Mountain Scholar :: Browsing by Author "Schumacher, Russ, committee member"

Browsing by Author "Schumacher, Russ, committee member"

Now showing 1 - 20 of 23

Open Access
A framework for real-time, autonomous anomaly detection over voluminous time-series geospatial data streams
(Colorado State University. Libraries, 2014) Budgaga, Walid, author; Pallickara, Shrideep, advisor; Pallickara, Sangmi Lee, advisor; Ben-Hur, Asa, committee member; Schumacher, Russ, committee member
In this research work we present an approach encompassing both algorithm and system design to detect anomalies in data streams. Individual observations within these streams are multidimensional, with each dimension corresponding to a feature of interest. We consider time-series geospatial datasets generated by remote and in situ observational devices. Three aspects make this problem particularly challenging: (1) the cumulative volume and rates of data arrivals, (2) anomalies evolve over time, and (3) there are spatio-temporal correlations associated with the data. Therefore, anomaly detections must be accurate and performed in real time. Given the data volumes involved, solutions must minimize user intervention and be amenable to distributed processing to ensure scalability. Our approach achieves accurate, high throughput classications in real time. We rely on Expectation Maximization (EM) to build Gaussian Mixture Models (GMMs) that model the densities of the training data. Rather than one all-encompassing model, our approach involves multiple model instances, each of which is responsible for a particular geographical extent and can also adapt as data evolves. We have incorporated these algorithms into our distributed storage platform, Galileo, and proled their suitability through empirical analysis which demonstrates high throughput (10,000 observations per-second, per-node) and low latency on real-world datasets.
Open Access
A new hurricane impact level ranking system using artificial neural networks
(Colorado State University. Libraries, 2015) Pilkington, Stephanie F., author; Mahmoud, Hussam, advisor; van de Lindt, John, committee member; Schumacher, Russ, committee member
Tropical cyclones are intense storm systems that form over warm water but have the potential to bring multiple related hazards ashore. While significant advancements have been made for forecasting of such extreme weather, the estimation for the resulting damage and impact to society is significantly complex and requires substantial improvements. This is primarily due to the intricate interaction of multiple variables contributing to the socio-economic damage on multiple scales. Subsequently, this makes communicating the risk, location vulnerability, and the resulting impact of such an event inherently difficult. To date, the Saffir-Simpson Scale, based off of wind speed, is the main ranking system used in the United States to describe an oncoming tropical cyclone event. There are models currently in use to predict loss by using more parameters than just wind speed. However, they are not actively used as a means to concisely categorize these events. This is likely due to the scrutiny the model would be placed under for possibly outputting an incorrect damage total. These models use parameters such as; wind speed, wind driven rain, and building stock to determine losses. The relationships between meteorological and locational parameters (population, infrastructure, and geography) are well recognized, which is why many models attempt to account for so many variables. With the help of machine learning, in the form of artificial neural networks, these intuitive connections could be recreated. Neural networks form patterns for nonlinear problems much as the human brain would, based off of historical data. By using 66 historical hurricane events, this research will attempt to establish these connections through machine learning. In order to link these variables to a concise output, the proposed Impact Level Ranking System will be introduced. This categorization system will use levels, or thresholds, of economic damage to group historical events in order to provide a comparative level for a new tropical cyclone event within the United States. Discussed herein, are the effects of multiple parameters contributing to the impact of hurricane events, the use and application of artificial neural networks, the development of six possible neural network models for hurricane impact prediction, the importance of each parameter to the neural network process, the determination of the type of neural network problem, and finally the proposed Impact Level Ranking System Model and its potential applications.
Open Access
A rhetorical storm: linguistic analysis of uncertainty in severe weather communication
(Colorado State University. Libraries, 2019) Rosen, Zoey, author; Long, Marilee, advisor; Demuth, Julie, committee member; Schumacher, Russ, committee member
Weather forecasts are a product with inherent uncertainty and a wide audience (Compton, 2018). Known as an example of prediction rhetoric (Morss, Demuth, & Lazo, 2008), weather forecasts have been found to be influenced by linguistic and cultural factors in case studies (Pennesi, 2007). However, forecasts are still rarely studied as articles of rhetoric (Compton, 2018). This study analyzed patterns amongst the linguistics of uncertainty expressions in Twitter forecasts during a cluster of tornadoes in March 2018 through a content analysis. Tornado hazard messaging, due to tornadoes' short-term threat and overarching potential for damage (Ripberger, Jenkins-Smith, Silva, Carlson, & Henderson, 2014), provides an opportunity to study uncertainty language during short-term hazardous scenarios. Across a five-day period, there were N = 2,459 severe weather forecast tweets from 146 Twitter users located in Alabama, Mississippi, Tennessee, and Georgia. Results indicate there were significant relationships between the source of a forecast (i.e., weather media, weather government, and non-weather government) and uncertainty expression. Weather media sources were significantly less likely than government sources (both weather and non-weather) to use uncertainty expressions in their forecast tweets. The state the Twitter source was located also influenced the amount of uncertainty expressed within a forecast. For example, tweets from areas with a greater number of tornadoes were significantly less likely to contain uncertainty expressions than were areas with fewer threats. Also, time (measured as the number of days before tornado touchdown) was shown to have a significant relationship with uncertainty expression, as the amount of uncertainty expressed decreased the closer in time the messages were to the tornadic event. Due to the large amount of uncertainty in weather prediction, meteorological forecasts during severe events provide a unique, fascinating area for future research on risk communication and public safety messaging.
