Faculty Advisor: Dr. Sally Pusede, University of Virginia | Research Mentor: Solianna Herrera, University of Virginia
Investigating the Vertical and Horizontal Variability in Air Pollutants in Los Angeles Using In-Situ and Remote-Sensing Aircraft Measurements
Fred Okafor, Prairie View A&M University
Atmospheric mixing exerts a large control over the concentrations of atmospheric pollutants. For example, pollutants are diluted during the growth of the daytime boundary layer in the morning and also carried with horizontal winds. To understand the impact of these mixing processes on air quality in the polluted city of Los Angeles, California, I use in situ observations of atmospheric vertical profiles collected onboard the NASA C-23 Sherpa and remote-sensing observations of the horizontal distribution of NO2 made by the GeoTASO instrument from the NASA UC-12 across the LA basin on 26 June 2017. I focus on chemical and meteorological data from the C-23 Sherpa collected during four missed approaches at LAX airport to interpret boundary layer growth throughout the morning. I take advantage of the fact that during missed approaches over LAX, the Sherpa descended over land and ascended over the ocean, which produced distinct vertical profiles for comparison. I use these vertical profiles and the GeoTASO data to investigate land-sea breeze effects, finding a very steep land-ocean gradient in the early morning.
Airport Emission Rates from Airborne Measurements of NO(X) and CO(2) (AERO-AIM)
Dallas Burke, University of North Florida
Commercial aviation, with a predicted carrier passenger growth rate of 1.9 percent per year over the next twenty years in the U.S., will continue to be one of the most dominant forms of transportation in the sector. It is becoming increasingly important to understand the engine emissions from aircraft at altitude and throughout the landing and takeoff cycle (LTO) at airports. This study uses a new, top-down approach to measure the total airport emission rates of the Los Angeles International Airport (LAX), as well as predictions for other major airports. Since 2012, the NASA Student Airborne Research Program has made in-situ measurements of carbon dioxide and the oxides of nitrogen (NOx) during airborne missed approaches at multiple airports within the Los Angeles Basin. We show that observationally derived NOx emission indices (grams of NOx emitted per kilogram of jet fuel consumed) from individual emission plumes sampled at low altitude along the runway are consistent with the values listed in the ICAO Aircraft Engine Emissions Databank. The total airport emissions rates determined by the top-down approach are compared to bottom-up aircraft emission rate estimations.
Inter-comparison of NO(2) Differential Slant Columns Obtained by the GeoTASO Instrument and Surface-Based Pandora Data
Joshua Coduto, Augustana College
Atmospheric nitrogen dioxide (NO2) plays a role in the formation of hazardous compounds such as tropospheric ozone, aerosols, and acid rain. To better understand these processes, we need precise and accurate measurements of atmospheric NO2 that capture variability across space and time relevant for the atmospheric lifetime of NO2. The Geostationary Trace gas and Aerosol Sensor Optimization (GeoTASO) airborne instrument is a test bed for the Tropospheric Emissions Monitoring of Pollution (TEMPO) and Geostationary Environment Monitoring Spectrometer (GEMS) satellites. This study compares the NO2 differential slant column data obtained from the GeoTASO instrument aboard the UC-12B King Air with the slant column data from six ground-based Pandora spectrometers during the June 27th and 28th, 2017 SARP flights over the Los Angeles Basin. The GeoTASO data were used to retrieve differential slant column densities of NO2 using Differential Optical Absorption Spectroscopy (DOAS) from a 435-460 nm window. This work evaluates how well these differential slant columns compare to the more well-established retrievals from Pandora. The collected coincident points for each of the five recorded flights show a strong correlation between the two instruments. Additionally, the NO2column densities at each Pandora site are investigated relative to the spatial distribution mapped by GeoTASO and compared to the more long-term dataset from Pandora collected over the month surrounding the SARP flight days. Results from this study will assist in evaluating the differential slant column retrievals from GeoTASO and show results on the spatial and temporal changes in NO2 in the LA Basin for the first time with these two instruments.
Investigating NO(2) Emissions from a Lightning Ignited Forest Fire in Sequoia National Park by Ground, Airborne and Satellite Measurements
Ryan Tumminello, Hendrix College
NO2 is a reactive gas directly linked to ambient air quality through its role as a precursor to tropospheric ozone. Natural emissions of NO2 are dominated by biomass burning and lightning strikes. Production of NO2through these means is challenging to constrain due to the highly variable nature of each phenomena. To better understand natural NO2production, this study focused on emissions generated by a lightning ignited forest fire located in Sequoia National Park known as the Schaeffer Fire in June 2017. NO2 columns were collected by GeoTASO onboard the NASA UC-12 King Air over the fire and the resulting smoke plume. Lightning data was obtained from the Earth Networks Total Lightning Network to determine the storm severity leading up to ignition of the fire in addition to measuring the lightning flash rate required for modeling of NO2 production. An empirical relationship was adapted from previously published work to estimate NO2 emissions from the fire with remote sensing data obtained from the MODIS instrument onboard the Aqua and Terra satellites for application in the model. Results for the Schaeffer Fire, which would have been too small to be visible by current satellite NO2 instruments, were compared to results derived from a global study.
An Investigation of Inequitable Distribution of NO(2), in the Los Angeles Basin using data from the GeoTASO
Aracely Navarro, Colorado College
In the U.S., poor air quality greatly affects the health of people living in cities. However, research has consistently shown that air pollution burdens are highest for low income and communities of color. Though, these studies often use spatially-averaged concentrations measured by surface monitors for whole communities, which may potentially lead to bias in exposure estimates. To address this issue, this study quantifies income and race-based patterns in distribution of the pollutant nitrogen dioxide (NO2) in Los Angeles, California utilizing NO2 slant columns measured by the Geostationary Trace gas and Aerosol Sensor Optimization (GeoTASO) airborne instrument onboard the NASA UC-12B King Air on 26 June and 27 2017. GeoTASO observations were combined with income and demographic data at the census tract spatial scale provided by the California Environmental Protection Agency Environmental Health Screening Report. This analysis found higher NO2associated with greater poverty and nonwhite communities. For example, NO2 concentrations were higher for the Hispanic community experiencing poverty. Communities that were 60-80% Hispanic with 60-80% in poverty were exposed to statistically significantly more NO2 than a community with 60-80% White at the same poverty level. It was generally the case that White, more wealthy communities had lower mean NO2 value. Poor Hispanic communities had the strongest association with NO2 exposure, i.e. as percent Hispanic increased and percent in poverty increased the mean NO2 also increased, displaying an R2 of .90-.95 over the two days of the study. A similar trend was found for Asian Americans, and, in certain cases, for African Americans. Second, I compare the results obtained with the high-resolution GeoTASO dataset to results derived from an identical analysis using interpolated surface monitor NO2 data. I find the surface monitor observations underestimate the association between the presence of NO2 with poverty compared to GeoTASO.
Ozone Production in Los Angeles: Investigating Spatial Differences in the Impacts of NO(X) Emission Control
Margarita Reza, University of Houston
Los Angeles has historically experienced the highest ozone in the U.S. Strategies to reduce high ozone have included reduction of emissions of both nitrogen oxides (NOx) and volatile organic compounds (VOC), the two main precursors to ozone. Observations of ozone on weekdays and weekends, as well as from many intensive atmospheric experiments in LA, have indicated that for decades, ozone production chemistry in LA has remained sensitive to VOC rather than NOx. However, very recent literature has suggested that LA ozone chemistry may be transitioning to a chemical regime that is sensitive to NOx, which would mean that LA air quality is positioned to benefit from NOx emission control on diesel truck traffic that are currently in the implementation phase. This study investigates this possible chemical-regime transition by first looking at the temporal trends in NOx concentrations at nine monitoring stations throughout LA. I find that there is spatial variability in the observed rate of decrease of NOx, suggesting that ozone has been affected differently throughout the LA basin. To investigate this potential spatial variability in the impacts of NOx changes, and, potentially, the change in ozone chemistry sensitivity to NOx, I use data gathered from the Geostationary Trace Gas and Aerosol Sensor Optimization (GeoTASO) airborne instrument on board the NASA UC12-B King Air on June 26th and 27th, 2017. I use an analytical model to simulate ozone production in LA to identify the sensitivity of ozone production to NOx. Understanding the dependence of ozone production on NOx is critical for developing effective control strategies in this area.
Isoprene as a Probe of Drought Impacts on Ozone Chemistry in the San Joaquin Valley of California
Emily Najacht, Saint Mary’s College/ University of Notre Dame
The California San Joaquin Valley (SJV) has recently (2012–2015) experienced the worst drought in recorded history. The SJV is also the location of some of the worst ozone pollution in the U.S. Ozone concentrations are a function of atmospheric mixing, the ozone lifetime, and ozone chemical production, and each of these terms are expected to vary with drought conditions. As a result, it is difficult to predict the impacts of drought on high ozone in the SJV, which is important, as global climate change is anticipated to alter drought frequency and severity. Tropospheric ozone production results from complex, nonlinear chemistry combining nitrogen oxides (NOx), the atmospheric oxidant the hydroxyl radical (OH), and volatile organic compounds (VOCs). One ozone-forming VOC is isoprene, a highly reactive biogenic VOC emitted from plants through the stomata. In the daytime, the concentration of isoprene in the atmosphere is determined by the rate of emissions and the abundance of OH. This study takes advantage of the short lifetime of isoprene (~10 minutes on a sunny day), before and during drought conditions (2009–2016), to test how the chemical sensitivity of OH and ozone production has changed as a result of the drought. To do this, I use isoprene and ozone data collected by the U.S. EPA. I also compare isoprene observations to emissions used to constrain the widely-used chemical transport model GEOS-Chem. Understanding these results is an integral step in making effective controls and policies for tropospheric ozone pollution.
Aircraft Observations of Nitrous Oxide (N2O) in the San Joaquin Valley of California
Shunta Muto, Tufts University
Agriculture is the largest source of anthropogenic nitrous oxide (N2O) in the U.S. While it is generally known which processes produce N2O, there is considerable uncertainty in controls over N2O emissions. Factors that determine N2O fluxes, such as soil properties and manure management, are highly variable in space and time, and, as a result, it has proven difficult to upscale chamber-derived soil flux measurements to regional spatial scales. Aircraft observations provide a regional picture of the N2O spatial distribution, but, because N2O is very long-lived, it is challenging to attribute measured concentrations of N2O to distinct local sources, especially over areas with complex and integrated land use. This study takes advantage of a novel aircraft N2O dataset collected onboard the low-flying, slow-moving NASA C-23 Sherpa in the San Joaquin Valley (SJV) of California, a region with a variety of N2O sources, including dairies, feedlots, fertilized cropland, and industrial facilities. With these measurements, we link observed N2O enhancements to specific sources at sub-inventory spatial scales. We compare our results with area-weighted emission profiles obtained by integrating detailed emission inventory data, agricultural statistics, and GIS source mapping.
