Theses and Dissertations from UMD
Permanent URI for this communityhttp://hdl.handle.net/1903/2
New submissions to the thesis/dissertation collections are added automatically as they are received from the Graduate School. Currently, the Graduate School deposits all theses and dissertations from a given semester after the official graduation date. This means that there may be up to a 4 month delay in the appearance of a give thesis/dissertation in DRUM
More information is available at Theses and Dissertations at University of Maryland Libraries.
Browse
113 results
Search Results
Item INVESTIGATION OF AMBIENT METHANE CONCENTRATION, SOURCES, AND TRENDS IN THE BALTIMORE-WASHINGTON REGION(2024) Sahu, Sayantan; Dickerson, Russell Professor; Chemistry; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Methane, an important and not yet fully understood greenhouse gas, has a global warming potential 25 times that of carbon dioxide over 100 years, although with an atmospheric lifetime much shorter than carbon dioxide. Controlling methane emissions is a useful way to avoid some of the adverse effects of climate change at least on short time scales. Natural sources include wetlands, ruminants, and wildfires, while anthropogenic sources include the production, transmission, distribution, and use of natural gas, livestock, and landfills. In the US, natural gas and petroleum systems, anthropogenic sources, are the second-largest source of methane emissions. Urban areas are a significant source of anthropogenic methane emissions, primarily fugitive emissions from natural gas distribution and usage.We studied methane observations from five towers in the Baltimore-Washington (BWR) region – two urban towers ARL (Arlington, VA), NEB (Northeast Baltimore, MD), and one rural tower, BUC (Bucktown, MD). Methane measurements from these three towers displayed distinct seasonal and diurnal cycles with maxima at night and in the early morning, which indicated significant local emissions. We concluded from our analysis that anthropogenic methane emissions dominate at the urban sites whereas wetland emissions dominate at the rural site. We compared observed enhancements (mole fractions above the 5th percentile) to simulated methane enhancements using the WRF-STILT model driven by two EDGAR inventories – EDGAR 4.2 and EDGAR 5.0. We did a similar comparison between model and observations with vertical gradients. We concluded that both versions of EDGAR underestimated the regional anthropogenic emissions of methane, but version 5.0 had a more accurate spatial representation. We ran the model with WETCHARTs to account for wetland emissions which significantly reduced the bias between model and observations especially in summer at the rural site. We investigated winter methane observations from three towers in the BWR including a ten-year record, 2013-2022, from BUC, located ~100 km southeast of these urban areas. We combined the observations with a HYSPLIT clustering analysis for all years to determine the major synoptic patterns influencing methane mixing ratios at BUC. For methane concentrations above global background, the cluster analysis revealed four characteristic pathways of transport into BUC – from the west (W), southwest (SW), northwest (NW), and east (E) and these showed significant differences in methane mixing ratios. We corroborated our conclusions from BUC using 2018-2022 data from towers in Stafford, Virginia (SFD), and Thurmont, Maryland (TMD); results confirmed the influence of synoptic pattern, typically associated with frontal passage, on methane. No significant temporal trend over the global background was detected overall or within any cluster. For BUC, low concentrations were observed for air off the North Atlantic Ocean (E cluster) and flowing rapidly behind cold fronts (NW cluster). High methane mixing ratios were observed, as expected, in the W cluster due to the proximity of the BWR and oil and gas operations in the Marcellus. Less expected were high mixing ratios for the SW cluster – we attribute these to agricultural sources in North Carolina. Swine production, ~500 km to the SW, impacts methane in eastern Maryland as much or more than local urban emissions plus oil and gas operations 100–300 km to the west; this supports the high end of emission estimates for animal husbandry and suggests strategies for future research and mitigation.Item DEEP LEARNING APPROACHES FOR ESTIMATING AND FORECASTING SURFACE DOWNWARD SHORTWAVE RADIATION FROM SATELLITE DATA(2024) Li, Ruohan; Wang, Dongdong; Geography; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Surface downward shortwave radiation (DSR) designates solar radiation with a wavelength from 300 to 4000 nm received at the Earth’s surface. DSR plays a pivotal role in the surface energy and radiation budget, serving as the primary driver for hydrological, ecological, and biogeochemical cycles (Liang et al., 2010, 2019), and the important input for various earth models (Huang et al., 2019; Liang et al., 2010; Stephens et al., 2012). Given the rising demand for renewable energy, as well as accelerated advancements in solar energy technologies on both utility-scale and