Skip to main content

2021 Earth/ Environmental Sciences, Technology and Engineering

The pre-recorded talks and posters on this page showcase the work of students who received NC Space Grant research funding for the 2019-20 and 2020-21 academic years. The menu at right provides links to pre-recorded talks and posters by other funded students on additional topics.

Patrick Gray 

2019-2020 NC Space Grant Graduate Research Fellow
Duke University
Graduate Student (Ph.D.), Marine Science

Biophysical interactions along the Gulf Stream front: are we missing fine-scale productivity and diversity that is important more broadly? 

Large scale niches and gradients in marine productivity are set by environmental contrasts at the basin scale, but they are rearranged by mesoscale flows, which in turn are reshaped by submesoscale dynamics. These meso- and submesoscale processes can stir existing gradients and drive fluxes across the ocean’s stratified vertical layers, linking the euphotic zone to the ocean’s nutrient rich interior. Both lateral stirring and vertical fluxes shape the diversity and productivity of phytoplankton, cascading changes throughout the ocean’s food web. These physical dynamics and biological responses produce the ocean’s characteristically extreme spatially and temporally heterogeneity which includes hotspots of biological productivity, diversity, and export. These biophysical interactions are complex, and despite decades of research there persist conflicting and incomplete theories concerning how and where they drive phytoplankton productivity, diversity, and export. The Gulf Stream is an intense western boundary current and the current’s sharp front off Cape Hatteras, North Carolina has substantial submesoscale processes and is a known biodiversity hotspot. To date, the distribution of phytoplankton productivity and diversity across the front has not been characterized below the scale of a few kilometers. The ecological and physical environment at this scale and how its dynamics might predict important broad scale properties is a critical gap in our understanding of the ocean and ability to predict its future state. Towards addressing this gap we present a novel observation approach using drones to complement satellite observations. First we outline our approach using drones to retrieve ocean color, and then we present preliminary data on productivity and diversity across the Gulf Stream front addressing two questions about biophysical interactions: 1) Does the front contain a phytoplankton community that is distinct from the regions on either side? 2) Do submesoscale fronts adjacent to the main Gulf Stream front also structure phytoplankton?

Faculty Advisor: David W. Johnston

Rebecca Hahn 

2020-2021 NC Space Grant Graduate Research Fellow
North Carolina State University
Graduate Student (Ph.D.), Geology 

Exploring the Morphology and Spatial Relations of Volcanoes on Venus 

NASA’s Magellan spacecraft mapped virtually the entire Venusian surface using synthetic aperture radar (SAR) as well as nadir-directed altimetry. The high-resolution radar data collected during this mission enabled the discovery and classification of volcanic features and structures across Venus far beyond the scope of earlier missions. Previous studies have documented volcanoes and volcanic structures across Venus (see, for instance, Crumpler et al., 1993). To build upon this earlier work, I utilized ESRI’s ArcMap to develop a new global catalog of Venusian volcanic edifices with the Magellan SAR FMAP (full-resolution radar map) left- and right-look global mosaics at 75 meter-per-pixel resolution. The completed dataset includes ~85,000 shield volcanoes, 45 times more than any earlier such databases for Venus. With this new dataset, I utilized both ArcMap and ArcGIS Pro to determine preliminary morphological properties for all edifices >5 km in diameter, including area, basal diameter, and aspect ratio. Future work will involve using a combination of tools in ArcGIS Pro ModelBuilder to determine the height and volume of edifices (where data resolution is sufficient). For edifices <5 km in diameter, I developed a quantitative approach of grouping these edifices based on proximity to build a global catalog of “shield fields”— areas with relatively high spatial concentrations of shield volcanoes. Within these shield fields, I employed statistical analysis in ArcGIS Pro to determine the degree of clustering and the overall orientation of the field, as well as the relationship between the field and any proximal tectonic or volcanic structures. Further analysis of my global catalog of shield volcanoes and my global catalog of shield fields will provide a new framework for understanding how volcanoes form and develop across Venus in particular, and large terrestrial planets generally.