Remote Sensing of the Coastal Ocean and Near Shore Processes
Faculty Advisor: Dr. Raphael Kudela, University of California Santa Cruz | Research Mentor: Henry Houskeeper, University of California Santa Cruz
Dried Out: Phytoplankton Response to Drought in San Francisco Estuary
Tyler Dawson, Northern Arizona University
Between 2012 and 2016, the state of California experienced one of the most severe multiyear droughts in nearly 120 years, causing a drastic reduction of freshwater flow to the San Francisco Estuary (SFE). During this period, the lack of winter rains and spring snow melt, coupled with high retention by dams, led to roughly a third less water reaching the SFE. Decreased freshwater flow to the bay alters the ecology of the SFE, for example by advancing the seasonal timing of phytoplankton blooms (Raimonet et al., 2016), and has been linked to phytoplankton plumes of different, and often more toxic, species. Phytoplankton functional type (PFT) methods, such as PHYDOtax (Palacios et al., 2015), enable the measurement of community composition, and has been validated in SFE. We will test the accuracy of the PHYDOtax algorithm during the drought period in SFE using matchups between in situ pigment measurements and remotely sensed reflectance spectra from the AVIRIS airborne sensor. We will present time series of salinity and phytoplankton composition in the SFE and evaluate the effects of the drought on the estuarine phytoplankton composition. California is expected to experience increased frequency of extreme weather events, such as drought, as a consequence of climate change. We evaluate the consequences of the drought on phytoplankton community composition to understand how future extreme weather events may alter the ecology or toxicity of SFE.
Analyzing Phytoplankton Composition Response to Oceanic Mesoscale Eddies
Nicole Cosenza, Pennsylvania State University
The area in the lee of the Hawaiian Islands is conducive to mesoscale eddy formation due to the interaction between the trade winds and the islands. Cyclonic activity promotes upwelling of deeper, nutrient-rich waters and has been shown to fuel phytoplankton blooms. In situ ocean measurements of a cyclonic mesoscale eddy, Cyclone Opal, were gathered in 2005 to investigate the physical properties of eddies in this region. Ocean color remote sensing has indicated increased chlorophyll-a concentration corresponding to Opal’s known position and dates. However, oceanic productivity and export are dependent on size-class and physiology of the phytoplankton community, and whether the heightened chlorophyll resulted in a phytoplankton community shift is unknown. Numerous methods to distinguish phytoplankton types have been validated in oceanic waters. Here we present a time series of MODIS chlorophyll-a images and the results of an abundance-based method for measuring phytoplankton composition (Hirata et al., 2011) to describe changes in community structure during mesoscale eddy propagation.
Application of Sentinel-3 to Detecting Cyanobacterial Blooms
Kimberly Dawid, North Central College
Certain species of cyanobacteria (blue-green algae) have been found to produce a hepatotoxin known as microcystin, which is linked to liver cancer and disease in humans (Kudela et al., 2014). Humans are exposed to microcystin by various mechanisms: ingesting tap water from areas of cyanobacterial harmful algal blooms (CyanoHABs), accidental ingestion during recreational activities, and through the consumption of seafood exposed to microcystin. Satellite and airborne sensing platforms are useful for detecting cyanobacterial blooms, but since only certain species of cyanobacteria produce microcystin toxin, an algorithm targeting phycocyanin pigment is used to differentiate between toxic and non-toxic blooms. High concentrations of phycocyanins, associated with CyanoHABs are detectable from remote sensing data using wavelengths 615-630 nm (Ogashawara et al., 2013). Frequency of CyanoHABs detected with the MERIS satellite platform has been linked to incidence of nonalcoholic liver disease (Zhang et al., 2015), but many inland water bodies conducive to CyanoHAB events are too small to be detected using legacy platforms like MERIS. This study evaluates the potential to measure phycocyanin in smaller water bodies using the newly operational OLCI sensor aboard Sentinel-3. We investigate the occurrence of non-alcoholic liver disease in counties with low estimated CyanoHAB events previously measured by legacy platforms to test potential benefits from improved spatial resolution to CyanoHAB monitoring for human health.
Remote Sensing Spectra of Oceanic Microplastics
Svea Southall, University of Alaska, Fairbanks
About 8 million tons of plastic are dumped into the ocean annually. Many plastics break down into smaller pieces, or microplastics, which release chemical contaminants into the water are consumed by marine wildlife. These microplastics are long-lived, but difficult to quantify. Obtaining an accurate estimation of the quantity of microplastics in the world’s oceans is difficult because in-situ plastic surveys are expensive, cover small areas, and cannot sample at the frequency required to monitor global plastic accumulation in the oceans. A method enabling ocean remote sensing of marine debris would allow the evaluation and monitoring of plastic pollution throughout the global ocean and the ability to track changes through time, but so far, no robust tool for measuring plastics in the oceans exists. Laboratory measurements of submerged microplastics suggest that high plastic concentrations may correspond to an elevated near-infrared signal in water-leaving radiances (Estess 2010). Here we evaluate the remote sensing spectra of oceanic microplastics by comparing non-aerosol-corrected satellite retrievals (MODIS aqua) with microplastic concentrations measured in the North Pacific subtropical gyre during a Sea Education Association research cruise. By comparing MODIS reflectance data with observed in-situ concentration of plastics, We test the feasibility of remotely sensing the presence of plastics in what has been called the “Pacific Garbage Patch,” where plastics concentrate in the Gyre. This would enable satellite remote sensing to be used on large scales to detect and quantify plastics abundance in open ocean water.
The Effect of the 2014-15 El Niño and The Blob on Giant Kelp Canopy in the Santa Barbara Channel, California
Sarita Quimpo Chiu, Smith College
Giant kelp (Macrocystis pyrifera) is a foundation species found in temperate coasts, forming thick forests, which provide shelter and food to numerous organisms. This makes them a vital component in food webs and structures of many coastal ecosystems, which in turn support commercial fisheries and food production. The Santa Barbara Channel, an upwelling ecosystem, hosts numerous patches of kelp forests where they thrive in the cool, nutrient-rich waters. A multiyear north Pacific warm anomaly followed by one of the strongest ENSO events on record heightened temperatures in the Santa Barbara channel and led to prolonged, decreased nutrient levels. Here we show a time series of kelp canopy measured using the Normalized Difference Vegetation Index from Landsat 5 and 7 imagery during the extended warm anomaly and subsequent El Niño. We examine how the extended period of heightened sea surface temperatures and decreased nutrients affected giant kelp canopy in the Santa Barbara Channel, and we test canopy cover against multiple basin-scale oscillations to evaluate their combined impacts on kelp survival and recruitment.
Mapping the decline of California Bull Kelp using Multiple Endmember Spectral Mixing Analysis of Landsat Data
Avery Renshaw, Towson University
Bull kelp (Nereocystis luetkeana), a species of canopy-forming brown macroalga dominant in the coastal US Pacific Northwest, provides critical ecological services such as habitat for a diverse array of marine species, nutrient regulation, photosynthesis, and regional marine carbon cycling. Starting around 2014, annual aerial surveys of bull kelp forests along California’s northern coastline conducted by the California Department of Fish and Wildlife (CDFW) have reported a sudden 93% reduction in bull kelp canopy area. Remote sensing using satellite imagery is a robust, highly accurate tool for detecting and quantifying the abundance of the canopy-forming giant kelp, Macrocystis pyrifera; however, it has not been successfully applied to measuring northern bull kelp forests. One of the main difficulties associated with bull kelp detection via satellite is the small surface area of bull kelp canopies. As a result, bull kelp beds often only constitute part of a satellite pixel, making it difficult to obtain a kelp reflectance signal significantly different than water’s reflectance signal. Here we test a novel method for detecting and quantifying bull kelp canopy area using a multiple endmember spectral mixing analysis (MESMA) applied to Landsat 5 and Landsat 8 imagery from 2003-2016. Water and kelp spectral endmembers are selected along the northern California coastline from Havens Neck cape to Point Arena. MESMA results are ground truthed with the CDFW aerial multispectral imagery data. This project will present a satellite-based time series of bull kelp canopy area and evaluate canopy change in a northern California kelp ecosystem.
The effects of the 2013-2015 warm water anomaly on the productivity of the Santa Barbara Channel
Shaina Wilburn, Mississippi State University
The 2013-2015 warm water anomaly, “The Blob,” was an atmospheric and oceanic phenomenon that resulted in shifts of the dynamics and biology of the Northeastern Pacific Ocean. This anomaly, caused by an area of persistent high pressure, stretched from the Gulf of Alaska to Baja California and impacted the phytoplankton productivity in the Santa Barbara Channel. As ocean temperatures fluctuate, anomalous warm conditions provide an opportunity to understand how phytoplankton might respond to future changes in oceanic and atmospheric coupling, such as nutrient availability via curl-driven or wind-driven upwelling. Because phytoplankton are the base of the oceanic food chain, changes in their abundance and composition alter the structure of higher trophic levels. We present sea surface temperature anomalies, surface wind fields, and chlorophyll-a anomalies during the Blob, and assess how physical shifts in an upwelling ecosystem alter phytoplankton productivity. This project will evaluate the effects of the 2014-15 high-pressure atmospheric anomaly on the sea surface wind fields across the upwelling-driven Santa Barbara channel ecosystem to understand how future weather anomalies could change the structure of the food chain in similar ecosystems.
Effect of changing sea surface temperature on phytoplankton composition in the Santa Barbara Channel
Alexandra Jones, Temple University
Since the mid-twentieth century the California Current System has experienced a long-term rise in water column temperature, along with recent signals of increased Chlorophyll-a (Chl-a) and dinoflagellate concentrations. Ocean color remote sensing enables the study of sea surface temperature and phytoplankton on a large scale, although coastal systems remain a challenge due to their optical complexity. The Santa Barbara Channel lies in the lee of Point Conception, which partially shelters the region from strong equatorward winds that flow along the central California coastline. We examine the impact of changing ocean temperatures on phytoplankton communities in this complex coastal system using a remote sensing abundance-based approach that partitions Chl-a concentrations by phytoplankton size class to estimate the underlying biomass composition in the Santa Barbara Channel (Hirata et al., 2011). We validate our remote sensing phytoplankton classification method using matchups with in situ time series of phytoplankton abundance, and perform a regional spatial analysis of Chl-a and phytoplankton composition in the Santa Barbara Channel to test how phytoplankton may respond to future ocean temperature shifts in coastal upwelling ecosystems.
