit circular, asteroid-like orbits with semi-major axes 𝑎 < 𝑎_J, but maintain comet-like characteristics such as a coma, tail, or 𝐽𝑢𝑝𝑖𝑡𝑒𝑟 visible mass loss. Since their initial discovery, MBCs have been inadvertently observed throughout the Main Belt, and many have attempted to put constraints on their number density. A 2013 study by Waszczak and others placed upper limits of 33 and 22 active MBCs per million main-belt asteroids down to∼1-km diameter (Waszczak et al, 2013). A more recent study in 2022, by Ferellec and others, deduced a MBC-to-asteroid ratio of <1:500 (Ferellec et al. 2022). For this study, we plan to utilize the data from the WISE/NEOWISE survey to search for MBCs. WISE/NEOWISE is a space-based wide-field infrared telescope that surveyed MB small bodies over the past ten years. The mission’s objective was to identify near-earth objects (NEOs) and MBAs; thus, it has not been analyzed for MBCs. A thorough search of this data set should yield a number of MBCs which can then be used to place constraints on their number density in the Main Belt.

The Pacific Ocean covers a large portion of the Earth’s surface. However, seismic constraints on its crust and upper mantle structure remain limited due to sparse station coverage. Ambient noise tomography provides an effective approach to improving spatial coverage by extracting empirical Green’s functions (EGFs) between station pairs, while also offering enhanced vertical constraints on crustal structure. In this study, I use continuous seismic data from both land-based stations and ocean-bottom seismometers (OBS) to measure Rayleigh wave group and phase velocity dispersion across the Pacific region, targeting the period 1990–2025. Preliminary results based on the 2005–2006 datasets demonstrate that reliable dispersion curves can be obtained for station pairs at appropriate interstation distances over a period range of 5–50 s. Ongoing work focuses on incorporating additional years of data and improving signal quality, particularly for OBS records, to enhance path coverage and support future tomographic imaging of oceanic structure.

The Mendocino Triple Junction is in a tectonically complex region where a multitude of physical mechanisms contribute to crustal deformation. To discern these physical drivers, our group (1) enhanced the Northern California Seismic Network (NCSN) earthquake catalog spanning the 2022 Ferndale and 2024 Offshore Cape Mendocino sequences using machine learning and template matching approaches, (2) extended a repeating earthquake catalog to study the contribution of aseismic slip to both earthquake sequences (3) analyzed changes in groundwater levels during the both sequences. In this talk, I will present our enhanced catalogs, which reveal ~5-10 times more events than the NCSN catalog, with a clear concentration of seismicity at the edges of the 2024 sequence’s coseismic slip patch. Additionally, I will highlight our study’s evidence for accelerating afterslip in the weeks following the 2024 sequence, and an elastic strain mechanism to explain the observed groundwater level changes.

Wildfire fundamentally alters hillslope and channel sediment transport by rapidly mobilizing dry ravel, which accumulates in channels immediately following burning and is subsequently flushed during the first post-fire rainfall. Using repeat high-resolution topographic surveys and channel-scale swath analyses within the Palisades burn scar, we examine how first-flush sediment storage and redistribution relate to channel steepness (k_{sn}) and underlying lithologic gradients. The Palisades wildfire provides a natural laboratory to examine these dynamics, as the study area spans a pronounced east–west lithologic gradient associated with systematically increasing channel steepness. We show that post-wildfire sediment transport and hillslope failure can be partitioned into two coupled but distinct processes: (1) dry ravel production, which occurs exclusively in steep, high-k_{sn} lower-order channels, and (2) precipitation-driven mobilization and redistribution of this material into channel trunks during early storm events. Dry ravel does not accumulate ubiquitously across the drainage network, but is instead restricted to channels that exceed a critical steepness threshold, indicating a strong topographic control on post-fire sediment supply. We further demonstrate that dry ravel accumulation in steeper lower order channels is spatially coincident with active scouring of the underlying channel bed, leading to sediment budgets that exceed existing infrastructural design capacities. Following rainfall, dominant patterns of erosion and deposition are negatively correlated with k_{sn}: deposition dominates in low-k_{sn} reaches, while high-k_{sn} channels experience net erosion and sediment export. Together, these observations suggest that first-flush sediment response following wildfire is governed not only by burn severity and rainfall intensity, but also by broader lithologic and topographic controls that regulate transport efficiency, sediment storage, and channel–hillslope coupling.

Static, long-term deformation generated by earthquakes provides essential constraints on fault slip, stress redistribution, and crustal rheology. However, measuring earthquake-induced static deformation remains challenging, particularly for moderate earthquakes, due to their small strain amplitudes. Traditional geodetic observations, such as high-frequency Global Navigation Satellite System (GNSS) and tiltmeter measurements, are susceptible to noise and limited in spatial resolution. Fortunately, Distributed Acoustic Sensing (DAS) offers a novel technology by converting pre-existing optical fiber into thousands of dense strain sensors, enabling high-resolution observations over large areas. This approach has been successfully applied to monitor static deformation associated with magma migration in volcanic systems.
Here, using DAS data from the Taiwan Milun Fault Drilling and All-inclusive Sensing (MiDAS) project, we demonstrate that static strain from more than a dozen aftershocks (Ml > 5) following the 2024 Mw 7.4 Hualien earthquake can be clearly recorded. Interestingly, at short spatial scales (~100 m), the strain field exhibits complex patterns, with even alternating extensional and compressional signals. Through numerical simulations, we hypothesize that this phenomenon is linked to anomalies in the medium’s Poisson’s ratio, which is consistent with the fractured structure of the Milun Fault. Overall, our results highlight DAS as a reliable complement to conventional geodetic observations, offering a useful tool for investigating seismic deformation fields.

Quantifying spatiotemporal dynamics of landslides, including occurrence, reactivation, and timing, is critical for understanding hazard risk in steep mountains. However, in regions like the Eastern Himalayas, mapping these spatiotemporal patterns is heavily restricted by persistent cloud cover, land-cover variability, and a scarcity of historical ground-truth inventories. To overcome these data and observational limitations, we utilized a novel unsupervised machine learning framework—the Stable Diffusion Change Detector (SDCD)—to extract continuous surface disturbance records directly from time-series optical satellite imagery. By bypassing the limitations of traditional manual labeling, this approach allows us to generate high-resolution landslide inventories in a data-scarce, cloud-prone region. Applying this method, we identify ~300 surface disturbance events, including landslides in hillslopes and channel-adjacent regions, over 1300 km2 and 10 years. We investigate temporal triggering of these events by analyzing daily precipitation and spatial patterns by comparing environmental conditions, including precipitation, long-term erosion rates, and topographic slope. Our time-series analyses indicate that abundant landslides in frontal basins are directly related to the magnitude of daily precipitation events, and the spatial density of overall landslides is highly consistent with independently measured long-term erosion rates. These results highlight the importance of high-resolution landslide datasets, both spatially and temporally, for understanding the complex controls on landscape change over time.