Open Access
Analyzing risk-related information seeking behavioral intention and risk perception of wildfires: the High Park Fire Burn Area
(Colorado State University. Libraries, 2019) Mokry, Melissa M., author; Trumbo, Craig, advisor; Kim, Jangyul, committee member; Abrams, Katie, committee member; Hoffman, Chad, committee member; Schumacher, Russ, committee member
This study assessed risk-related information seeking behavioral intention and dual-process risk perception within the context of wildfires. Particularly, the study focused on utilizing a combined risk-related information seeking model with concepts originating from the planned risk information seeking model (PRISM), a framework of risk information seeking (FRIS), and the risk information seeking and processing model (RISP). The key concepts utilized included: past risk-related information seeking, self-efficacy, response efficacy, dual-process risk perception (affective and cognitive risk perception, perceived hazard knowledge, information needs, and behavioral intention. A survey (N=432; 60.8% response rate) was disseminated to the High Park Fire Burn Area, west of Fort Collins, Colorado which experienced a wildfire in 2012. The survey revealed the importance of including dual-process risk perception in risk-related information seeking models and highlighted its influence on past risk-related information seeking and risk-related information seeking behavioral intention. Response efficacy was correlated with self-efficacy, following suit to other risk-related information seeking studies. Cognitive risk perception was correlated with affective risk perception, suggesting a bi-directional relationship between the two concepts. Individuals were more likely to seek wildfire information in the past if they did not have enough knowledge about the hazard. Moreover, individuals are more likely to base their risk perception on their emotions, particularly when facing a wildfire. The results from the survey revealed that the exploratory path had a better model fit than the confirmatory path model, yet both provided important findings related to risk-related information seeking behavioral intention and dual-process risk perception. This study reaffirmed the need for theoretical improvement related to current information needs, particularly in relation with perceived hazard knowledge and risk-related information seeking behavioral intention. There were inconsistencies with current information needs throughout the study, following suit with the literature and calls for further refinement of the concept. Implications and future research efforts are also noted and discussed such as the importance of tailored messaging and a communication campaign.
Open Access
Application of an interpretable prototypical-part network to subseasonal-to-seasonal climate prediction over North America
(Colorado State University. Libraries, 2024) Gordillo, Nicolas J., author; Barnes, Elizabeth, advisor; Schumacher, Russ, committee member; Anderson, Chuck, committee member
In recent years, the use of neural networks for weather and climate prediction has greatly increased. In order to explain the decision-making process of machine learning "black-box" models, most research has focused on the use of machine learning explainability methods (XAI). These methods attempt to explain the decision-making process of the black box networks after they have been trained. An alternative approach is to build neural network architectures that are inherently interpretable. That is, construct networks that can be understood by a human throughout the entire decision-making process, rather than explained post-hoc. Here, we apply such a neural network architecture, named ProtoLNet, in a subseasonal-to-seasonal climate prediction setting. ProtoLNet identifies predictive patterns in the training data that can be used as prototypes to classify the input, while also accounting for the absolute location of the prototype in the input field. In our application, we use data from the Community Earth System Model version 2 (CESM2) pre-industrial long control simulation and train ProtoLNet to identify prototypes in precipitation anomalies over the Indian and North Pacific Oceans to forecast 2-meter temperature anomalies across the western coast of North America on subseasonal-to-seasonal timescales. These identified CESM2 prototypes are then projected onto fifth-generation ECMWF Reanalysis (ERA5) data to predict temperature anomalies in the observations several weeks ahead. We compare the performance of ProtoLNet between using CESM2 and ERA5 data. We then demonstrate a novel approach for performing transfer learning between CESM2 and ERA5 data which allows us to identify skillful prototypes in the observations. We show that the predictions by ProtoLNet using both datasets have skill while also being interpretable, sensible, and useful for drawing conclusions about what the model has learned.
Open Access
Atmospheric reactive nitrogen in Rocky Mountain National Park
(Colorado State University. Libraries, 2018) Shao, Yixing, author; Collett, Jeffrey L., advisor; Schumacher, Russ, committee member; Jathar, Shantanu, committee member; Benedict, Katherine, committee member
The Front Range urban corridor in Colorado, located east of Rocky Mountain National Park (RMNP), includes a variety of urban sources of nitrogen oxides, while high emissions of ammonia are found in agricultural sources on the eastern plains of Colorado. The spatial distribution and temporal variation of ammonia and other reactive nitrogen species in the region is not well characterized. Periods of upslope flow can transport atmospheric reactive nitrogen from the Front Range and eastern Colorado, contributing to nitrogen deposition in the park. Deposition of excess atmospheric reactive nitrogen in Rocky Mountain National Park poses threats to sensitive ecosystems. It is important to characterize temporal variation and spatial distribution of reactive nitrogen in the region to better understand the degree to which emission sources in the northeastern plains of Colorado impact RMNP and how meteorological conditions are associated with transport of ammonia to the park. Mobile and in-situ measurements of reactive nitrogen gases and particles were made between 2015 and 2016 in northeastern Colorado and RMNP. Gaseous ammonia was measured with high-time resolution instruments (a Picarro cavity-ring down spectrometer and an Air Sentry ion mobility analyzer); 24-hr integrated concentrations of trace gases and PM2.5 chemical composition in RMNP were measured by URG denuder/filter systems coupled with lab analysis; wet nitrogen deposition was collected with an automated precipitation collector followed by lab analysis. Model outputs from The Hybrid Single Particle Lagrangian Integrated Trajectory Model (HYSPLIT) was also included for examining transport of ammonia source plumes. Diurnal and seasonal variability of ammonia concentrations and some other reactive nitrogen species were characterized with high time-resolution measurement data. Repeating diurnal cycles were found in Greeley and RMNP. Ammonia concentrations usually increase in the morning and reach maxima around noon in RMNP, while at Greeley ammonia builds up during the night followed by a rapid decrease after sunrise. A seasonal pattern of ammonia levels was also revealed, with higher concentrations observed during summer. When combined with wind data it is clear that elevated ammonia levels in RMNP were associated with easterly transport from the eastern plains of Colorado. The median daily averaged ammonia concentrations measured in Greeley, Loveland and RMNP are 26.2 ppb, 6.3 ppb and 1.1 ppb respectively. Considerable ammonia variability was found in NE Colorado with higher concentrations measured close to CAFOs and source regions. This was particularly clear in mobile NH3 observations where distinct plumes of ammonia were observed away from confined animal feeding operation (CAFOs) sources. Spatial variations, particularly in the north-south direction, were observed to be strongly dependent on meteorology as highlighted by HYSPLIT back trajectories. This study also evaluates the pilot Early Warning System which informs agricultural producers of impending upslope events that are likely to transport nitrogen from eastern Colorado to the park, so that management practices may be implemented to reduce nitrogen emissions. The performance of the meteorological forecasting was evaluated using continuous measurements of atmospheric ammonia concentrations in the RMNP, as well as the wet nitrogen deposition data from 2015. It was found that the model showed skill in capturing some large wet nitrogen deposition events in the park.