Vegetation Drought Response in the Visible-Shortwave-Infrared and Thermal Infrared
Faculty Advisor: Dr. Dar Roberts, University of California Santa Barbara | Research Mentor: David Miller, University of California Santa Barbara
Effect of Drought on Riparian Vegetation in the Santa Clara River Basin Using Imaging Spectroscopy
Claire Schmidt, Knox College
Riparian forests are groundwater-dependent ecosystems that are highly sensitive to changes in the water table. As climate change continues, droughts are likely to become more frequent and severe in Southern California, threatening the persistence of these ecosystems. From 2012 to 2017, California experienced the most severe drought in the past century, providing a case study to assess drought impacts. Using imagery collected by the Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) from 2013 and 2016, we evaluated changes in riparian forest health in a central section of the Santa Clara River Basin, a multi-use river located just north of Los Angeles. We used Iterative Endmember Selection (IES) to select image endmembers of green vegetation (GV), non-photosynthetic vegetation (NPV), and rock/soil. We then used Multiple Endmember Spectral Mixture Analysis to estimate endmember fractions. We assessed changes in GV and NPV cover and canopy water content at key monitoring sites using cover fractions, the Normalized Difference Water Index, and the Water Band Index. Results showed a decrease in fractional cover of GV and an increase in fractional cover of NPV in riparian forests. Individual forest stands reacted differently to the drought, where larger drops in water table correlated with less fractional cover of GV. These results demonstrate the potential for severe droughts to decimate riparian forests, urging the creation of more sustainable groundwater management plans that consider the health of riparian forests.
Analyzing Vegetation Cover Change in Response to Drought in Southern California using Imaging Spectrometer Data
Amber Golini, California Polytechnic State University San Luis Obispo
The recent drought in California from 2012 to 2017 was the most severe drought event in the last century, driven by record high temperatures and significantly reduced precipitation. Such prolonged, intense drought conditions can have adverse impacts on plant health and can destabilize ecosystem functioning. Airborne imaging spectrometer data provide the opportunity to develop spatially detailed assessments of the impacts of drought on terrestrial vegetation at regional scales through repeated acquisitions. This imagery data can be used to quantify biophysical properties of the land surface, providing a greater understanding of ecosystem processes and improving ecological modeling methods. In this study, we analyzed vegetation cover change in a region of Santa Barbara County by comparing vegetation cover in 2011 (pre-drought) and in 2017 (post-drought). We used reflectance data in 7.5 meter pixels from the Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) to create a post-drought plant species classification map to compare with an existing 2011 classification dataset. A spectral library was built based on the unique spectral profile of each plant species in the imagery and was optimized using Iterative Endmember Selection (IES). We used spectral mixture analysis to generate a map of the 16 dominant plant species and rock and soil classes, as well as a fractional cover map to identify green vegetation (GV), non-photosynthetic vegetation (NPV), and soil. The results showed pronounced changes in the spatial distribution and appearance of plant species throughout the study area, and changes in fractional cover of GV and NPV due to the drought were observed. This study emphasizes the need for further analysis to quantify land cover changes and identify species most impacted by severe drought conditions. Examining vegetation changes in response to drought can provide valuable information for managing natural resources and can provide greater insights into the effects of drought on terrestrial ecosystems.
Patterns in Burn Severity by Pre-fire Plant Functional Type in the 2016 Rey Fire
Sarah Thalheimer, Smith College
Wildfires can have dramatic and lasting effects on vegetation at landscape scales. With fire frequency increasing due to drying climatic conditions in places such as California, it is becoming increasingly important to understand the spatial impacts of wildfires over time. Remote sensing datasets can show land cover composition and changes that are time-intensive and difficult to assess from the ground at the large scales needed for wildfires. This study examines differences in burn severity and its relationship to pre-fire vegetation cover for the 2016 Rey Fire near Santa Barbara, California. We developed a pre-fire fuels map of vegetation divided by similar resource usage and other properties as classified by plant functional types (such as evergreen broadleaf trees). Airborne Visible / Infrared Imaging Spectrometer (AVIRIS) imagery was spectrally unmixed with Multiple Endmember Spectral Mixture Analysis (MESMA) using a spectral library optimized by Iterative Endmember Selection (IES). The differenced Normalized Burn Ratio (dNBR) was derived from Landsat 8 Operational Land Imager (OLI) imagery to assess the immediate post-fire burn severity. Burn severity was also assessed on the same imagery by post-fire land cover class fractions (green vegetation (GV), soil, non-photosynthetic vegetation (NPV) and ash) as determined from MESMA using a spectral library optimized by Constrained Reference Endmember Selection (CRES). Within the fire perimeter, dNBR values ranged from -200, indicating plant growth in an unburnt region, to 1100, indicating extreme burn severity, but more typical sections of the burn scar dNBR values ranged from 100 to 600. Initial analyses showed that evergreen broadleaf and needleleaf shrubs were more intensely burned relative to evergreen broadleaf trees or annual herbs.
An Active Fire Temperature Retrieval Model Using Hyperspectral Remote Sensing
Keegan Quigley, Brown University
Wildfire is both an important ecological process and a dangerous natural threat that humans face. In situ measurements of wildfire temperature are notoriously difficult to collect due to dangerous conditions. Imaging spectrometry data has the potential to provide some of the most accurate and highest temporally-resolved active fire temperature retrieval information for monitoring and modeling. Recent studies on fire temperature retrieval have used have used Multiple Endmember Spectral Mixture Analysis (MESMA) applied to Airborne Visible / Infrared Imaging Spectrometer (AVIRIS) bands to model fire temperatures within the regions marked to contain fire. However, these methods are less effective at coarser spatial resolutions as linear mixing methods are degraded by saturation within the pixel. The assumption of a distribution of temperatures within pixels allows us to model pixels with an effective maximum and likely minimum temperature. This assumption allows a more robust approach to modeling temperature at different spatial scales. In this study, instrument-corrected radiance is forward-modeled for different ranges of temperatures, with weighted temperatures from an effective maximum temperature to a likely minimum temperature contributing to the total radiance of the modeled pixel. Effective maximum fire temperature is estimated by minimizing the Root Mean Square Error (RMSE) between modeled and measured fires. The model was tested using AVIRIS data collected over the 2016 Sherpa Fire in and the 2007 Zaca Fire, both in Santa Barbara County, California. While only in situ experimentation would be able to confirm active fire temperatures, the fit of the data to modeled radiance can be assessed, as well as the similarity in temperature distributions estimated at different spatial resolutions. Results show that this model improves upon current modeling methods in producing similar effective temperatures at multiple spatial scales as well as a similar modeled area distribution of those temperatures.
Ten Years of Post-Fire Vegetation Recovery following the 2007 Zaca Fire using Landsat Satellite Imagery
Jacklyn Hallett, University of Utah
Forest fires play a key role in shaping ecosystems. The risk to vegetation depends on the fire regime, fuel conditions (age and amount), fire temperature, and plant adaptations such as bark thickness and stem diameter. The 2007 Zaca Fire (24 km NE of Buellton, Santa Barbara County, California) burned 826.4 km2 over the course of 2 months. In this study, we used a time series of Landsat 5 Thematic Mapper (TM) and Landsat 8 Operational Land Imager (OLI) imagery to evaluate plant burn severity and post fire recovery as defined into classes of above average recovery, normal recovery, and below average recovery. We spectrally unmixed the images into green vegetation (GV), non-photosynthetic vegetation (NPV), soil surface (SOIL), and ash with a spectral library developed using Constrained Reference Endmember Selection (CRES). We delineated the fire perimeter using the differenced Normalized Burn Ratio (dNBR) and evaluated changes in this index and the Normalized Difference Vegetation Index (NDVI) through time. The results showed an immediate decline in GV and NPV fractions, with a rise in soil and ash fractions directly following the fire and a slow recovery in GV fraction and a loss of bare soil cover at later dates. There was a sharp increase in the ash fraction following the fire and gradual decrease in the year after. Most areas have recovered as of 2017, with prominent recovery around the perimeter of the fire scar and reduced recovery in the center. These results indicate how post-fire vegetation varies based on initial burn severity and pre-fire GV and NPV fractions.
Heat Reduction Capabilities and Air Pollution Mitigation of Urban Tree Species
Renee Baker, North Park University
Cities are predicted to house 61% of the world’s population within the next 30 years and their ecological footprints extend much farther than the extent of developed land. They trap 15-30% more radiation than their rural counterparts making urban areas 0.6 – 1.3° C warmer. In this study, we looked at the thermal effects and pollution mitigation potential of tree species in Santa Barbara, California using data from the Airborne Visible Infrared Imaging Spectrometer-Next Generation (AVIRIS-NG) on June 3, 2014. We mapped the 33 most common tree species in the city area in addition to non-photosynthetic vegetation (NPV), soil, and impervious surfaces using Multiple Endmember Spectral Mixture Analysis (MESMA). We developed a spectral library using image endmembers from reference polygons, and optimized the library using Iterative Endmember Selection (IES). We compared the urban tree map with land surface temperature data acquired on June 4, 2014 from the MODIS/ASTER airborne simulator (MASTER) to evaluate patterns between dominant tree species and temperature reduction. We found that increases in the fractional cover of urban tree species such as coast live oak (Quercus agrifolia), blue gum eucalyptus (Eucalyptus globulus),California sycamore (Platanus racemose), and avocado (Persea americana) were correlated with temperature reductions in surrounding areas. In addition, we compared species cover and air pollution reduction using literature values by species and plant functional type to provide examples of the direct effects of urban vegetation on the ecology of cities.
Characterizing Drought Impacted Soils in the San Joaquin Valley of California Using Remote Sensing
Leila Wahab, Rice University
California’s San Joaquin Valley is an extremely agriculturally productive region of the country, and understanding the state of soils in this region is an important factor in maintaining this high productivity. In this study, we quantified changing soil cover during the drought and analyzed spatial changes in salinity, organic matter, and moisture using unique soil spectral characteristics. We used data from the Airborne Visible / Infrared Imaging Spectrometer (AVIRIS) from Hyperspectral Infrared Imager (HyspIRI) campaign flights in 2013 and 2014 over the San Joaquin Valley. A mixture model was applied to both images that identified non-photosynthetic vegetation (NPV), green vegetation (GV), and soil cover fractions through image endmembers. We optimized the spectral library used to identify these classes with Iterative Endmember Selection (IES), and the images were unmixed using Multiple Endmember Spectral Mixture Analysis (MESMA). Maps of soil electrical conductivity, organic matter, soil saturated moisture, and field moisture were generated for the San Joaquin Valley based on indices developed by Ben-Dor et al. (2002). Representative polygons were chosen to quantify changes between years. Maps of spectrally distinct soils were also generated for 2013 and 2014 to determine their spatial distribution and temporal dynamics between years. We estimated that soil cover increased by 16% from 2013 to 2014. Six spectrally distinct soil types were identified for the region, and we determined that the distribution of these soil types was not constant for most areas between 2013 and 2014. We observed similar changes in soil pH, electrical conductivity, and soil moisture in the region between 2013 and 2014.