residential scale, the precision and resolution in estimating and forecasting DSR have become indispensable for planning and administering solar power plants (Gueymard, 2014; Jiang et al., 2019). This dissertation delves into the potential of integrating deep learning with satellite observations to address the deficiencies in current DSR estimation and forecasting methods, aiming to cater to the evolving needs of solar radiation estimation. The research begins by examining current DSR satellite products, emphasizing their limitations, particularly concerning spatial resolution and performance in snowy, cloudy, and high-latitude areas. In such regions, challenges arise from the degradation of radiative transfer models, band saturation, the pronounced effects of 3D cloud dynamics, and temporal resolution constraints (Li et al., 2021). Identifying these gaps, the study introduces the concept of transfer learning to tackle cases where physical methods degrade and limited training data is available. By combining data from physical simulations and ground observations, the proposed models enhance both the accuracy and adaptability of DSR predictions on a global scale. The investigation further reveals the influence of training data volume on model performance, illustrating how transfer learning can ameliorate these effects (Li et al., 2022). Moreover, the dissertation compares the application of DenseNet, Gated Recurrent Unit (GRU), and a hybrid of Convolutional Neural Network (CNN) and GRU (CNNGRU) to geostationary satellite data, achieving precise and timely DSR estimates. These models underscore their prowess in tackling 3D cloud effects and reducing dependency on additional data sources by the spatial and temporal structure of DL (Li et al., 2023b). Finally, the dissertation introduces the SolarFormer, a space-time transformer neural network adept at forecasting solar radiation up to three hours in advance at 15-minute intervals. By harnessing solely geostationary satellite imagery without the need for ground measurements, this model facilitates expansive DSR predictions, which are crucial for optimizing solar energy distribution at both utility and micro scales. This chapter also highlights the Transformer model's potential for extended forecasting due to its computational and memory efficiency.Item DEVELOPMENT AND APPLICATION OF PROPINQUITY MODELING FRAMEWORK FOR IDENTIFICATION AND ANALYSIS OF EXTREME EVENT PATTERNS(2024) kholodovsky, vitaly; Liang, Xin-Zhong; Atmospheric and Oceanic Sciences; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Extreme weather and climate events such as floods, droughts, and heat waves can cause extensive societal damage. While various statistical and climate models have been developed for the purpose of simulating extremes, a consistent definition of extreme events is still lacking. Furthermore, to better assess the performance of the climate models, a variety of spatial forecast verification measures have been developed. However, in most cases, the spatial verification measures that are widely used to compare mean states do not have sufficient theoretical justification to benchmark extreme events. In order to alleviate inconsistencies when defining extreme events within different scientific communities, we propose a new generalized Spatio-Temporal Threshold Clustering method for the identification of extreme event episodes, which uses machine learning techniques to couple existing pattern recognition indices with high or low threshold choices. The method consists of five main steps: construction of essential field quantities, dimension reduction, spatial domain mapping, time series clustering, and threshold selection. We develop and apply this method using a gridded daily precipitation dataset derived from rain gauge stations over the contiguous United States. We observe changes in the distribution of conditional frequency of extreme precipitation from large-scale, well-connected spatial patterns to smaller-scale, more isolated rainfall clusters, possibly leading to more localized droughts and heatwaves, especially during the summer months. Additionally, we compare empirical and statistical probabilities and intensities obtained through the Conventional Location Specific methods, which are deficient in geometric interconnectivity between individual spatial pixels and independent in time, with a new Propinquity modeling framework. We integrate the Spatio-Temporal Threshold Clustering algorithm and the conditional semi-parametric Heffernan and Tawn (2004) model into the Propinquity modeling framework to separate classes of models that can calculate process level dependence of large-scale extreme processes, primarily through the overall extreme spatial field. Our findings reveal significant differences between Propinquity and Conventional Location Specific methods, in both empirical and statistical approaches in shape and trend direction. We also find that the process of aggregating model results without considering interconnectivity between individual grid cells for trend construction can lead to significant variations in the overall trend pattern and direction compared with