Open Access
Blending model output with satellite-based and in-situ observations to produce high-resolution estimates of population exposure to wildfire smoke
(Colorado State University. Libraries, 2016) Lassman, William, author; Pierce, Jeffrey, advisor; Fischer, Emily, committee member; Schumacher, Russ, committee member; Magzamen, Sheryl, committee member; Pfister, Gabriele, committee member
In the western US, emissions from wildfires and prescribed fire have been associated with degradation of regional air quality. Whereas atmospheric aerosol particles with aerodynamic diameters less than 2.5 μm (PM 2.5 ) have known impacts on human health, there is uncertainty in how particle composition, concentrations, and exposure duration impact the associated health response. Due to changes in climate and land-management, wildfires have increased in frequency and severity, and this trend is expected to continue. Consequently, wildfires are expected to become an increasingly important source of PM 2.5 in the western US. While composition and source of the aerosol is thought to be an important factor in the resulting human health-effects, this is currently not well-understood; therefore, there is a need to develop a quantitative understanding of wildfire-smoke-specific health effects. A necessary step in this process is to determine who was exposed to wildfire smoke, the concentration of the smoke during exposure, and the duration of the exposure. Three different tools are commonly used to assess exposure to wildfire smoke: in-situ measurements, satellite-based observations, and chemical-transport model (CTM) simulations, and each of these exposure-estimation tools have associated strengths and weakness. In this thesis, we investigate the utility of blending these tools together to produce highly accurate estimates of smoke exposure during the 2012 fire season in Washington for use in an epidemiological case study. For blending, we use a ridge regression model, as well as a geographically weighted ridge regression model. We evaluate the performance of the three individual exposure-estimate techniques and the two blended techniques using Leave-One-Out Cross-Validation. Due to the number of in-situ monitors present during this time period, we find that predictions based on in-situ monitors were more accurate for this particular fire season than the CTM simulations and satellite-based observations, so blending provided only marginal improvements above the in-situ observations. However, we show that in hypothetical cases with fewer surface monitors, the two blending techniques can produce substantial improvement over any of the individual tools.
Open Access
Changes in the snowpack of the Upper Colorado River basin in a warmer future climate
(Colorado State University. Libraries, 2023) Sherman, Erin Alexys, author; Rasmussen, Kristen, advisor; Schumacher, Russ, committee member; Fassnacht, Steven, committee member
Water is a crucial factor to sustaining life on Earth. Snow acts as a reservoir for water, providing storage during the cold seasons and freshwater resources throughout the warmer months. Streamflow in the upper Colorado River Basin is primarily contributed by seasonal mountain snowmelt that provides critical freshwater resources to humans and wildlife, effectively connecting ecological, hydrological, and atmospheric systems. Global Climate Models (GCMs) and regional climate models do not represent the complex processes that can impact snowpack growth, evolution, and melting, thus they often rely on parameterizations to represent such processes. SnowModel is a high-resolution snowpack-evolution modeling system that can simulate processes such as blowing snow redistribution and sublimation, forest canopy interception, and snow-density evolution. To investigate how snowpack in the Upper Colorado Basin may change in a future warmer climate, high-resolution convection-permitting regional climate atmospheric model simulations at 4-km horizontal grid spacing are used to provide input conditions to drive SnowModel at 100-m in the current and future climate for 13 years. Results show that the average snow season will be shorter in the future, reducing the days that the snowpack can accumulate. In addition, analysis of the characteristics of precipitation in the simulations shows a ~150% increase in convective precipitation frequencies in the winter months, indicating shifts in the character of precipitation in a future climate. Liquid precipitation in winter increases ~200% in a future climate as a result of warmer air temperatures. In contrast, solid precipitation stays roughly the same in the winter, but decreases about 25 percent in the fall and spring. A case study analysis of the high-impact snowstorm on 17-19 March 2003 that delivered between 30-70 inches of snow along the Colorado Front Range in a current and future climate shows a shift from a snow-dominant to a rain-dominant event, as well as increases in moisture and convective precipitation frequencies. The simulated changes in the snowpack of the Upper Colorado River Basin will likely have detrimental impacts on freshwater resources and food production in a future climate that will undoubtedly impact a multitude of humans and ecosystems in the western United States.