Drought-induced Bishop Pine Mortality on Santa Cruz Island, California
Madison Lichak, Barnard College at Columbia University
Widespread plant mortality has been observed following the California drought of 2012-2017, and many conifer trees experienced massive delayed mortality events due to a gradual decline in water available in deep soils. Bishop Pine (Pinus muricata) is a drought-intolerant species endemic to the western coast of the USA that has been particularly affected by the most recent drought. In this study, we sought to map the extent of Bishop Pine mortality on Santa Cruz Island to better understand mortality patterns of this rare species in a unique maritime ecosystem. Using Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) data from the HyspIRI campaign, and Multiple Endmember Spectral Mixture Analysis (MESMA), we mapped the change in the spatial range of living and dead Bishop Pine between 2014 and 2016. Results showed extensive mortality in Bishop Pine stands across Santa Cruz Island, even in areas previously thought to be better able to withstand the effects of drought. The maritime environment of Santa Cruz Island provides a buffer against summer drought, but the prevailing climatic conditions were likely insufficient to prevent widespread die-off during the prolonged drought. These results suggest that future droughts have the potential to cause widespread mortality in mainland stands of Bishop Pine further north, regardless of climatic buffers present.
Whole Air Sampling
Faculty Advisor: Dr. Donald Blake, University of California, Irvine | Research Mentor: Stacey Hughes, University of California, Irvine
Estimating Potential Ozone Formation from Agricultural Crop, Dairy, and Oil and Natural Gas Emissions in the San Joaquin Valley, CA
Natasha Dacic, The College of Idaho
The San Joaquin Valley (SJV) in California is a major contributor to the state’s oil and natural gas (ONG) production and to United States agriculture (dairy and crops). Despite these sectors’ economic success, they emit volatile organic compounds (VOCs) which indirectly help to produce tropospheric ozone (O3) that is harmful to human health and vegetation. The SJV is a nonattainment area and had 112 exceedances for national O3 standards in 2016. Thus, these sources and location merit further study of local contribution to tropospheric O3. Whole air samples collected aboard the NASA C-23 Sherpa during the Student Airborne Research Program (SARP) were analyzed for various VOC mixing ratios and used to estimate potential O3 formation by calculating hydroxyl (OH) reactivity. We assume the emissions from Tulare and Kings County are similar because both are agriculturally analogous and don’t have many active ONG wells whereas an i-pentane/n-pentane ratio of ~1 is indicative of ONG influence which separates Kern County (1.43 ± 0.45) from the two. The calculated OH reactivity (ROH) for agriculture was ROH = 1.77 s-1 ± 0.19 s-1 and for ONG was ROH = 1.95 s-1 ± 0.63 s-1. In addition, agricultural contribution to OH reactivity was 36% VOCs, 17% methane, and 47% carbon monoxide, while ONG contribution was 48% VOCs, 16% methane, and 36% carbon monoxide, suggesting that the primary source of O3 in the SJV varies greatly with wind transport from north to south as supported by the NOAA HYSPLIT wind trajectories. Quantifying OH reactivity will illuminate the local source O3 production throughout the Central Valley and can provide insight into future mitigation strategies to improve air quality within this region.
Examining Dimethyl Sulfide Emissions in California's San Joaquin Valley
Daniel Huber, University of California-Davis
Dimethyl Sulfide (DMS) is a sulfur-containing compound that leads to the formation of aerosols which contribute to the formation of haze and fog. Whole air samples were collected on board the NASA C-23 Sherpa aircraft during the 2017 Student Airborne Research Program (SARP) over dairies and agricultural fields in the San Joaquin Valley. Analysis of the samples indicate average DMS concentrations of 23 +/- 9 pptv, with a maximum concentration of 49 pptv. When compared with DMS concentrations from previous SARP missions (2009-2016), 2017 by far had the highest frequency of elevated DMS in this region. For this study, agricultural productivity of this region was analyzed to determine whether land use could be contributing to the elevated DMS. Top down and bottom up analysis of agriculture and dairies were used to determine emission rates of DMS in the San Joaquin Valley. Correlations to methane and ethanol were used to determine that DMS emissions were strongly linked to dairies, and resulted in R2 values of 0.61 and 0.43, respectively. These values indicate a strong correlation between dairies and DMS emissions. Combined with NOAA HySPLIT back trajectories and analysis of air samples collected on the research aircraft and on the ground, results suggest that the contribution of dairies to annual DMS emissions in the San Joaquin Valley exceeds those from DMS-emitting crops.
Investigating Elevated Concentrations of Dimethyl Sulfide in the Coachella Valley
Charles Fite, University of North Carolina at Charlotte
Dimethyl sulfide (DMS) is primarily emitted to the atmosphere by marine environments, agriculture, and dairy farms, and is important to study because its oxidation has been shown to contribute to the formation of sulfate aerosols. Whole air samples were collected in Southern California during the 2017 NASA Student Airborne Research Program and an average DMS concentration of 12 ± 11 pptv was observed during flights over land. However, elevated DMS concentrations as high as 63 pptv were observed over an area of dense agriculture in the northern area of the Coachella Valley. Wind trajectories originating from the Los Angeles Basin were determined using the NOAA HYSPLIT model. Since DMS has a short lifetime of 1-2 days, results indicate that despite HYSPLIT data originating from LA, the DMS was emitted locally. Comparisons to local marine trace gases imply that the elevated DMS was not solely emitted from the nearby Salton Sea. Correlations with acetone and ethyl nitrate suggests that the elevated DMS was likely also a result of various terrestrial sources. Through analyzing DMS and methane correlations, dairies were excluded as a potential source. However, investigations into potential land use change in the region since the 2014 NASA SARP campaign confirmed an increase in crop acreage for citrus plants, grapes, and aquaculture. A notable increase was also observed in DMS concentrations, from 10 ± 8 pptv in 2014 to 30 ± 20 pptv in 2017. Further analysis using ground sample comparisons between DMS with alpha-pinene and beta-pinene also exclude citrus farms in the area as a potential source. Findings confirm that the elevated DMS was influenced by both the Salton Sea and the local agriculture. Ruling out the unlikely sources provided a better understanding of terrestrial emissions, which gives insight into the understanding of the impact of sulfate aerosol formation and its effect on the radiative forcing budget of Earth’s climate.
Investigating the Air Pollution in Joshua Tree National Park
Kathryn Bumiller, University of South Carolina
Joshua Tree’s ozone levels have risen above the EPA’s daily maximum standard of 65 ppb, 55 times in 2017, so far. It is imperative to understand why the park is so polluted and from where the pollution is originating to preserve our National Parks and improve local air quality. During the 2017 Student Airborne Research Program (SARP), students collected in situ measurements at different locations throughout the park. Whole air samples were collected and analyzed in an attempt to understand the pollution affecting Joshua Tree. Maximum ozone measurements on the western side of the park were reported as 84 ± 9 ppb, compared to 59 ± 6 ppb on the eastern side of the park on the day that in-situ measurements were taken. A gradient of pollution has been seen to form in the park from west to east. Time-series of ground samples at Black Rock and Keys View show that as the day progresses, alkane concentration decreases, and alkyl-nitrate concentrations increase, indicating photolysis may occur throughout the day. Additionally, non-methane hydrocarbon OH reactivity was calculated for the ground samples across the park. Oxygenates were calculated to account for 54% of the total OH reactivity in the park. Biogenics were calculated for 30% of total OH reactivity in the park, indicating that the park could be affected by an agricultural area. HySPLIT models validate these assumptions with wind backtrack trajectories leading to the Salton Sea area.
Estimating oil and natural gas influence in the LA Basin
Haberly Kahn, University of Rochester
In an effort to reduce California’s carbon footprint from electricity generation, natural gas has become a more prevalent fossil fuel. Los Angeles is an urban area that uses a significant amount of oil and natural gas. To study the emissions in the city and the contribution from oil and natural gas, whole air samples (WAS) were collected onboard the NASA C-23 Sherpa during the 2017 Student Airborne Research Program (SARP) mission on June 26, 2017. Within the boundary layer in the Los Angeles International Airport (LAX) region, the average methane mixing ratio (2100 ± 200 ppbv) and ethane mixing ratio (6 ± 5 ppbv) were significantly higher than their background concentrations, 1800 ± 100 ppbv and 1 ± 2 ppbv, respectively. This suggests that natural gas is leaking into the atmosphere from the Los Angeles basin. Additionally, the area around LAX yielded an i/n pentane ratio of 2.5 ± 0.1, similar to that of urban areas in California. Although there are urban emissions in the area, natural gas leakage in the LAX region is significant enough to alter the urban trace gas composition. Using the California Air Resources Board (CARB) carbon monoxide estimations for 2015, a top-down approach was conducted for the LA basin and determined that 0.2 ± 0.3 Tg/y of natural gas leaks in the LA basin (2 ± 3% of the total natural gas delivered annually and twenty times higher than the leakage rate reported by SoCalGas).
Investigating Elevated Concentrations of H2 in the LAX region
Philip Rund, University of Maryland, Baltimore County
The growing interest in hydrogen (H2) fuel cell vehicles has created a need to study the atmospheric H2 budget. While there is resounding agreement that hydrogen would escape into the atmosphere due to fuel transport/storage processes, there is disagreement over the amount that would be leaked in a hydrogen fuel economy. Leakage rate estimates range from 2% to 10% for total hydrogen production and transport. A hydrogen based energy infrastructure seems a viable clean alternative to oil because, theoretically, the only waste products are H2O and heat. However, hydrogen leads to the formation of water vapor, polar stratospheric clouds, and a decrease in stratospheric temperatures, which contribute to the depletion of stratospheric ozone. Whole air samples (WAS) collected aboard the NASA Sherpa C-23 during the Student Airborne Research Program (SARP) showed elevated concentrations of hydrogen near LAX (950 ± 110 ppbv) compared to global average concentrations of 560 ± 20 ppbv. Trace gas analysis along with wind trajectories obtained with the NOAA HySPLIT models indicate that the source of elevated mixing ratios was leakage from H2 fuel stations in the surrounding areas. Correlation and ratio analyses eliminate the potential for common photochemical sources of H2 in the LAX area. This project could elucidate new and potential factors that contribute to the global atmospheric hydrogen budget.