models that do account for interconnectivity. Based on these results, we recommend avoiding such practices and instead adopting the Propinquity modeling framework or other spatial EVA models that take into account the interconnectivity between individual grid cells. Our aim for the final application is to establish a connection between extreme essential field quantity intensity fields and large-scale circulation patterns. However, the Conventional Location Specific Threshold methods are not appropriate for this purpose as they are memoryless in time and not able to identify individual extreme episodes. To overcome this, we developed the Feature Finding Decomposition algorithm and used it in combination with the Propinquity modeling framework. The algorithm consists of the following three steps: feature finding, image decomposition, and large-scale circulation patterns connection. Our findings suggest that the Western Pacific Index, particularly its 5th percentile and 5th mode of decomposition, is the most significant teleconnection pattern that explains the variation in the trend pattern of the largest feature intensity.Item Combining Physics-based Modeling, Machine Learning, and Data Assimilation for Forecasting Large, Complex, Spatiotemporally Chaotic Systems(2023) Wikner, Alexander Paul; Ott, Edward; Physics; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)We consider the challenging problem of forecasting high-dimensional, spatiotemporally chaotic systems. We are primarily interested in the problem of forecasting the dynamics of the earth's atmosphere and oceans, where one seeks forecasts that (a) accurately reproduce the true system trajectory in the short-term, as desired in weather forecasting, and that (b) correctly capture the long-term ergodic properties of the true system, as desired in climate modeling. We aim to leverage two types of information in making our forecasts: incomplete scientific knowledge in the form of an imperfect forecast model, and past observations of the true system state that may be sparse and/or noisy. In this thesis, we ask if machine learning (ML) and data assimilation (DA) can be used to combine observational information with a physical knowledge-based forecast model to produce accurate short-term forecasts and consistent long-term climate dynamics. We first describe and demonstrate a technique called Combined Hybrid-Parallel Prediction (CHyPP) that combines a global knowledge-based model with a parallel ML architecture consisting of many reservoir computers and trained using complete observations of the system's past evolution. Using the Kuramoto-Sivashinsky equation as our test model, we demonstrate that this technique produces more accurate short-term forecasts than either the knowledge-based or the ML component model acting alone and is scalable to large spatial domains. We further demonstrate using the multi-scale Lorenz Model 3 that CHyPP can incorporate the effect of unresolved short-scale dynamics (subgrid-scale closure). We next demonstrate how DA, in the form of the Ensemble Transform Kalman Filter (ETKF), can be used to extend the Hybrid ML approach to the case where our system observations are sparse and noisy. Using a novel iterative scheme, we show that DA can be used to obtain training data for successive generations of hybrid ML models, improving the forecast accuracy and the estimate of the full system state over that obtained using the imperfect knowledge-based model. Finally, we explore the commonly used technique of adding observational noise to the ML model input during training to improve long-term stability and climate replication. We develop a novel training technique, Linearized Multi-Noise Training (LMNT), that approximates the effect of this noise addition. We demonstrate that reservoir computers trained with noise or LMNT regularization are stable and replicate the true system climate, while LMNT allows for greater ease of regularization parameter tuning when using reservoir computers.Item IMPROVING SWAT MODELING AND INTEGRATING OBSERVATIONAL ANALYSIS TO ASSESS THE IMPACTS OF CLIMATE AND LAND USE CHANGE ON HYDROLOGICAL PROCESSES IN THE UNITED STATES(2023) dangol, sijal; Liang, Xin-Zhong; Atmospheric and Oceanic Sciences; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Climate and land use have been changing significantly. Climate and land use changes are intricately linked to the hydrological processes that have a significant impact on water resources. The availability of water resources is crucial for food and energy production. In the U.S., about 42% of water is used for agricultural irrigation to improve the productivity of grain crops, including corn and soybeans. About 30% of the corn produced in the U.S. is used for ethanol production, which raises serious concerns regarding the food-fuel competition and detrimental environmental impacts. The expansion of corn production competes for water resources, as corn requires irrigation. To address these issues, the cultivation of bioenergy crops on marginal lands is proposed as a solution that also alleviates water demand, as they do not require irrigation. This cultivation could involve