Faculty Advisor: Paul K. Byrne

Ian McGregor

2020-2021 NC Space Grant Graduate Research Fellow
North Carolina State University
Graduate Student (Ph.D.), Geospatial Analytics

Leveraging multi-source data to improve near real-time forest disturbance monitoring

Remote sensing has been used for decades to understand forest disturbance across the world, especially in the Tropics. The majority of studies using remote sensing data until now have focused on retroactive analyses, where prior imagery is used to see how the landscape has already changed. With the availability of data from Google Earth Engine, recent studies have been shifting toward a near real-time (NRT) monitoring paradigm to assess forest change almost as it’s happening. Despite these advances, it remains rare to accurately, consistently, and quickly identify small-scale forest disturbances (~30m) after occurrence. This is not ideal for forest managers facing logging pressures in Myanmar, which is one of the most forested nations in SE Asia with high levels of biodiversity. Logging events near Chatthin Wildlife Sanctuary, for example, are typically short-lived. Thus, we wanted to use multi-source data to simultaneously increase spatial accuracy and decrease temporal latency of identified disturbances.

To do so, we expanded on recent efforts of combining remote sensing data by aggregating Sentinel (1,2 [top of atmosphere (TOA)]), Landsat (8, TOA), and MODIS (MOD09GA) data for the normalized difference vegetation index (NDVI) across training points. Specifically, we combined the z-scores from each sensor time series and used multiobjective optimization to apply rolling weights, where the most recent observations were given larger influence. Initial results yield consistent disturbance alerts along with daily probabilities of disturbance across the study area. We focused on Chatthin Wildlife Sanctuary in north-central Myanmar due to our partnership with the park staff and the Smithsonian. The next steps of the research will focus on increasing the training dataset, incorporating Bayesian priors, and using field work to optimize the Sentinel 1 interpretation. To our knowledge, this study represents one of the first instances of a large multi-source data method applied to NRT monitoring.

Faculty Advisor: Josh Gray

Shannon Ricci 

2020-2021 NC Space Grant/NC Sea Grant Graduate Research Fellow
North Carolina State University
Graduate Student (Ph.D.), Geospatial Analytics

Since the mid-1800s, artificial reefs have been constructed in coastal waters to benefit resident fish species, offset habitat loss, and provide recreational and educational opportunities to surrounding communities. The state of North Carolina’s Division of Marine Fisheries currently maintains 43 artificial reefs in its coastal waters. While artificial reefs serve a dual purpose of providing both essential habitat and recreational opportunities, the majority of research centers on reef ecology, resulting in a lack of research evaluating the socioeconomic aspects of the reefs. Historically, recreational use of artificial reefs has been assessed through a combination of on-water observations and various methods of boater interviews. These methods are often biased by low response rates and have limited spatial and temporal extent due to cost and time constraints. To overcome these limitations, we developed and deployed an object detection model (YOLOv3) to detect small boats (greater than ~9m) within Planet’s high resolution (~3 m/pixel) imagery at four pilot artificial reefs over a period of one year (2019). Detections revealed seasonal patterns in visitation, with more boats in the spring and summer months. Spatial patterns revealed more popular reef sites, and within reefs, boats were primarily clustered around sunken vessels as opposed to other reef materials. Leveraging the growing availability of high temporal and spatial resolution satellite imagery, this model can be used to detect boats around other areas of interest, and for other years. Such a dataset can be useful in evaluating reef effectiveness and help in the design of future reefs to support local coastal economies and recreation.

Faculty Advisor: Del Bohnenstiehl

Chloe Schneider 

NASA DEVELOP Internship Award – Summer 2020 
The University of North Carolina at Chapel Hill
Recent Graduate (December 2020), Environmental Science

Drought and Flood Monitoring using NASA Earth Observations 

The Missouri River Basin provides irrigation water for a substantial part of the domestic agricultural sector within the United States. Drought events pose a significant threat to economic livelihoods of dependent individuals, industries, and ecosystems (e.g. farmers, local tribes, hydroelectric power, wildlife). In 2017 alone, the Missouri River Basin experienced a severe drought that resulted in a $2.6 billion loss to the U.S. Northern Plains. In response to such events, organizations throughout the basin, such as the Montana Climate Office, have dedicated efforts for drought monitoring and communicating relevant information to local stakeholders. In an effort to aid regional decision-making capabilities, the project team partnered with the Montana Climate Office, NOAA National Weather Service (NWS) Missouri Basin River Forecast Center, and NOAA Regional Climate Services, Central Region to create a monthly Composite Moisture Index (CMI) that relies on NASA Earth observations from the Terra Moderate Resolution Imaging Spectroradiometer (MODIS) and the Soil Moisture Active Passive (SMAP) mission. From these satellites, as well as the NOAA NWS National Operational Hydrologic Remote Sensing Center Snow Data Assimilation System (SNODAS), our team aggregated climate datasets including soil moisture, snow cover, snow depth, and snow water equivalent to compute a CMI that indicates regional moisture conditions during the winter months. The March CMI values produced over the Missouri River headwater subbasin strongly correlate (r = 0.75) with spring and early summer stream discharge, demonstrating the use of this metric to indicate moisture conditions for the snowmelt and growing seasons.