Investigation of varying methyl bromide concentrations in the LAX area
Mario Autore, George Mason University
Methyl Bromide (CH3Br) is a volatile organic compound (VOC) that is highly regulated due to its toxicity and its photolysis that leads to the formation of bromine; which is known to cause mass degradation of stratospheric ozone at a rate that is 100 time more efficient than chlorine. Whole air samples containing elevated levels of methyl bromide were collected over the Los Angeles International Airport (LAX) region. Over the runway, 31 ± 2 pptv of methyl bromide was observed, compared to background concentrations, 7 ± 2 pptv. Methyl bromide showed correlations to the following gases that can be used to identify specific sources: polybromomethanes, CO2, CO, and CH3CCl3. LAX is surrounded by both anthropogenic and biogenic sources of methyl bromide; but, from the correlations observed three likely sources are responsible for elevating CH3Br concentrations above the background levels. The three sources are saltwater marshes, biomass burning, and the Hyperion water treatment plant. Additional sources of methyl bromide in the LAX area are marine, which hosts major micro/macro algal blooms that produce CH3Br, and fumigation of crop fields. However, crop field fumigation leading to a spike in methyl bromide levels seem unlikely as HySPLIT trajectories indicate there would be considerable dilution of air masses coming from the nearest sites of possible fumigation. LAX is surrounded by multiple methyl bromide sources, but large fluxes from saltwater marshes, Hyperion water treatment processes, and biomass burning have the potential to cause the variation in methyl bromide concentrations observed over LAX.
Investigating Enhanced Coastal Methyl Nitrate Concentrations Near LAX
Sean Leister, University of New Mexico
Methyl nitrate (CH3ONO2) is an important atmospheric trace gas that plays a significant role in the reactive nitrogen budget and ozone production. It is primarily produced through two main mechanisms: outgassing of ocean surface waters and atmospheric photochemistry. During the 2017 NASA Student Airborne Research Program (SARP), whole air samples were collected onboard the NASA C-23 Sherpa over California’s Central Valley and LA Basin. Samples collected over the LAX region during missed approaches had elevated concentrations of methyl nitrate, up to 20 pptv. Correlations between methyl nitrate concentrations and other marine trace gases throughout the day suggest that marine emissions played a role in the observed enhancements of methyl nitrate near LAX. In the summer of 2016, the first leg of the Atmospheric Tomography Mission (ATom 1) flew around the world to study the impacts of anthropogenic pollution on the remote environment. Global methyl nitrate concentrations were analyzed using whole air samples, and average background concentrations measured over deep Pacific waters (6 ± 1 pptv) were found to be much lower than those examined in the coastal marine regions of Southern California during SARP (13 ± 4 pptv). The vast disparity between measured methyl nitrate concentrations over deep ocean waters and coastal marine regions indicate that measurements near coastal environments overestimate background methyl nitrate concentrations, which can potentially result in inaccurate models, studies, and errors in the reactive nitrogen budget. This study will provide insight into these discrepancies to further improve global monitoring of marine trace gases and the understanding of the reactive nitrogen budget.
Fire and Flood: Analysis of 4 years of Drought, Major Fires and Catastrophic Debris Flows
Faculty Advisor: Dr. Dar Roberts, UC Santa Barbara | Research Mentor: Alana Ayasse, UC Santa Barbara
Fire Risk Assessment on the Wildland Urban Interface (WUI) of Goleta, California pre-Holiday Fire (2018) using Imaging Spectrometer Data
Corey Walker, Berea College
Fire hazard maps are important for promoting the safety of homeowners and firefighters and providing useful information to policymakers. In 2008, the California Department of Forestry and Fire Protection (CAL FIRE) created a fire hazard map for the state. This map considered broad variables of fire hazard such as weather patterns, slope and fuel levels. On July 6th, 2018 the Holiday Fire occurred on the Wildland Urban Interface (WUI) of Goleta, California burning over 40 hectares including 28 structures. However, CAL FIRE’s 2008 map did not identify Goleta as having high fire hazard. In this study, I created a fire risk assessment of Goleta using Airborne Visible / Infrared Imaging Spectrometer (AVIRIS) data taken 16 days prior to the Holiday Fire with 6.5m resolution. I built a spectral library that was optimized using Iterative Endmember Selection (IES). I then used Multiple Endmember Spectral Mixture Analysis (MESMA) to map land cover and classify the area into low, medium and high areas of risk. Areas of low fire risk were mapped around roads and soil. Areas of medium fire risk were mapped around green vegetation (GV). High levels of fire risk were mapped around non-photosynthetic vegetation (NPV). The Holiday Fire parcels showed high levels of fire risk. Medium regional values of fire risk were found north of the WUI while low values were found to the south. This study shows that highly specific and valid assessments can be developed around fire risk using a multifaceted remote sensing approach.
Mapping Impervious Surface Fraction of the Santa Barbara Front Range
Richelle Cabatic, University of Oregon
Understanding the dynamics of a watershed is critical for estimating ground water change, stream flow, and ocean runoff. The amount of runoff occurring in a system is highly correlated with the surface properties of the watershed because different surfaces have different infiltration capacities. Impervious surfaces do not allow any infiltration and therefore increase runoff. Urban impervious surfaces, such as pavement, are well studied and understood, however, less attention has been given to the natural impervious surfaces such as exposed rock. We predict that given a consistent rainfall rate, more natural impervious surfaces such as exposed rock, would correlate to more overall runoff in a watershed. In addition, accurate maps of exposed rock are rarely produced and not often used in hydrological models. In this study, we used Airborne Visible/Infrared Imaging Spectrometer Next Generation (AVIRIS NG) data collected in 2014 to map exposed rock in the Santa Barbara front range. We implemented Multiple Endmember Spectral Mixture Analysis (MESMA) to compute fractional cover of rock, soil, green vegetation (GV), and non-photosynthetic vegetation (NPV). We then used Long Term Ecological Research (LTER) watershed data for precipitation, discharge rates, slope, and area of each watershed within our study site and compared the LTER to our impervious surface fractions. We found a positive correlation between impervious surface cover and runoff rates per watershed. We speculate that there are other parameters such as green vegetation cover, and slope that may also affect runoff rates in a given watershed.
Mapping the Fire Front of the Thomas Fire Using MODIS, Landsat, and AVIRIS Active Fire Data
David McLeod-Warrick, Principia College
California is subject to numerous wildfires every year, and recently the fire season has been increasing. In December of 2017, the Thomas Fire burned over 114,078 hectares and destroyed 1,063 structures while claiming 15 lives. In order to understand the dynamics of these fires, we need to study fire movement on a day-to-day basis. One of the unique qualities about the Thomas Fire was that multiple sensors, both satellite and airborne, were used to take measurements of the fire. In this project, I set out to map the fire front of the Thomas Fire using Airborne Visible / Infrared Imaging Spectrometer (AVIRIS) at a 16.5-m resolution , the Operational Land Imager (OLI) at 30-m resolution, and the Moderate Resolution Imaging Spectroradiometer (MODIS) which has 1-km resolution in the thermal infrared. All 3 of my sensors have different temporal resolution. MODIS had daily measurements, OLI had 16-day and AVIRIS which has variable sampling. I created an active fire product for both AVIRIS and Landsat-OLI and used NASA’s active fire product for MODIS. I compared the active fire product for AVIRIS on December 6th and 7th, Landsat OLI on December 9th and 25th to MODIS on the same dates. MODIS overestimated the maximum area of the fire by over 600% when compared to both Landsat-OLI and AVIRIS. However, the temporal resolution of MODIS allowed us to track the fire movement every day. This shows that we need a combination of high spatial and high temporal resolution sensors to accurately map fire fronts.
Detecting Hotspots Ahead of Wildfires and their Correlation to Wind and Fuel Types
Corey Sargent, Old Dominion University
Hot spots are characteristics of a fire in which multiple subsequent ignitions are present ahead of the primary fireline. This can increase the fires burn rate, ignite fires in locations the fire would not have naturally traversed, and seriously complicate firefighting efforts. Our objective was to study this phenomenon and evaluate potential factors promoting spotting, including wind conditions and fuel type. We manually mapped hot spots using the Airborne Visual/Infrared Imaging Spectrometer (AVIRIS) data acquired during the Thomas Fire with the hyper-spectral fire detection index (HFDI) and a radiance threshold in the short-wave infrared (SWIR). These hot spot locations were combined with a land cover classification model built from LANDSAT data to determine the associated fuel types, and wind data from local weather stations were plotted and analyzed as well. Our results conclude that wind conditions, rather than fuel types, are the primary indicators of hot spot probability in our study area.
A Partial Recovery Analysis of the Thomas Fire Scar
Meredith Grames, University of Missouri
The Thomas Fire in the winter of 2017-2018 was the largest wildfire in California history. As wildfires have become more common in recent years, it is important to better understand the impact that these fires have on the environment, and to gain an understanding of how well the environment can recover after disturbance by wildfire. This study aims to assess the partial recovery of an area that burned in the Thomas Fire and determine whether the amount of recovery has any correlation with burn severity or dominant plant type in an area. Data from before the fire (June 26, 2017), immediately after the fire (December 21, 2017), and then six months later (June 20, 2018) were used to study the change in vegetation cover as a result of the fire and environmental recovery. Images of the study area were taken with the Airborne Visual/Infrared Imaging Spectrometer (AVIRIS). The land cover and green vegetation fractions before and after the fire were modeled using Multiple Endmember Spectral Mixture Analysis (MESMA). Burn severity was calculated using Difference Normalized Burn Ratio (dNBR). Fractional cover change was used to analyze the amount of recovery that had taken place in the six months following the fire. For the pre-fire data set, the vegetation cover was classified into chaparral, trees, and grassland and burn severity and recovery were analyzed in the context of dominant plant type in an area. Analysis of the data shows grassland has the highest amount of recovery in the six months following the fire and lowest burn severity. In contrast, areas of chaparral and oak woodland had higher burn severities and lower recovery amounts, although recovery did take place in those areas.
Determining Biomass for Wildfires Using Hyperspectral Imaging
Tyler Welty, University of Alabama in Huntsville
Wildfires are one of the most destructive and common threats to people and property in Southern California. An important attribute of wildfires is the fuel available. Knowing the amount of fuel present helps predict fire intensity and post fire recovery. Since fires burn organic matter, measuring biomass can be an effective way to know the potential fuel available. Remote sensing can be used to estimate biomass values over a large region and aid in understanding wildfire and its potential destruction. To study the feasibility of remotely measuring biomass for wildfires, we used Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) data from before and after the 2017 Thomas Fire in Santa Barbara and Ventura counties, CA. The pre-fire image was classified into Chaparral, Oak Woodland, Senesced (non-photosynthetic) Grassland, and Soil classes using a 2-endmember spectral mixture analysis (SMA). Known biomass values for each class were drawn from literature. The Normalized Difference Vegetative Index (NDVI), Normalized Differenced Water Index (NDWI), and Water Band Index (WBI) were calculated for the pre- and post-fire imagery. Linear regression was then used to create models based on each index for estimating biomass. Burn severity was also estimated using the differenced Normalized Burn Ratio (dNBR). The results of the NDVI analysis showed that an estimated 88% of biomass within the study area was lost due to the fire. The NDWI and WBI analyses, however, estimated a 26%-35% loss and a gain of 80%-100% after the fire, respectively. Since much of the study area had high burn severity, the results for NDWI and WBI indicate that these are likely poor indices for estimating biomass. We also found that the dNBR did not correlate well with amount of biomass lost. Further study needs to be done to confirm the accuracy of remotely sensed biomass estimates as well as the most effective index for measuring biomass loss after fires.