complex interactions among climate, crops, and hydrological processes, which may lead to significant environmental consequences. This necessitates an understanding of how land use and climate changes will affect hydrological processes and, subsequently, food production.To understand these processes, I use the hydrological model, Soil and Water Assessment Tool (SWAT). SWAT has a distinct advantage in its capability to simulate water and sediment transport, crop growth, agricultural management, and land management practices, and has specialized extensions for groundwater simulation (SWAT-MODFLOW). In this study, I focus on large agricultural watersheds in the U.S. Northern High Plains (NHP) Aquifer region. The NHP is a vast and critical agricultural area that relies heavily on groundwater irrigation, which has resulted in streamflow and groundwater level decline. Although the SWAT model can quantify the environmental impacts of climate and land use changes, its utility is limited by: (1) deficiencies in the auto-irrigation algorithms in SWAT that continue irrigation during the non-growing season, (2) optimization of the SWAT model with streamflow data only, without considering other hydrological components such as evapotranspiration and soil moisture, (3) failure to explicitly account for shoot and root biomass development when simulating bioenergy crops, and (4) standalone climate and land use change simulation that do not consider the influence of climate-crop feedback on hydrological processes. To provide an accurate assessment of the impact of climate and land use changes on hydrological processes in the NHP, my first effort involved improving the irrigation algorithms in both SWAT and its extension SWAT-MODFLOW to provide an improved representation of irrigational water use conditions. The use of SWAT-MODFLOW provides an improved representation of hydrological processes, particularly when considering the significant role of groundwater dynamics in the overall water budget of the NHP. The modified SWAT and SWAT-MODFLOW were applied to the NHP which exhibited improved performance in simulating groundwater irrigation volume, groundwater level, and streamflow in the NHP. I also examined the effects of groundwater irrigation on the water cycle. Based on simulation results from SWAT-MODLFOW, historical irrigation has increased surface runoff, evapotranspiration, soil moisture, and groundwater recharge by 21.3%, 4.0%, 2.5%, and 1.5%, respectively. Irrigation also improved crop water productivity by nearly 27.2% for corn and 23.8% for soybean. To optimize the SWAT model for addressing climate and land use change issues in the NHP, I tried to identify the best approach to calibrate the SWAT. Traditionally, SWAT is calibrated using streamflow only. I hypothesize that calibrating SWAT with streamflow only, without considering broad hydrological processes like evapotranspiration and soil moisture causes model overfitting. I utilized the best available remotely sensed data, including Atmosphere–Land Exchange Inverse (ALEXI) Evapotranspiration (ET), Moderate Resolution Imaging Spectroradiometer (MODIS) ET, and Soil MERGE (SMERGE), to examine if multi-variate calibration could improve model performance. Contrary to my hypothesis, using remotely sensed data does not improve model performance because of two reasons: (a) remotely sensed data may lack physical constraints and has large uncertainty, and (b) SWAT may not provide an accurate representation of hydrological processes due to deficiencies in its model algorithms. Further investigation is required to address these issues. SWAT utilizes standalone climate change data without considering climate-crop feedback in environmental impact assessments of climate and land use change scenarios. To address this gap, I used another tool, the Climate-Weather and Research Forecasting (CWRF) model, both in standalone mode and coupled with the BioCro model. This approach incorporates the climate-crop feedback associated with the cultivation of perennial grasses on marginal lands. The climate data from these two scenarios: (i) Climate-only without climate-crop feedback, and (ii) Climate-crop feedback turned on, were used in SWAT to investigate hydrological responses to simulated climate-crop feedback in the NHP. The incorporation of climate-crop feedback produced cooler and wetter conditions in the NHP. These changes in temperature and precipitation have substantial environmental consequences. The cultivation of miscanthus on marginal lands increases evapotranspiration and decreases surface runoff, soil moisture, and percolation. At the watershed scale, surface runoff substantially increases during the growing season, both in the present (1985-2014) and the future (2031-2060). The differences in the extent of marginal land use for miscanthus cultivation between the Platte (4%) and Republican (20%) River basins result in different responses in streamflow and nitrogen loading. For example, in the future, annual streamflow and nitrogen loading increases by 5.5% and 2.3%, respectively, in the Platte River basin. In contrast, the Republican River basin exhibits negligible changes in