Open Access
Characteristics of current and future flood-producing storms in the continental United States
(Colorado State University. Libraries, 2020) Dougherty, Erin M., author; Rasmussen, Kristen, advisor; Schumacher, Russ, committee member; Maloney, Eric, committee member; Morrison, Ryan, committee member
Understanding the changes to extremes in the hydrologic cycle in a future, warmer climate is important for better managing water resources and preventing detrimental impacts to society. The goal of this dissertation is to contribute to this understanding by examining the precipitation characteristics of flood-producing storms in the current climate over the continental United States (CONUS) and how these will change in a future, warmer climate. Numerous storm types are responsible for floods over the CONUS, so quantifying how their characteristics will change among a large number of flood-producing storms in the future provides a spectrum of possible changes and impacts to flood-prone regions across the country. To understand flood-producing storms in the current climate over the CONUS, a climatology of these storms from 2002–2013 is created by merging storm reports, streamflow-indicated floods, and Stage-IV precipitation data (Chapter 2). From this climatology, it is observed that flash flood-producing storms preferentially occur in the warm-season in the Mississippi River Basin, with intense rain rates and short durations. Slow-rise floods occur mostly during the cool-season, concentrated in the Ohio River Valley and Pacific Northwest, and are long-duration, low-intensity rainfall events. Hybrid floods, having characteristics of both flash and slow-rise flood-producing storms, tend to occur in the spring and summer notably in the central CONUS and Northeast, with moderate durations and rain rates. Examining these floods on a sub-basin scale in the Wabash and Willamette basins, precipitation and instantaneous streamflow correlations are spatially variable, with strong positive correlations in areas of complex terrain and urbanization (Chapter 3). These studies show that in the current climate, flood-producing storm precipitation characteristics and their hydrologic response is nuanced, which is critical to document in order to understand their behavior in a future climate. A subset of nearly 600 flash flood-producing storms from the Chapter 2 climatology are examined using high-resolution convection-permitting simulations over the CONUS to understand how these historical storms might change in a future, warmer climate (Chapter 4). Both precipitation and runoff show widespread increases in the future over the CONUS, increasing by 21% and 50%, respectively, with maximum hourly rain rates becoming more intense by 7.5% K−1. In California, 45 flood-producing storms associated with atmospheric rivers also display a future increase (decrease) in precipitation (snow water equivalent) leading to increased runoff, particularly over the Sierra Nevada Mountains, implying a shift in future water resources in California (Chapter 5). In the Mississippi River Basin–a flash flood hotspot in the CONUS––nearly 500 flash flood-producing storms exhibit a 17% average increase in precipitation and 32% average increase in runoff primarily associated with warm-season convection, and to a lesser extent, tropical cyclones (Chapter 6). When stratified by vertical velocity, the storms with the strongest vertical velocity in the current climate exhibit the greatest (least) increase (decrease) in future rainfall (vertical velocity), suggesting a potential role of storm dynamics in modulating future rainfall changes.
Open Access
Estimates of sublimation in the Upper Colorado River basin
(Colorado State University. Libraries, 2013) Phillips, Morgan, author; Cotton, William, advisor; Stednick, John, committee member; Schumacher, Russ, committee member
Snowpack stored in mountain environments is the primary source of water for the population of much of the western United States, and the loss of water through direct evaporation (sublimation) is a significant factor in the amount of runoff realized from snow melt. A land surface modeling study was carried out in order to quantify the temporal and spatial variability of sublimation over the Upper Colorado River basin through the use of a spatially distributed snow-evolution model known as SnowModel. Simulations relied on forcing from high resolution atmospheric analysis data from the North American Land Data Assimilation System (NLDAS). These data were used to simulate snow sublimation for several years over a 400 by 400 km domain in the Upper Colorado River Basin at a horizontal resolution of 250 m and hourly time-steps. Results show that total volume of sublimated water from snow varies 68% or between 0.95 x 107 acre feet in WY 2002 to the maximum of 1.37 x 107 acre feet in WY 2005 within the ten years of the study period. On daily timescales sublimation was found to be episodic in nature, with short periods of enhanced sublimation followed by several days of relatively low snowpack water loss. The greatest sublimation rates of approximately 3 mm/day were found to occur in high elevation regions generally above tree line in conjunction with frequent windblown snow, while considerable contributions from canopy sublimation occurred at mid-elevations. Additional sensitivity runs accounting for reduced canopy leaf area index as a result of western pine beetle induced tree mortality were also carried out to test the models sensitivity to land surface characteristics. Results from this comparison show a near linear decrease in domain total sublimation with reduced LAI. Model performance was somewhat satisfactory, with simulations underestimating precipitation and accumulated SWE, most likely due to biases in the precipitation forcing and errors in determining precipitation phase.
Open Access
Features based assessments of warm season convective precipitation forecasts from the high resolution rapid refresh model
(Colorado State University. Libraries, 2017) Bytheway, Janice L., author; Kummerow, Christian, advisor; Schumacher, Russ, committee member; Randall, David, committee member; Chandrasekar, V., committee member; Alexander, Curtis, committee member
Forecast models have seen vast improvements in recent years, via increased spatial and temporal resolution, rapid updating, assimilation of more observational data, and continued development and improvement of the representation of the atmosphere. One such model is the High Resolution Rapid Refresh (HRRR) model, a 3 km, hourly-updated, convection-allowing model that has been in development since 2010 and running operationally over the contiguous US since 2014. In 2013, the HRRR became the only US model to assimilate radar reflectivity via diabatic assimilation, a process in which the observed reflectivity is used to induce a latent heating perturbation in the model initial state in order to produce precipitation in those areas where it is indicated by the radar. In order to support the continued development and improvement of the HRRR model with regard to forecasts of convective precipitation, the concept of an assessment is introduced. The assessment process aims to connect model output with observations by first validating model performance then attempting to connect that performance to model assumptions, parameterizations and processes to identify areas for improvement. Observations from remote sensing platforms such as radar and satellite can provide valuable information about three-dimensional storm structure and microphysical properties for use in the assessment, including estimates of surface rainfall, hydrometeor types and size distributions, and column moisture content. A features-based methodology is used to identify warm season convective precipitating objects in the 2013, 2014, and 2015 versions of HRRR precipitation forecasts, Stage IV multisensor precipitation products, and Global Precipitation Measurement (GPM) core satellite observations. Quantitative precipitation forecasts (QPFs) are evaluated for biases in hourly rainfall intensity, total rainfall, and areal coverage in both the US Central Plains (29-49N, 85-105W) and US Mountain West (29-49N, 105-125W). Features identified in the model and Stage IV were tracked through time in order to evaluate forecasts through several hours of the forecast period. The 2013 version of the model was found to produce significantly stronger convective storms than observed, with a slight southerly displacement from the observed storms during the peak hours of convective activity (17-00 UTC). This version of the model also displayed a strong relationship between atmospheric water vapor content and cloud thickness over the central plains. In the 2014 and 2015 versions of the model, storms in the western US were found to be smaller and weaker than the observed, and satellite products (brightness temperatures and reflectivities) simulated using model output indicated that many of the forecast storms contained too much ice above the freezing level. Model upgrades intended to decrease the biases seen in early versions include changes to the reflectivity assimilation, the addition of sub-grid scale cloud parameterizations, changes to the representation of surface processes and the addition of aerosol processes to the microphysics. The effects of these changes are evident in each successive version of the model, with reduced biases in intensity, elimination of the southerly bias, and improved representation of the onset of convection.