Remote Sensing Analysis of the Montecito Debris Flows
Sara Nethercutt, Tulane University
The recent Thomas Fire in December 2017 was the largest wildfire in modern California history, burning over 114,000 hectares in Santa Barbara and Ventura counties before it was fully contained in January 2018. On January 9th, 2018, a cold front moved over Montecito, CA and dropped a record 18.54 mm of rain in 15 minutes which initiated massive debris flows that led to the loss of 23 lives and over $421 million in property damage. In order to better understand the factors that created these destructive flows, we utilized Airborne Visual/Infrared Imaging Spectrometer (AVIRIS) data from two days: a December 21, 2017 post-fire pre-debris flow image and a January 11, 2018 post-debris flow image, both covering the Montecito area. We used Multiple End Member Spectral Analysis (MESMA) to determine, on a sub-pixel level, the fractional cover of soil, ash, rock, green vegetation, and non-photosynthetic vegetation in the images. Fractional cover of ash calculated from MESMA was used to quantify burn severity in the December image. To estimate debris flow severity, a change detection was calculated on the soil fractional cover between the December and January images. An increase in soil fraction between the two images was used as a proxy for the mass movement of soil/mud, and this was successfully used to map multiple tracks of the debris flows. We then compared the burn severity, debris flow severity, watershed characteristics, and precipitation in Montecito to adjacent watersheds in Summerland/Carpenteria (where debris flows were less severe). The Montecito watersheds showed greater burn severity, greater ash fractional cover, greater 5-minute precipitation levels, and greater percent area of the watershed burned, pointing towards these variables as potential factors in determining the occurrence and severity of post-fire debris flows in the Santa Barbara County area.
Faculty Advisor: Dr. Sally Pusede, University of Virginia | Research Mentor: Laura Barry, University of Virginia
Observing Boundary Layer Heights over Mountainous Terrain Using Aircraft Vertical Profiles
Dallas McKinney, Western Kentucky University
The boundary layer height separates turbulently mixed air and pollutants emitted at the ground from the free troposphere above and is an important parameter in numerical weather prediction. Discerning the boundary layer height over mountainous terrain is difficult due to complex interactions with upper level winds, venting of humidity and aerosols into the free troposphere, and large spatiotemporal variability. Mountain boundary layers can closely follow the terrain, be flat, or be shallower than valleys below depending on the time of day, season, and synoptic conditions. To determine the extent to which the boundary layer height follows terrain, I used aircraft meteorological and trace gas data collected onboard the NASA DC-8 during accents and descents over mountains across Central and Southern California. Meteorological parameters considered include water vapor, potential temperature, and turbulence. Carbon monoxide, methane, and carbon dioxide enhancement ratios were used to consider when observed atmospheric layers were last in contact with the ground surface. Vertical profiles over the southern Sierra Nevada, San Emigdio, and San Bernardino Mountains indicate that boundary layer heights closely follow the topography of these areas, being higher over ridges and lower over valleys.
Calculating Surface Energy Budgets Utilizing the Airborne Eddy Covariance Method
Sierra Laltrello, Mississippi State University
Global climate change may alter the magnitude and dynamics of energy fluxes between the surface and atmosphere. The airborne eddy-covariance technique gives an opportunity to observe surface energy fluxes over large regions and in locations without established research towers. To increase observational constraints on the surface energy balance, I derive sensible and latent heat fluxes using the airborne eddy-covariance technique over two different land cover types in the San Joaquin Valley of California, a large agricultural area and the city of Bakersfield, and in the late morning and afternoon. I use aircraft observations of 3-D winds, temperature, and water at 20Hz, 10Hz, and 1Hz, respectively, collected onboard the NASA DC-8 and interpret observed differences in the surface energy budget as a function of location and time of day.
Using Enhancement Ratios for Source Apportionment of Greenhouse Gas Emissions near Bakersfield, CA
Chris Wright, Pomona College
In California’s San Joaquin Valley, anthropogenic sources of long-lived greenhouse gases (CO2, CH4, and N2O) are various, distributed, and, consequently, difficult to distinguish. Due to the region’s complex land use, sources from livestock operations, farming, oil extraction, and urban traffic are all co-located on spatial scales of tens of kilometers. This study aims to understand the contribution of several source types have on atmospheric composition. Sources can be identified through enhancements ratios, as they emit distinct ratios of these long-lived greenhouse gases and other pollutants. Using in-situ airborne data from the NASA DC-8 aircraft, I identify enhancement ratios for agriculture, urban traffic sources, and biomass burning and adapt a multivariate mixing model to quantify the contributions of these sources along the DC-8 flight track. I then use HYSPLIT backward dispersion trajectories to calculate dilution factors over known source areas to augment the spatial resolution of my apportionment.
N(2)O Emissions as a Function of Animal Agriculture Density in the California San Joaquin Valley
Laura Paye, University of Maine
Nitrous oxide (N2O) is a long-lived greenhouse gas with a global warming potential almost 300 times that of carbon dioxide and the dominant destroyer of stratospheric ozone. While the free troposphere concentration of N2O is well-constrained, there are large uncertainties in both the identity of N2O sources and in controls over variability in these sources. N2O emissions are particularly difficult to quantify because of their high spatiotemporal heterogeneity, with sources often characterized as hotspots and/or hot moments. Because of this variability, it has proven difficult to upscale research conducted using chamber measurements, which gather information at meter to tens of meter spatial scales. Aircraft data are useful because they provide insight into N2O sources on kilometer spatial scales. I use data collected onboard the NASA DC-8 to look at spatial patterns in N2O concentrations and its agricultural sources over the California San Joaquin Valley, a place of dense agricultural activities including crop cultivation, feedlots, and dairies. I correlate measured N2O concentrations with methane (CH4) and feedlot and dairy location to determine relationships between N2O and cattle operations and discuss factors driving the observed correlations.
Observational Constraints on Atmospheric Chemical Production of Formaldehyde over Bakersfield, CA
Madeline Miles, Appalachian State University
Formaldehyde (CH2O) plays an important part role in atmospheric oxidation chemistry, including the production of the harmful air pollutant tropospheric ozone (O3). O3 production is driven by the oxidation of numerous different gaseous organic compounds, most of which will at some point be oxidized to CH2O. Because the total concentration and composition of gaseous organic compounds is often unknown, CH2O provides valuable insight into O3 production rates. In this study, I use airborne CH2O data collected from onboard the NASA DC-8 over the city of Bakersfield, California, one of the most O3-polluted cities in the U.S. I observationally derive the CH2O production rates using changes in CH2O along upwind to downwind transects across the Bakersfield urban plume. Comparisons of morning and afternoon transects, as well as data at four different altitudes (500m, 800m, 1200m, and 1800m) are also used. I link inferred CH2O production rates to O3 production over this area.
Investigating the Health Impacts of Changes in Ozone in Central and Southern California
Hannah Zuercher, Lewis University
Ozone (O3) air pollution is harmful to human health, affecting rates and exacerbation of respiratory diseases, hospital admissions, and mortality. O3 health impacts are predicted to worsen, as O3 production in Central and Southern California is temperature-dependent and global climate change is projected to increase regional temperatures. Simultaneously, as California’s population ages, residents become more vulnerable to respiratory diseases. Understanding how O3 affects mortality is critical to effectively shaping pollution control policies. Aircraft data is valuable in such assessments as it provides high spatial resolution data necessary for evaluating impacts below the county level. Using airborne data collected on the NASA C-23 Sherpa in 2017, I examine the potential health effects due to changes in O3 exposure and age in Central and Southern California using the EPA’s Environmental Benefits Mapping and Analysis Program (BenMAP). First, I model future mortality as a result of a 5 ppb increase in O3 concentrations and an artificially-aged population. Second, I decrease high O3 concentrations with the artificially-aged population to the World Health Organization recommendation of 50 ppb to simulate the results of a policy change on mortality incidence.
Reducing ozone air pollution through an urban forestry project in Bakersfield, California
Arie Feltman-Frank, University of Denver
Tropospheric ozone (O3) is a harmful pollutant regulated by the U.S. Environmental Protection Agency. O3 is formed by complex chemical reactions that involve sunlight, volatile organic compounds (VOCs), and nitrogen oxides (NOx). O3 formation is non-linear and O3 chemistry within a region can be classified as either NOx– or VOC-limited. In NOx-limited regions, reductions in NOx lead to reductions in O3. In VOC-limited regions, reductions in NOx lead to increases in O3. Therefore, to effectively regulate O3, it is important to know whether local O3 chemistry is NOx or VOC sensitive. Formaldehyde (HCHO) is a short-lived oxidation product of many VOCs, making it a useful indicator of total VOC reactivity. The ratio of HCHO to NO2 is an accurate indicator of O3 sensitivity to NOx and VOCs. Low ratios suggest VOC-limited areas and high ratios suggest NOx-limited areas. Trees remove O3 and NOx by uptake through their leaf stomata and by reactions on surface waxes. In order to reduce O3 pollution, tree planting in urban and surrounding areas may be effective, but also has the potential to worsen O3 pollution as trees emit O3-forming VOCs. In this study, I use aircraft and ground-based data to examine the local variability of the HCHO to NO2 ratio over Bakersfield, California. The HCHO to NO2 ratio and wind patterns are used to determine the location of a NOx-limited region downwind of Bakersfield optimal for tree planting. I then quantify NO2 and O3 reductions for low, moderate, and high tree planting scenarios based on percentages of suitable land cover within this NOx-limited region. My work suggests that the HCHO to NO2 ratio can play a key role in urban forestry planning and associated land management decisions, especially in regions with O3 pollution problems.