annual streamflow and an 8.6% decrease in nitrogen loading during this period. These results highlight the importance of including climate-crop feedback in addition to the impact from land use change when assessing the impacts of climate and land use changes on hydrological processes. The SWAT model does not explicitly account for shoot and root biomass development. I integrated the grass growth sub-model from the DAYCENT into SWAT (SWAT–GRASSD) and further modified it by considering the impact of leaf area index (LAI) on potential biomass production (SWAT–GRASSM). Based on testing at eight sites in the U.S. Midwest, SWAT–GRASSM generally outperformed both SWAT and SWAT–GRASSD in simulating switchgrass biomass yield and the seasonal development of LAI. Additionally, SWAT–GRASSM can more realistically represent root development, which is key for the allocation of accumulated biomass and nutrients between aboveground and belowground biomass pools. From these model experiments and analyses, I realized that there are still a lot of model and data gaps that need to be addressed. Consequently, the results of the impacts of the climate and land use changes on hydrological processes could be subjective due to inherent model deficiencies and calibration issues because of uncertainty in observed data. I, therefore, revisited the observed data screening approaches to identify near-natural streamflow data that are representative of actual watershed processes. Assuming that the hydrologic response of the watershed is primarily driven by precipitation and runoff processes, I applied lagged correlation analysis between precipitation and streamflow at various temporal scales, complemented by a systematic screening process using basin properties to select streamgages with near-natural streamflow. The screening method introduced here is objective and easy to implement in other data-sparse regions. The selected streamgages serve as valuable calibration and validation points for hydrological models, thereby it could help enhance the accuracy and confidence in model projections. In summary, my work improved the SWAT auto-irrigation algorithm for a better representation of irrigation water use and integrated the DAYCENT grass growth sub-model to enhance biomass yield predictions. The hydrological consequences of land use change for miscanthus cultivation highlighted the importance of incorporating climate-crop feedback into hydrological modeling. This integration had a substantial impact on the hydrological processes in the NHP. Despite the model improvements through changes in model algorithms and optimization, my work reveals that model and data gaps result in subjectivity in the study’s findings. To address this, I propose an objective screening process for streamgage selection that reflects natural watershed processes. The enhancements introduced to the modeling framework, coupled with the objective streamgage screening approach, will assist water resource managers and the regional hydrologic modeling community in the credible assessment of the hydrological impacts resulting from climate and land use changes.Item Bridging Gaussian and non-Gaussian Data Assimilation for High-Dimensional Geophysical Models(2023) Kurosawa, Kenta; Poterjoy, Jonathan; Atmospheric and Oceanic Sciences; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Strongly nonlinear model dynamics and observation operators can induce bias in Gaussian-based data assimilation methods commonly used for numerical weather prediction, such as ensemble Kalman filters (EnKFs) and the 4D variational method (4DVar). This limitation is apparent for multiscale weather prediction systems that exhibit large uncertainty in smaller scales, or when observations are sensitive to cloud processes. Several methods have been proposed for improving data assimilation performance in nonlinear regimes. Examples include the adoption of an "outer loop" in variational methods, which helps reduce bias caused by linear assumptions. Likewise, numerous "iterative ensemble methods" exist, which periodically re-linearize model and measurement operators in the same manner. While the convergence properties of the latter methods are not completely known, numerical experiments performed by several previous studies suggest they can provide accurate solutions for mildly nonlinear problems. Another strategy that has gained momentum in recent years is to apply dimension-reduction procedures (namely, localization) to particle filters (PFs). PFs avoid the parametric estimation of Bayesian posterior densities, thus providing great flexibility for solving non-Gaussian data assimilation problems. However, these methods are more easily affected by sampling error than Gaussian-based methods—even when using localization. My research introduces new approaches that bridge Gaussian and non-Gaussian data assimilation for geophysical models. To begin, the first part of this study investigates intrinsic limitations in data assimilation methods that are currently used for nonlinear applications in geoscience. We then propose novel data assimilation strategies for combining PFs with Gaussian-based methods that are