Open Access
GeoLens: enabling interactive visual analytics over large-scale, multidimensional geospatial datasets
(Colorado State University. Libraries, 2015) Koontz, Jared, author; Pallickara, Sangmi, advisor; Pallickara, Shrideep, committee member; Schumacher, Russ, committee member
With the rapid increase of scientific data volumes, interactive tools that enable effective visual representation for scientists are needed. This is critical when scientists are manipulating voluminous datasets and especially when they need to explore datasets interactively to develop their hypotheses. In this paper, we present an interactive visual analytics framework, GeoLens. GeoLens provides fast and expressive interactions with voluminous geospatial datasets. We provide an expressive visual query evaluation scheme to support advanced interactive visual analytics technique, such as brushing and linking. To achieve this, we designed and developed the geohash based image tile generation algorithm that automatically adjusts the range of data to access based on the minimum acceptable size of the image tile. In addition, we have also designed an autonomous histogram generation algorithm that generates histograms of user-defined data subsets that do not have pre-computed data properties. Using our approach, applications can generate histograms of datasets containing millions of data points with sub-second latency. The work builds on our visual query coordinating scheme that evaluates geospatial query and orchestrates data aggregation in a distributed storage environment while preserving data locality and minimizing data movements. This paper includes empirical benchmarks of our framework encompassing a billion-file dataset published by the National Climactic Data Center.
Open Access
Grassland responses to seasonal shifts in water availability
(Colorado State University. Libraries, 2023) Hajek, Olivia Louise, author; Knapp, Alan K., advisor; von Fischer, Joseph, committee member; Cusack, Daniela, committee member; Schumacher, Russ, committee member
Climate change is altering seasonal dynamics across a wide range of ecosystems with consequences that include shifts in phenology, timing of nutrient availability, and changes in plant community composition. Current research has primarily focused on temperature as the key driver for these shifts because of the strong directional trend with climate warming, however, alterations in the availability of water across seasons is an unappreciated aspect of climate change that can significantly influence ecosystem functioning. While changes in the seasonal availability of water are expected to be globally pervasive, grasslands may be particularly vulnerable because these ecosystems are often water-limited and have species with distinct seasons of growth. Therefore, my dissertation examined how seasonal patterns of water availability may shift with climate change in the grasslands of the US Great Plains and the ecological consequences of these shifts. I first explored several mechanisms by which climate change is altering the seasonal water balance, using the Great Plains as a case study. Building on that, I then designed two field experiments in semi-arid grasslands that altered seasonal patterns of water availability to understand how these shifts affected ecosystem function and structure (primarily C3 vs C4 grasses). Overall, the results from both field experiments suggest that shifts in the seasonality of water availability with climate change will alter carbon cycling dynamics, shift seasonal patterns of canopy albedo, and differentially impact C3 vs. C4 species in the semi-arid grasslands of the US Great Plains. Thus, my research confirms the importance of this aspect of climate change and provides evidence that seasonal shifts in water availability can alter ecosystem processes and drive compositional changes. Since grasslands provide many economically and ecologically valuable services, understanding how climate change will impact these systems is critical for land managers and policymakers to make informed decisions.