Remote Sensing of the Coastal Ocean and Near-Shore Processes
Faculty Advisor: Dr. Raphe Kudela, UC Santa Cruz | Research Mentor: Henry Houskeeper, UC Santa Cruz
Historical Declines of Northern California Bull Kelp Canopy Following El Niño Southern Oscillation Events
Dennis Finger, University of California, Berkeley
While field surveys and aerial imagery can provide detailed biological data, satellite remote sensing enables consistent, long-term imagery over a large spatial scale. Such imagery enables us to track changes in vegetation over time, monitoring how environmental conditions have affected ecosystems. Long-term time series from satellite imagery have been developed for giant kelp in Southern California, but there have been no such surveys for bull kelp. Bull kelp is a Northern California brown algae species that supports fisheries and provides a habitat for keystone species. Decreases in bull kelp canopy cover are believed to result from warm waters, low nutrient levels, and intense storms, conditions typically associated with El Niño Southern Oscillation (ENSO). Using Multiple Endmember Spectral Mixture Analysis (MESMA) of Landsat imagery, we test bull kelp classification methods using California Department of Fish and Wildlife (CDFW) aerial bull kelp surveys for validation. Of MESMA models using 30, 19, 7, and 3 seawater endmembers, we find that the 7-endmember model records the lowest Root-Mean-Square Error (0.2352; 3.69%) during CDFW validation. Using the 7-endmember model on Landsat imagery, we construct a 36-year time series of bull kelp canopy coverage in Northern California. Canopy reductions and recovery are compared to ENSO and other basin-scale oscillations. We find that bull kelp decreased sharply during historical strong ENSO events, notably during 1982-1983, 1997-1998, and 2015-2016. We also found, however, that bull kelp recovered quickly from these events, as years with large decreases in canopy coverage were frequently followed by years with large bull kelp canopy increases.
Remote Sensing of Ocean Microplastics in the Near-Infrared (NIR): A Feasibility Study
Jandlyn Bentley, Bridgewater State University
Since plastic pollution began entering the ocean in the 1950s, trends have shown that this pollution has been and will continue to increase in the future. Once plastics reach the ocean they never fully biodegrade; instead they break down into smaller and smaller pieces until they reach the microplastic stage. The only methods to gather ocean plastic data are by plankton net tows, which are costly, inefficient, and time consuming, and computer-generated models. A reliable method to locate floating ocean plastic via remote sensing has yet to be discovered. Although the ocean absorbs nearly all light beyond the 700 nm range, suspended microplastics are expected to backscatter light in the near-infrared (NIR) range resulting in non-negligible signals. This project investigates the potential to detect ocean microplastics with remote sensing based on elevated reflectance within the NIR region. We present modeled inherent optical properties of plastics in water derived from above-water radiance measurements to evaluate expected signals for water with high concentrations of microplastics. We also develop matchups between satellite imagery and in-situ plastic abundance measurements from the Sea Education Association’s dataset in a region that has a statistically higher prevalence of microplastics.
The Spatial Distribution of Dinoflagellates in Retention Zones of the California Coastline
Jessica Ganim, University of Delaware
Dinoflagellates constitute 75% of all harmful algal species and are the phytoplankton functional type responsible for producing the vast majority of toxic red tides. Areas along the California coastline, known as retention zones, are popular seeding sites for dinoflagellates, particularly during the autumn, from August to November. Retention zones are confined areas downstream of capes or headlands where cold, nutrient-rich water becomes trapped beneath a shallow, distinct thermocline, in lieu of a contiguous active upwelling system. The stratification and localized mixing of these “upwelling shadows” traps heat and microorganisms. Dinoflagellates’ ability to vertically migrate deeper into the water column to retrieve nutrients at depth provides them with a competitive advantage in areas where nutrient-rich water is not upwelled to the sea surface. Airborne and satellite measurements (AVIRIS and HICO) of retention zones in Monterey Bay, Santa Barbara, Point Reyes, and Santa Monica, with Newport Beach used as a non-retention baseline, were processed using PHYDOtax, an algorithm that discriminates phytoplankton functional types. We present advection models derived using the General NOAA Operational Modeling Environment (GNOME) software and compare with sea surface temperature and salinity profiles to evaluate the potential of retention zones for seeding broader regions of the California coastline.
Evaluating phytoplankton size class shifts in Monterey Bay using remote sensing and in-water imaging platforms
Mackenzie Devine, Piedmont College
Phytoplankton are the foundation of the oceanic ecosystem, and the composition of different phytoplankton functional groups is relevant to many ocean biogeochemical cycles. Remote sensing methods provide novel tools to measure phytoplankton size, a key metric for understanding ecological structure. Using remote sensing methods, we are able to measure and analyze a larger area than when using in-situ methods, but with lower accuracy. Different phytoplankton cell shapes and sizes are potential sources of error for satellite size class estimates. We compare multiple remote sensing algorithms that discriminate phytoplankton sizes and groups (e.g. diatoms, dinoflagellates) with in-water imagery of phytoplankton cells to evaluate whether changes in phytoplankton size and shape correspond to shifts in remote sensing size retrievals. We validate the size class algorithms derived using MODIS-Aqua and AVIRIS imagery from Monterey Bay during June 2018 using matchups with the Imaging Flow Cytobot, an in-situ flow cytometer that collects water from the environment and takes images of particles containing chlorophyll. Snapshots from the Imaging Flow Cytobot allow real time observation of phytoplankton cells during environmental perturbations and provide a rich dataset for evaluating remote sensing algorithms.
Tracking the Montecito Mudslide Plume and Response via Remote Sensing in the Santa Barbara Channel
Jack Dechow, Knox College
In January 2018, Santa Barbara, California suffered from a devastating combination of rainfall, historic drought, and fires, leading to the Montecito Mudslide. The year preceding the mudslide had anomalously low precipitation, and during December 2017 the Thomas fire killed off large amounts of vegetation that normally protected soils on the sloping hills of Santa Barbara County. When the torrential January rains arrived, local rivers and creeks were overwhelmed, and large amounts of debris ended up on the CA 101 highway and in residential neighborhoods. To remove the debris, several hundred thousand cubic meters of debris were dumped on local Santa Barbara County beaches, with the expectation that ocean waves and tides would remove the debris. Here we present remote sensing imagery from MODIS-Aqua to track the plume of the dumped debris. We test how long the perturbation persisted in the region and estimate whether the retention and flushing observed would be comparable if the dumping occurred during different months or in different locations. Our results suggest that the plume dispersed within the Santa Barbara channel within weeks of the dumping event. Advection by currents within the Santa Barbara Channel is compared with dynamics of other, more retentive coastal zones.
From Fire to Flora: Effects of Ash from the Thomas Fire on the Biology of the Santa Barbara Channel in December 2017
Ariana Castillo, Texas A&M University
The Thomas Fire was the biggest fire in California state history, burning an astounding 201,893 acres. Based on NASA’s MODIS (Moderate Resolution Imaging Spectroradiometer) Aqua imagery, the event filled the Santa Barbara Channel (SBC) with a smoke plume, transporting ash towards the marine environment. Although the effect of ash from fires on marine ecosystems is unknown, volcanic ash has been shown to fertilize oceanic environments with nutrients and stimulate the growth of diatoms and other phytoplankton. Nutrient over-enrichment, which may occur from ash deposition, may favor specific phytoplankton groups. But little work has been done to study the implications that ash from wildfires has on the biology of nearby ocean regions. Since ash deposition brings nutrients to the oceans, we use the Thomas Fire as an opportunity to evaluate the effects of the changing atmospheric and land environments on the ocean ecosystem. Due to the lack of in situ data, ash deposition is difficult to quantify accurately. Hence, the smoke plume track was evaluated with MODIS imagery and with the NOAA model HYSPLIT (Hybrid Single Particle Lagrangian Integrated Trajectory). Advection of deposited ash in the ocean was estimated using the GNOME (General NOAA Operational Modeling Environment) model, which utilizes dispersion properties of surface-associated materials (originally designed for oil) while taking meteorological conditions and SBC currents into consideration. Furthermore, a cruise coincidentally collected data under the location where the smoke plume was prominent, confirmed by satellite observations and model results of the plume itself. An Imaging Flow Cytobot deployed onboard observed species uncharacteristic for the region during that season. We analyze phytoplankton community during the ash fertilization event and compare with phytoplankton climatology for the region to test whether the ash conditions led to biological anomalies in the SBC. Because fires frequently deposit ash into rivers, lakes, and coastal waters, understanding the role of ash in fertilizing phytoplankton blooms is important to understanding the importance of fires on biogeochemical processes of marine and aquatic ecosystems.
Environmental Analysis of Proposed Changes to the Los Angeles/ Long Beach Traffic Separation Scheme
Kasey Castello, Rensselaer Polytechnic Institute
As the only remaining supertanker berth on the West Coast, the Port of Long Beach is one of the largest marine oil traffic routes in the United States. Each day, at least 50 million gallons of oil arrive by ship somewhere in the Los Angeles County region. This vessel traffic transits heavily along the Western Approach and Southern Approach elements of the Traffic Separation Scheme (TSS) set for the area. Both of these port access approaches require that vessels transit near areas deemed by federal, state and local governments as Marine Protected Areas (MPAs). These MPAs include the Channel Islands National Marine Sanctuary, Catalina Island South Coast MPA, and beaches in Santa Barbara, Ventura, Los Angeles and Orange counties. Currently, proposals to modify the TSS regions exist, and if implemented would alter the proximity of some ship traffic to the MPAs. This study evaluates potential environmental impacts and benefits from modification of the TSS regions, for example in the case of a major oil spill event. Using a spill trajectory experiment within the General NOAA Operational Modeling Environment (GNOME) software, we compare modeled spill trajectories between existing and proposed TSS regions, and find reduced likelihood of beaching of spilled particles after modification of the TSS lanes. We then analyze the GNOME data using ArcMap geoprocessing tools to determine which protected areas would be at risk for a spill at each location. Finally, we test adherence of large tanker ships to current TSS requirements using Automatic Identification System (AIS) data. The findings provide additional environmental rationale for modifications to the TSS in order to protect MPAs in the event of an oil spill within the Western Approach element of the Long Beach / Los Angeles TSS.
Whole Air Sampling
Faculty Advisor: Dr. Donald Blake, University of California, Irvine | Research Mentor: Chris Woods, University of California, Irvine
Hydroxyl-Radical Removal: Los Angeles and Oildale
Katie Kloska, University of Kentucky
As pollutants are emitted, removal processes prevent their accumulation in the atmosphere. One removal process is oxidation via the hydroxyl radical, OH. Commonly referred to as the “detergent of the atmosphere” the hydroxyl radical is highly reactive and is a major sink for greenhouse gases and pollutants such as carbon monoxide and methane. Previous studies indicate that CO and methane are the main removal processes for OH. Thus, by increasing carbon monoxide and methane emissions, it is likely that the lifetime of other OH removed gases may increase due to the loss of an oxidizing sink. In this study, the relative reaction rates of 36 compounds measured on NASA airborne science missions from 2012-2018 in Los Angeles and Oildale, California are compared in order to determine the effect that different compounds have on the reactivity of OH. Calculated at STP, relative reaction rates are combined with mixing ratios to determine the percent of OH consumed by each compound. These results show that at altitudes less than 3500 feet, carbon monoxide emissions react with hydroxyl on average 6% more in Los Angeles than in Oildale. In addition, at altitudes less than 3500 feet, methane emissions react with hydroxyl on average 7% more in Oildale. Acetaldehyde, a compound used in the manufacture of acetic acid and also the product of ethanol oxidation in the atmosphere, was found to react with 14% and 20% of hydroxyl in Los Angeles and Oildale. Altitudinal data depict significant changes in relative hydroxyl consumption. These data may aid in the determination of effective emission reduction proposals. Future estimations should include OH photolysis and reactions with NOx, which are additional sinks of hydroxyl that will increase the accuracy of these estimations.