Advisor: Andrew Shannon (NASA DEVELOP) 

Olivia Sessoms

2020-2021 NC Space Grant Undergraduate Research Scholar
East Carolina University
Undergraduate Student (Senior), Environmental Engineering

Biohydrochemical Computational Modeling of Nitrogen Species Transport at Lake Mattamuskeet Waterfowl Impoundments 

Managing the nitrogen cycle is one of the 14 Grand Challenges for Engineering in the 21st Century. Addressing nitrogen pollution is critical, as it directly affects water, soil, and air quality. To minimize pollution risks, it is important to be able to predict nitrogen contributions from various contamination sources. This can be done through quantitative simulation of the transformation and transport of nitrogen species in soil zones and aquifers. Many previous models, however, have been found to oversimplify nitrogen reactions and/or are extremely computationally demanding. This project details the development of a model in COMSOL Multiphysics for nitrogen transformation and transport in variably saturated soil and its application to waterfowl impoundments at Lake Mattamuskeet. COMSOL Multiphysics software enables simulation of coupled physics problems, which reduces computational burden, as well as the creation of exportable, easy-to-use apps. Lake Mattamuskeet is the largest natural lake in North Carolina and center of the Mattamuskeet National Wildlife Refuge. The lake’s managed wetlands are habitat to large amounts of migratory bird populations, and nutrients from these impoundments have been hypothesized to contribute to nutrient pollution of the lake. This model simulates the effects of low, medium, and high levels of nitrogen sources -such as bird feces transport, agricultural runoff, impoundment drawdown (in which water and nutrients are removed and sent downstream), and atmospheric deposition- to the production and transformation of nitrogenous compounds in the soil. The model has also been converted to a user-friendly simulation app that allows for visualization of reactive nitrogen transport in soil even by non-expert users. Ultimately, the model will provide a novel perspective and valuable information towards improving nitrogen management processes.

Faculty Advisor: Ali Vahdati

Bethany Sutherland

2020-2021 NC Space Grant Graduate Research Fellow
North Carolina State University
Graduate Student (Ph.D.), Atmospheric Science

A Comparison of aerosol type specific optical parameters with AERONET retrievals across North America 

In situ measurements, remote sensing, and models have been widely used to characterize the radiative effects of aerosols, yet the direct radiative effect (DRE) of aerosols, or the amount that aerosols in the atmosphere affect the radiation budget, remains one of the leading contributors to climate prediction uncertainty. This is due partially to our inability to carry out continuous global retrieves of the vertically resolved optical properties of aerosol necessary to do the DRE estimates, namely single scattering albedo (ω) and asymmetry parameter (g). As satellite direct retrievals of ω and g are unlikely to become available in the near future, we have proposed a methodology for using High Spectral Resolution Lidar (HSRL)-derived aerosol types to quantify the direct radiative effect of aerosols. Using aerosol type-based values for ω and g presents a promising avenue for narrowing uncertainty in DRE estimates as several schemes for determining aerosol types from lidar measurements have become available. HSRL aerosol types are derived using aerosol intensive properties and are therefore expected to best represent the aerosol microphysics that influences the DRE calculations. 

     Here we investigate the feasibility of using type-specific ω and g values derived using the Creating Aerosol Types from Chemistry (CATCH) algorithm, by comparing type-based ω and g to level 1.5 retrievals from AERONET site across North America for the year 2014. CATCH is used to determine simultaneous aerosol types from the GEOS-Chem 3-D atmospheric chemistry model version 12.7.1 simulations for the same year. AERONET retrievals are screened to include only days where a single type of aerosol dominates. Results of this study suggest that using type specific optical properties will be able to produce DRE estimates with lower uncertainty.

Faculty Advisor: Nicholas Meskhidze