Open Access
Gravity wave and microphysical effects on bow echo development
(Colorado State University. Libraries, 2012) Selin, Rebecca Denise Adams, author; Johnson, Richard, advisor; van den Heever, Susan, committee member; Bienkiewicz, Bogusz, committee member; Schumacher, Russ, committee member
Numerical simulations of the 13 March 2003 bow echo over Oklahoma are used to evaluate bow echo development and its relationship with gravity wave generation and microphysical heating profile variations. The first part of the research is directed at an explanation of recent observations of surface pressure surges ahead of convective lines prior to the bowing process. Multiple fast-moving n = 1 gravity waves are generated in association with fluctuations in the first vertical mode of heating in the convective line. The surface impacts of four such waves are observed in Oklahoma mesonet data during this case. A slower gravity wave is also produced in the simulation, which is responsible for the pre-bowing pressure surge in the model. This gravity wave is generated by an increase in low-level microphysical cooling associated with an increase in rear-to-front flow and low-level downdrafts shortly before bowing. The wave moves ahead of the convective line and is manifested at the surface by a positive pressure surge ahead of the convective line. The low-level upward vertical motion associated with this wave, in conjunction with higher-frequency gravity waves generated by the multicellularity of the convective line, increases the immediate pre-system CAPE by approximately 250 J kg-1. Two-dimensional heating profiles from this idealized, full-physics bow echo simulation are placed as a constant heat source in another simulation without moisture, to evaluate what type of gravity waves are produced by a heating profile from a given instance in time. A one-dimensional vertical mean heating profile is calculated from each two-dimensional profile, and a statistical method is used to evaluate the significance of each vertical mode. A number of gravity waves are produced in the dry simulation despite their vertical mode lacking statistical significance in the one-dimensional profile, suggesting that horizontal variations in the heating profile are important to consider. Microphysical sensitivity tests further elucidate the importance of the horizontal distribution of the microphysical heating profile. The tests used variations in the graupel parameter to evaluate its effect on bowing development and related forecasting parameters. Idealized and case study simulations showed that simulations using a larger, heavier, more "hail-like" graupel parameter with faster fallspeeds have decreased evaporation and melting rates concentrated closer behind the convective line, compared to simulations with a smaller, slower-falling, more "graupel-like" graupel parameter. This resulted in increased precipitation efficiency but a smaller stratiform region, weaker cold pool, weaker downdrafts and surface wind gusts, rear-to-front flow that remained elevated until close behind the convective line, and delayed bowing development in the "hail-like" simulations. Output from the case study sensitivity tests were compared to data from the Oklahoma Mesonet, which showed "hail-like" microphysical variations can cause significant variations in simulated forecasting parameters, including a 90 minute delay in onset of bowing, 150% weaker surface wind gusts, and a 600% increase in storm-total precipitation. Results from this work emphasize the importance of microphysical heating and cooling profiles in development of bow echoes, be it through the generation of multiple gravity waves and their feedback to the convection, or through direct modification of convective features such as the rear-inflow circulation and the cold pool strength. The pressure surge gravity wave generated by low-level cooling prior to bowing, and associated destabilization of the environment immediately in advance of the system, improves understanding of the cause of convective intensification as the system bows. However, the strong connection shown between bow echo development and microphysical processes, and the highly diverse nature of microphysical parameterizations, presents a challenge to the prediction of these severe weather phenomena.
Open Access
Joint tail modeling via regular variation with applications in climate and environmental studies
(Colorado State University. Libraries, 2013) Weller, Grant B., author; Cooley, Dan, advisor; Breidt, F. Jay, committee member; Estep, Donald, committee member; Schumacher, Russ, committee member
This dissertation presents applied, theoretical, and methodological advances in the statistical analysis of multivariate extreme values, employing the underlying mathematical framework of multivariate regular variation. Existing theory is applied in two studies in climatology; these investigations represent novel applications of the regular variation framework in this field. Motivated by applications in environmental studies, a theoretical development in the analysis of extremes is introduced, along with novel statistical methodology. This work first details a novel study which employs the regular variation modeling framework to study uncertainties in a regional climate model's simulation of extreme precipitation events along the west coast of the United States, with a particular focus on the Pineapple Express (PE), a special type of winter storm. We model the tail dependence in past daily precipitation amounts seen in observational data and output of the regional climate model, and we link atmospheric pressure fields to PE events. The fitted dependence model is utilized as a stochastic simulator of future extreme precipitation events, given output from a future-scenario run of the climate model. The simulator and link to pressure fields are used to quantify the uncertainty in a future simulation of extreme precipitation events from the regional climate model, given boundary conditions from a general circulation model. A related study investigates two case studies of extreme precipitation from six regional climate models in the North American Regional Climate Change Assessment Program (NARCCAP). We find that simulated winter season daily precipitation along the Pacific coast exhibit tail dependence to extreme events in the observational record. When considering summer season daily precipitation over a central region of the United States, however, we find almost no correspondence between extremes simulated by NARCCAP and those seen in observations. Furthermore, we discover less consistency among the NARCCAP models in the tail behavior of summer precipitation over this region than that seen in winter precipitation over the west coast region. The analyses in this work indicate that the NARCCAP models are effective at downscaling winter precipitation extremes in the west coast region, but questions remain about their ability to simulate summer-season precipitation extremes in the central region. A deficiency of existing modeling techniques based on the multivariate regular variation framework is the inability to account for hidden regular variation, a feature of many theoretical examples and real data sets. One particular example of this deficiency is the inability to distinguish asymptotic independence from independence in the usual sense. This work develops a novel probabilistic characterization of random vectors possessing hidden regular variation as the sum of independent components. The characterization is shown to be asymptotically valid via a multivariate tail equivalence result, and an example is demonstrated via simulation. The sum characterization is employed to perform inference for the joint tail of random vectors possessing hidden regular variation. This dissertation develops a likelihood-based estimation procedure, employing a novel version of the Monte Carlo expectation-maximization algorithm which has been modified for tail estimation. The methodology is demonstrated on simulated data and applied to a bivariate series of air pollution data from Leeds, UK. We demonstrate the improvement in tail risk estimates offered by the sum representation over approaches which ignore hidden regular variation in the data.