Dimethyl Sulfide Chemistry at the Salton Sea, CA
Mara Nutt, Mills College
Formed anthropogenically in 1905 the Salton Sea, CA is a saline endorheic lake. Dimethyl sulfide (DMS) is an organic sulfide compound, mainly emitted by marine organisms, which oxidizes and forms reactive sulfur compounds. The reaction of DMS + OH → Products is known to draw-down hydroxyl (OH) radical concentrations in the atmosphere and produce sulfuric acid (H2SO4). Ground samples were taken downwind of the Salton Sea in five locations using 2-Liter stainless steel cans by the National Aeronautics and Space Administration (NASA) Student Airborne Research Program whole air sampling group in June 2018. These samples were then analyzed at the Rowland-Blake Laboratory at University of California, Irvine. The Salton Sea was found to have mixing ratios of 80-280 pptv DMS, greatly enhanced compared to airborne samples. In order to determine if DMS is drawing down the OH radical, a relative rate percentage was calculated with R[DMS]/R[CO]. A minimum reaction ratio of 5% indicated that DMS does draw down the OH radical at the Salton Sea via the DMS oxidation reaction, which allows for the formation of H2SO4. Using pseudo-first order assumptions, the amount of H2SO4 created by the DMS on the ground was calculated to be a maximum of 0.8 µg/m3 or 205 pptv. This is lower than the California Air Resources Board’s twenty-four hour average ambient air quality standard for sulfates of 25 µg/m3. Hence, the H2SO4 that is a product of the DMS oxidation reaction at the Salton Sea is not harmful to the local community. Future studies should take whole air samples at ground locations around, as well as on, the Salton Sea, every hour for two to three days in order to determine DMS diurnal cycles. In addition, NASA DC-8 flights should take airborne samples to track DMS chemistry at higher altitudes.
Impact of Dimethyl Sulfide Emitted from the Salton Sea on Regional Air Quality
Katie Cush, Emory University
On June 25, 26, and 27, 2018, the NASA Student Airborne Research Program collected whole air samples throughout Southern California using stainless steel canisters while aboard NASA’s DC-8. Ground samples were also collected around the Salton Sea, located in Southern California, on June 28 and 29, 2018. Samples were analyzed in the Rowland-Blake Laboratory at the University of California, Irvine using gas chromatography and mass spectrometry. Historically, large mixing ratios of dimethyl sulfide (DMS) were found along the shores of the Salton Sea. However, in 2018 DMS was greatly enhanced relative to previous Imperial Valley sampling. Dimethyl sulfide is a biological sulfur compound produced indirectly by phytoplankton. In the air, sulfate aerosols produced by DMS photochemistry can negatively impact human respiratory health and act as cloud condensation nuclei, influencing climate. In this study, we used HYSPLIT modeling (NOAA) to track the forward trajectory and the dispersion of sulfate aerosols so as to better understand how they travel through the atmosphere and impact the Salton Sea region. We found that DMS produced by the Salton Sea does not travel over any populated areas during the summer months; however, the HYSPLIT model showed that during the winter, DMS moves directly over populated areas in the region. We calculated that the DMS produced by the Salton Sea contributes an upward bound of 4% to the total Salton Sea’s annual regional aerosol load of 8 ug/m3, which has been recorded by California’s Air Resource Board. Because of the high mixing ratios of DMS and the potential for Salton Sea DMS to impact regional health and climate, future studies should take more rigorous sampling at the Salton Sea, ensuring that samples are collected along the entire perimeter, during different seasons, upwind and downwind, as well as farther from the shore of the Salton Sea.
Tracing Sources of Light Alkyl Nitrates Observed at High Altitudes during SARP 2018
Kiersten Johnson, University of Alaska Fairbanks
Light alkyl nitrates, such as methyl nitrate (MeONO2) and ethyl nitrate (EtONO2), are commonly attributed to being carriers of reactive nitrogen species that help to regulate tropospheric ozone. In previous studies, MeONO2 and EtONO2 have been traced back to a common oceanic source with data showing the ratio of MeONO2 to EtONO2 changes based on latitudinal position. In order to observe MeONO2 and EtONO2 mixing ratios, samples were collected on board the Student Airborne Research Program (SARP) DC-8 during the fourth research flight using the UC Irvine Whole Air Sampler. Samples taken during the flight were analyzed at the UC Irvine Rowland/Blake laboratory using gas chromatography and mass spectrometry. Results showed that at low altitudes (<26K feet) and the top of the Total Carbon Column Observing Network (TCCON) spiral the mixing ratios of MeONO2 to EtONO2 are consistent with expected data. However, during the TCCON descent there was an increase in the mixing ratio of MeONO2 and EtONO2 between ~36K feet and 25K feet. The ratio of MeONO2 to EtONO2 in this plume was ~7:1. The C2H2/CO atmospheric processing scale for each sample in the plume was calculated to determine the air samples were found to be <1 (ppt/ppb) indicating aged air. Through the utilization of HYSPLIT a back trajectory was developed to locate sources of air over a number of time intervals. The back trajectory coupled with an understanding of photolysis rate it was determined that the elevated light alkyl nitrate ratio was a result of aged well mixed air moving up from the equator and mixing with air from the Gulf of Mexico.
Estimating Long Range Transport Times of Air Masses Using Hydrocarbon Concentrations
Sujay Rajkumar, Temple University
Analyzing whole air samples from 2014 revealed elevated levels of pollutants 7.5-9.5 km above sea level at the Total Carbon Column Observation Network (TCCON) site at Edwards Air Force Base (Kern County, California). Gas chromatography analysis of these samples revealed the presence of common trace tropospheric gases such as carbon monoxide, carbon dioxide, methane and non-methane hydrocarbons (NMHC’s) at higher levels. The presence of chloroform at 740 ± 24 pptv in five samples was a good indicator that the source was coal. Bituminous coal has always been and still continues to be a cheap energy source. Despite its benefits, combustion of it produces greenhouse gases and volatile organic compounds (VOC’s). Using a model for depicting long range transport of hydrocarbons (Rudolph and Johnen, 1990), the times of transport of gases including ethane, ethyne, propane, butane and benzene were estimated. Using literature values for Chinese coal burning emissions and comparing ratios of the aforementioned hydrocarbons to another hydrocarbon, such as ethene, five separate transit times for the five individual samples were determined. Interestingly, the variance for the individual points was very small and in most cases less than 10%. There was a slight difference in the transit times for the air masses that contained the samples collected. The tight agreement in transit times for the hydrocarbons indicates that these gases were from the same source and they provided parameters for determining the locations of the emissions through HYSPLIT modeling. The results of HYSPLIT model suggests that the emissions originated from Eastern and Central China. Solvents known to be used in China (eg. 1,2-DCE) were also enhanced in the plume samples further suggesting the air mass was affected by Chinese emissions. This study indicates that surface emissions of VOCs from China can be transported to US airspace in less than one week.
Can Airborne Whole Air Sampling be used as a Compliance Tool for Air Quality Regulations?
Jack Biscupski, University of Iowa
Hydrochlorofluorocarbons (HCFCs) were developed as intermediate compounds to help wean the world off of chlorofluorocarbons (CFCs). Mandated through the Montreal Protocol and Clean Air Act (CAA), the EPA established guidelines and phase out plans for the usage of HCFCs in response to the rapid rise of HCFC use throughout the globe. Whole air samples collected on board the DC-8 during the NASA Student Airborne Research Program (SARP) 2018 campaign contained compelling data for HCFC-22, HCFC-141b, and HCFC-142b; three compounds considered by the EPA to be the worst offenders out of all HCFCs due to their high ozone depletion potentials (ODP). Investigation into local air quality data can be used as a regulatory tool to check ‘problem areas’ for compliance to EPA standards. For example, air data collected at Mt. Wilson in Los Angeles County during 2014 showed highly elevated concentrations in HCFC-142b until the start of January 2015, which directly coincided with the Clean Air Act’s 2015 phase out goal for HCFC-142b and HCFC-22. This is an excellent example of compliance. However, while analyzing this year’s data, concentration enhancements in mixing ratios for HCFC-22 and HCFC-141b were found around 33.8°N and 35.5°N. These elevated levels were located near the downtown LA area and Bakersfield. HCFC-22 levels reached up to 800 ppt; much higher than the 240 + 10 ppt background levels. HCFC-141b levels reached 50 ppt in certain areas, also much higher than its 24 + 2 ppt background levels. These values in close proximity to urban hubs suggest that these areas were emitting HCFCs. Emission information for these gases can be nebulous, and enforcement of this kind of policy has, historically, been difficult. In the future, local and/or federal regulatory agencies could utilize this sampling and analysis plan to improve their compliance monitoring techniques.
Potential Sources of Increased Halon-1301 in the Southern Central Valley of California
Christian Poutré, University of Massachusetts Amherst
Halon-1301 is a potent ozone depletor that is mainly used for fire suppression in aviation and data centers. Its production was phased out by January 1, 1994 under the Montreal Protocol, but stocks made before then are legal and still in use today. “Good-faith” regulations have been put in place by the EPA to manage Halon 1301’s use, storage and disposal. Over the course of June 25-27, 2018, whole air samples of the atmosphere were collected in and around the California Central Valley on board the NASA DC-8 during four SARP 2018 research flights. These samples were then analyzed in the Rowland-Blake laboratory at University of California Irvine using gas chromatography and mass spectrometry. Enhancements of Halon-1301 (~2 ppt above ~3.5 ppt background level) were found in the southern part of the Central Valley spanning a range of ~150 km N-S from Bakersfield to Five Points, CA. This enhancement is most likely either from civilian aviation in the Bakersfield area or military activity in the valley. Although Halon-1301 concentrations were high in and around Bakersfield, amounts of other refrigerants and CFC compounds normally associated with civilian aviation were not present in proportional concentrations to Halon-1301 outside of Bakersfield. The nonlinear association indicates that civilian aviation and airports are not the cause of this increase throughout the southern part of the valley. A number of Air Force and Navy bases are located in California, including the new Master Jet Base at Naval Air Station Lemoore which stations more than 200 aircraft and is currently expanding. We flew within 10 kilometers of it and observed increased concentrations of Halon-1301. Increased Halon-1301 concentrations could be attributed to the expansion of the Lemoore Naval Base, and be an indicator of increased military activity