more robust to sampling error. We demonstrate that the approaches have significant value within modern high-resolution regional atmospheric modeling systems, which are designed specifically for predicting tropical cyclones and severe convective storms. We further emphasize that this research has general implications for data assimilation within Earth-system models.Item ATMOSPHERIC ORGANIC AEROSOLS: THE EFFECT OF PHYSIOCHEMICAL PROPERTIES ON HYGROSCOPICITY(2023) Malek, Kotiba; Asa-Awuku, Akua; Chemical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Aerosols, tiny solid or liquid particles, are ubiquitous in the atmosphere yet their impact on climate remains poorly understood. One prominent way aerosols are able to impact the climate is through their ability to uptake water and form clouds. The chemical diversity and aerosol interactions in the atmosphere can greatly complicate the investigation of aerosol-cloud interactions. This complexity is expressed with a large uncertainty associated with aerosols’ role on climate change. This dissertation investigates the aerosol-cloud interaction by measuring the water uptake of atmospherically relevant aerosols. Our results highlight the importance of accounting for various physiochemical properties when exploring the water uptake of atmospheric aerosols. One such property is liquid-liquid phase separation (LLPS) in ternary mixtures. Our work offers new evidence, insight, and a paradigm shift to the contribution of LLPS to supersaturated droplet activation. We complemented this finding with a theoretical model, that incorporates solubility, O:C ratio, and LLPS, for predicting κ-hygroscopicity of ternary mixtures. Another physiochemical property that was shown to play a key role in droplet activation of polymeric aerosols is chemical structure. Our study shows that polycatechol is more hygroscopic than polyguaiacol and the difference in hygroscopicity is attributed to the density of hydroxyl groups in both structures. Polycatechol has a higher density of hydroxyl groups than polyguaiacol, resulting in polycatechol having stronger water uptake affinity than polyguaiacol. When maintaining the same structural makeup by investigating the water uptake of two isomeric compounds, we discovered that solubility was the driving force in water uptake. The more soluble isomer o-aminophenol was more hygroscopic than p-aminophenol. Hence, a small change in the position of functional groups can impact solubility which in turn influence hygroscopicity. Lastly, we explored the presence of gas-phase organics on the water uptake of isomers with a wide range of solubilities. Our work highlights that gas-phase organics, specifically ethanol, can influence the water uptake of aerosols. Ethanol was shown to increase water uptake efficiencies based on solubility, with the least soluble compound showing stronger affinity to water uptake. Overall, this thesis advances our knowledge and understanding of aerosol-cloud interactions and its implications on climate change.Item Photochemistry of Exoplanet Atmospheres: Modelling alien chemistry accurately and self-consistently(2023) Teal, Dillon James; Kempton, Eliza; Astronomy; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Exoplanets offer unique physical and chemical laboratories experiencing entirely alien environments compared to the Solar System planets. Their atmospheres, governed by the same laws of physics, display remarkable diversity and complexity. They serve as the most complex planetary phenomena we can directly observe, coupled to the planet's interior processes, formation environment, the properties of the host star, and complex chemical ecosystems. The art of modelling these systems is a rich field of study, and in this work I study the nature of photochemical models and what understanding they can provide for us based on the quality and breadth of their inputs. By characterizing the implicit uncertainty chemical models have without a well-characterized host star, I quantify the importance of host star characterization to chemical modelling, showing their sensitivity under different reaction schemes and microphysical models. I then apply this to recent observations of known exoplanet host stars LHS 3844 and AU Microscopii. Finally, I cover work to model sub-Neptune atmospheres across a wide parameter space aimed at understanding the influence of a planet's environment and unknowns on haze formation and observational prevalence in emission and transmission spectroscopy.Item Exploring the Observation Impacts and Enhancing the Predictability for Ensemble-Based Coupled Data Assimilation(2023) Chang, Chu-Chun; Kalnay, Eugenia EK; Atmospheric and Oceanic Sciences; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)This research aims to explore the observation impacts in coupled data assimilation (CDA) and improve the predictability of coupled systems by advanced DA approaches. Three topics are discussed in this dissertation: (1) An enhanced application of the correlation cutoff method (Yoshida and Kalnay, 2018) as a spatial localization is introduced. We investigated the feasibility and characteristics of the traditional