Open Access
Micrometeorological studies of a beef feedlot, dairy, and grassland: measurements of ammonia, methane, and energy balance closure
(Colorado State University. Libraries, 2018) Shonkwiler, Kira Brianne, author; Collett, Jeffrey L., advisor; Ham, Jay M., committee member; Kreidenweis, Sonia, committee member; Schumacher, Russ, committee member; Archibeque, Shawn, committee member
Ammonia emissions from concentrated animal feeding operations (CAFOs; most of which are beef feedlots) near the Colorado Front Range are suspected to be a large regional input of reactive nitrogen which has been found to accumulate and cause deleterious effects in nearby downwind Class I areas like Rocky Mountain National Park. Methane (CH4) is a strong greenhouse gas (GHG) emitted in large amounts from dairy anaerobic lagoons used for liquid manure management. Lagoon systems account for over half of the manure management-based CH4 emissions from agriculture in the US. There is a strong need for more emissions measurements from CAFOs like feedlots and dairies. For these data to be trusted, well-developed techniques must be utilized at emissions measurement sites and such techniques should be validated in ideal scenarios. Three micrometeorological studies were performed involving measurement of emissions using micrometeorological methods in the surface layer. The first study involved estimating summertime NH3 emissions from a 25,000-head beef feedlot in Northern Colorado. Two different NH3 sensors were used: a cavity ring down spectroscopy analyzer collected data at a single point while a long-path FTIR collected data along a 226-m long transect, both deployed along the same fenceline. Concentration data from these systems were used with two inverse dispersion models (FIDES, an inverse solution to the advection dispersion equation; and WindTrax, a backward Lagrangian stochastic model). Point sensor concentrations of NH3 were similar to line-integrated sensor concentrations suggesting some spatial uniformity in emissions. Emissions had a diurnal pattern (i.e., afternoon peak with minimum in early morning) that was driven by temperature. Emissions predicted by WindTrax were 25.2% higher than those from FIDES. Point vs. long-path measurements of NH3 had minimal effect on predicted emissions. The mean NH3 emission factor (EF) was 80 ± 39 g NH3 hd−1 d−1, with 40.0% of dietary-N emitted as NH3. The second study involved using eddy covariance and WindTrax to quantify CH4 emissions from a 3.9-ha anaerobic lagoon serving a 1400-head dairy in northern Colorado. Methane emissions followed a strong seasonal pattern correlated with temperature of the organic sludge layer on the bottom of the lagoon. Fluxes started increasing in late spring (May; ~10°C), increased rapidly in Jun (10-15°C) peaked in the summer (Jul/Aug; ~18-20°C) and remained high until mid-autumn (late Oct/early Nov; ~10°C). Fluxes then decreased and remained consistently low (up to 10 times less than peak emissions) until microbial activity ramped up again in May. The EC signal was very dependent on wind direction, with highest concentrations and fluxes associated with the direction of the lagoon. Gap-filled data showed a slight diurnal pattern to all seasons, with tenfold increases in diurnal values for summer over winter. Additionally, EFs for the lagoon varied by season with lows in the winter and highs in the summer with an annual mean of 819 ± 774 g CH4 hd-1 d-1. WindTrax overestimated EC for the lagoon (1163 ± 1049 g CH4 hd-1 d-1 versus 819 ± 774 g CH4 hd-1 d-1), but this difference may be attributable to differences in the sampling footprint and stability conditions. IPCC Tier 2-calculated EFs were extremely close to EC-based measurements and WT-based estimates. The third study involved using eddy covariance in an ideal environment (tallgrass prairie in Kansas) to test the reasons behind the "energy balance (EB) closure problem" at two landscape positions. This problem can cast uncertainty on flux measurements made by EC. One upland and one lowland EC tower each were used to measure EB components (i.e., net radiation, Rn; soil heat flux, G; total change in heat storage, deltaS; and sensible and latent heat fluxes, H and λE) during the summers of 2007 and 2008. To maximize closure, special attention was given to reduce all forms of instrumentation error and account for heat storage and photosynthesis between the soil and the reference height. Landscape position had little effect on G, H, and Rn; differences were ≤ 2% between sites. Lowland λE was 8% higher than upland λE because of greater biomass and soil moisture. On average, EB closure (i.e., Σ[λE+H] / Σ[Rn–G–ΔS]) was 0.88 and 0.94 at the upland and lowland sites, respectively. Closure was not correlated with friction velocity or the stability of the surface boundary layer. Given high confidence in Rn, G, and ΔS, turbulent fluxes depend directly on vertical velocity (w), and the fact that a systematic underestimation of w was recently found in literature, lack of closure may have resulted largely from anemometer-based underestimates of w.
Open Access
Nonstationary flood risk assessment in coastal regions under climate change
(Colorado State University. Libraries, 2021) Ghanbari, Mahshid, author; Arabi, Mazdak, advisor; Ettema, Robert, committee member; Schumacher, Russ, committee member; Bhaskar, Aditi, committee member
Coastal cities are exposed to multiple flood drivers including high tide, storm surge, extreme rainfall, and high river flows. The occurrence of these flood drivers, either in isolation or in combination, can cause significant risk to property and human life. Climate change is placing greater pressure on coastal communities by increasing frequency and intensity of flood events through sea level rise (SLR) and more extreme rainfall and storm events. Therefore, effective adaptation strategies are essential to reduce future flood risk in exposed communities. The planning and implementation of effective adaptation strategies require a comprehensive understanding of future flood hazards and risks under future climate conditions and adaptation options. The overarching goal of this dissertation is to improve the capacity to understand, estimate and mitigate future flood hazards and risks in coastal areas under uncertain climate change. To achieve this goal, first, a nonstationary mixture probability model was developed that enables simultaneous characterization of minor and major flood events under future sea level conditions. The probability model was used to estimate minor and major flooding frequency at 68 locations along the coasts of the Contiguous United States (CONUS). The results showed a significant increase in frequency of both minor and major flood events under future sea level conditions. However, the frequency amplification of minor and major flooding varied by coastal regions. While regions in the Pacific and southeast Atlantic coast are likely to be exposed to higher frequency amplification in major flooding, the Gulf and northeast Atlantic coastal regions should expect the highest minor flood frequency amplification. Second, the proposed mixture probability model was employed in a flood risk assessment framework to enable assessing future acute and chronic coastal flood risks under different SLR and adaptation levels. The HAZUS-MH flood loss estimation tool was used to estimate property damage. The application of the framework in Miami-Dade County revealed that as sea level rises, chronic risks from repetitive nonextreme flooding may exceed acute risks from extreme floods. Third, a nonstationary bivariate flood hazard assessment method was developed that enables estimation of future frequency of compound coastal-riverine flooding with consideration of impacts of climate change including SLR and variations in extreme river flows. The proposed method was employed at 26 paired tidal-riverine stations along the CONUS coast. Specifically, the joint return period of compound major coastal-riverine flooding, defined based on flood impact thresholds, was explored by mid-century. The results showed that under current climate conditions the northeast Atlantic and western part of the Gulf coasts are exposed to the highest compound major coastal-riverine flood probability. However, considering future SLR, emerging high compound major flooding probability was evident in the southeast Atlantic coast. The impact of changes in extreme river flows was found to be negligible in most of the locations. Finally, four stormwater intervention scenarios including gray (i.e., conventional centralized conveyance systems and water treatment plants) and green (i.e., decentralized infiltration measures) infrastructure systems, were assessed in New York City (NYC). The results revealed that in developed and urbanized cities like NYC, green systems should not be considered as a substitute for gray systems. Complementary benefits on flood and combined sewer outflow (CSO) reduction can be gained through integration of green and gray systems.
Open Access
Numerical simulation diagnostics of a flash flood event in Jeddah, Saudi Arabia
(Colorado State University. Libraries, 2014) Samman, Ahmad, author; Cotton, William R., advisor; Schumacher, Russ, committee member; Fontane, Darrell G., committee member
On 26 January 2011, a severe storm hit the city of Jeddah, the second largest city in the Kingdom of Saudi Arabia. The storm resulted in heavy rainfall, which produced a flash flood in a short period of time. This event caused at least eleven fatalities and more than 114 injuries. Unfortunately, the observed rainfall data are limited to the weather station at King Abdul Aziz International airport, which is north of the city, while the most extreme precipitation occurred over the southern part of the city. This observation was useful to compare simulation result even though it does not reflect the severity of the event. The Regional Atmospheric Modeling System (RAMS) developed at Colorado State University was used to study this storm event. RAMS simulations indicted that a quasi-stationary Mesoscale convective system developed over the city of Jeddah and lasted for several hours. It was the source of the huge amount of rainfall. The model computed a total rainfall of more than 110 mm in the southern part of the city, where the flash flood occurred. This precipitation estimation was confirmed by the actual observation of the weather radar. While the annual rainfall in Jeddah during the winter varies from 50 to 100 mm, the amount of the rainfall resulting from this storm event exceeded the climatological total annual rainfall. The simulation of this event showed that warm sea surface temperature, combined with high humidity in the lower atmosphere and a large amount of convective available potential energy (CAPE) provided a favorable environment for convection. It also showed the presence of a cyclonic system over the north and eastern parts of the Mediterranean Sea, and a subtropical anti-cyclone over Northeastern Africa that contributed to cold air advection bringing cold air to the Jeddah area. In addition, an anti-cyclone (blocking) centered over east and southeastern parts of the Arabian Peninsula and the Arabian Sea produced a low level jet over the southern part of the Red Sea, which transported large water vapor amounts over Jeddah. The simulation results showed that the main driver behind the storm was the interaction between these systems over the city of Jeddah (an urban heat island) that produced strong low-level convergence. Several sensitivity experiments were carried out showed that other variables could have contributed to storm severity as well. Those sensitivity experiments included several simulations in which the following variables were changed: physiographic properties were altered by removing the water surfaces, removing the urban heat island environment from the model, and changing the concentration of cloud condensation nuclei. The results of these sensitivity experiments showed that these properties have significant effects on the storm formation and severity.
Open Access
Radar and lightning analyses of gigantic jet-producing storms
(Colorado State University. Libraries, 2012) Meyer, Tiffany C., author; Rutledge, Steven A., advisor; Lang, Timothy, committee member; Robinson, Raymond, committee member; Schumacher, Russ, committee member
An analysis of the storm structure and evolution associated with six gigantic jets was conducted. Three of these gigantic jets were observed within detection range of very high-frequency lightning mapping networks. All six were within range of operational radars and two-dimensional lightning network coverage: five within the National Lightning Detection Network and one within the Global Lighting Detection network. Most of the storms producing the jets formed in a high CAPE, low lifted index environments and had maximum reflectivity values of 54 to 62 dBZ and 10-dBZ echo tops reaching 14-17 km. Most storms were near the highest lighting flash rate and peak storm intensity with an overshooting echo top just before or after the time of the jet. The overshooting top and strong intensification may have indicated a convective surge which allowed the upper positive charge to mix with a negatively charged screening layer that became depleted. Intra-cloud lightning initiating in the mid-level negative region could have exited upward through the recently depleted positive region, producing a gigantic jet.
Open Access
Storm microphysics and kinematics at the ARM-SGP site using dual polarized radar observations at multiple frequencies
(Colorado State University. Libraries, 2014) Matthews, Alyssa A., author; Rutledge, Steven, advisor; Chandrasekar, Venkatachalam, committee member; Schumacher, Russ, committee member; Cabell, Brenda, committee member
This research utilizes observations from the Atmospheric Radiation Measurement (ARM) Climate Research Facility at the Southern Great Plains location to investigate the kinematic and microphysical processes present in various types of weather systems. The majority of the data used was collected during the Mid-latitude Continental Convective Cloud Experiment (MC3E), and utilizes the network of scanning radars to arrive at a multi-Doppler wind retrieval and is compared to vertical wind measurements from a centrally located profiling radar. Microphysical compositions of the storms are analyzed using a multi-wavelength hydrometeor identification algorithm utilizing the strengths of each of the radar wavelengths available (X, C, S). When available, a comparison is done between observational analysis and simulated model output from the Weather Research Forecasting model with Spectral-bin Microphysics (WRF-SBM) using bulk statistics to look at reflectivity, vertical motions, and microphysics.