distance-dependent (Gaspari and Cohn, 1999) and the correlation-dependent localizations preliminary on the Lorenz (1996) model with the local ensemble transform Kalman filter (LETKF). We further discussed the potential of integrative localization strategies and the application of the correlation cutoff method on Mars DA. (2) We found that the surface sea temperature (SST) relaxation operationally used in the Climate Forecast System version 2 (CFSv2) is not effective in reducing existing SST biases. To address this issue, we replaced the SST relaxation with the weakly coupled data assimilation (WCDA) of satellite-retrieved SST products. A series of experiments with real observations were conducted on the CFSv2-LETKF (Sluka et al., 2018) to investigate the impacts of SST WCDA on the CFSv2 analysis and the forecasts. (3) The Ensemble Forecast Sensitivity to Observations (EFSO, Kalnay et al., 2012) is a powerful tool to identify the beneficial or detrimental impact of every observation and has been widely used in atmospheric ensemble-based DA. However, EFSO has not yet been applied to any ocean or coupled DA due to the lack of a proper error norm for oceanic variables. This study first introduces a novel density-based error norm that simultaneously includes sea temperature and salinity forecast errors, by which EFSO becomes available to ocean DA for the first time. We implemented the oceanic EFSO on the CFSv2-LETKF for quantifying the individual impact of ocean observations and explored the great potential of EFSO to be extended as a data selection criterion to improve the CFSv2 forecasts.Item REMOTE SENSING OF ATMOSPHERIC TRACE GASES FROM SPACEBORNE UV MEASUREMENTS(2022) Huang, Xinzhou; Yang, Kai; Dickerson, Russell R.; Atmospheric and Oceanic Sciences; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Satellite measurements of atmospheric trace gases provide continuous long-term information for monitoring the atmospheric chemical environment and air quality at local, regional, and global scales. Trace gas retrievals play a critical role in chemical data assimilation, air quality modeling and forecast, and regulatory decision-making. In this dissertation, I present retrievals of three trace gases species (O3, SO2, and NO2) from measurements of Ultraviolet (UV) radiation made from the imaging spectrometers onboard operational satellites, including the Earth Polychromatic Imaging Camera (EPIC) onboard the Deep Space Climate Observatory (DSCOVR), the Ozone Mapping and Profiler Suite - Nadir Mapper (OMPS-NM) onboard Suomi-NPP (SNPP), and the OMPS-NM onboard NOAA-20 satellite. The retrievals of the trace gas vertical columns are achieved through the Direct Vertical Column Fitting (DVCF) algorithm, which is designed to maximize the absorption signature from the Earth’s atmosphere in the UV spectral range. This dissertation first demonstrates the theoretical basis and mathematical procedures of the DVCF algorithm used for retrieving total vertical columns of ozone (O3) and sulfur dioxide (SO2) from DSCOVR EPIC. We describe algorithm advances, including an improved O3 profile representation that enables profile adjustments from multiple spectral measurements and the spatial optimal estimation (SOE) scheme that reduces O3 artifacts resulted from EPIC’s band-to-band misregistrations. Furthermore, we present detailed error analyses to quantify retrieval uncertainties from various sources, assess EPIC-observed volcanic plumes, and validate O3 and SO2 retrievals with correlative data. The second part of this dissertation presents a suite of efforts to retrieve the tropospheric and stratospheric NO2 vertical columns from the new NOAA-20 OMPS hyperspectral Ultraviolet-Visible (UV-Vis) instrument, covering retrieval algorithm, Stratosphere-Troposphere Separation (STS) scheme, measurement sensitivity assessment, inter-comparison with the Ozone Monitoring Instrument (OMI), evaluation with ground-based Pandora spectrometers, as well as a case study of drastic NO2 changes during COVID-19 pandemic. The third part of my dissertation focuses on validation and algorithm improvements for the tropospheric NO2 retrievals from SNPP OMPS UV measurements. OMPS column NO2 was validated against coincidence measurements from two ground-based MAX-DOAS spectrometers deployed in eastern China. To achieve higher retrieval accuracy, we developed and implemented a series of algorithm improvements, including an explicit aerosol correction scheme to account for changes in measurement sensitivity caused by aerosol scattering and absorption, the replacement of climatological a priori NO2 profile with more accurate NO2 vertical distribution from high-resolution CMAQ model simulations, and the application of model-derived spatial weighting kernel to account for the effect of heterogeneous subpixel distribution. These improvements yield more accurate OMPS NO2 retrievals in better agreement with MAX-DOAS NO2 measurements. The analysis concluded that explicit aerosol correction and a priori profile adjustment are critical for improving satellite NO2 observations in highly polluted regions and spatial downscaling is helpful in resolving NO2 subpixel variations.