Repository logo
 

Theses and Dissertations

Permanent URI for this collection

Browse

Recent Submissions

Now showing 1 - 20 of 350
  • ItemOpen Access
    Assimilation of geostationary, infrared satellite data to improve forecasting of mid-level, mixed-phase clouds
    (Colorado State University. Libraries, 2009) Seaman, Curtis J., author; Vonder Haar, Thomas H., advisor
    Mid-level, mixed-phase clouds (altocumulus and altostratus) are difficult to forecast due to the fact that they are generally thin and form in areas of weak vertical velocity where operational models typically have poor vertical resolution and poor moisture initialization. This study presents experiments designed to test the utility of assimilating infrared window and water vapor channels from the Geostationary Operational Environmental Satellite (GOES) instruments, Imager and Sounder, into a mesoscale cloud-resolving model to improve model forecasts of mid-level clouds. The Regional Atmospheric Modeling Data Assimilation System (RAMDAS) is a four-dimensional variational (4-DVAR) assimilation system used to test the viability of assimilating cloudy scene radiances into a cloud-free initial model state for one case of a long-lived, isolated altocumulus cloud over the Great Plains of the United States. Observations from one observation time are assimilated and significant innovations are achieved. Three experiments are performed: (1) assimilation of the 6.7 μm (water vapor) and 10.7 μm (window) channels from GOES Imager, (2) assimilation of the 7.02μm (water vapor) and 12.02 μm (window) channels from GOES Sounder, and (3) assimilation of the 6.7 μm channel from GOES Imager and the 7.02 μm channel from GOES Sounder. It is shown that the GOES Sounder channels provide more useful information than the GOES Imager channels due to increased sensitivity to the mid-troposphere. The decorrelation lengths and variance used in the background error covariance are varied and the impact on the results of the experiments is discussed. The effect of constraining the surface temperatures during assimilation of the window channels is also discussed. It is found that, in a cloud-free initial model state, the adjoint sensitivities are calculated on the assumption that there is no cloud, even with cloud in the satellite observations. This has a significant impact on the success of other 4-DVAR satellite data assimilation experiments.
  • ItemOpen Access
    Quasi-stationary, extreme-rain-producing convective systems associated with midlevel cyclonic circulations
    (Colorado State University. Libraries, 2008) Schumacher, Russ Stanley, author; Johnson, Richard H., advisor
    Observations and numerical simulations are used to investigate the atmospheric processes responsible for initiating, organizing, and maintaining quasi-stationary mesoscale convective systems (MCSs) that form in association with midlevel mesoscale convective vortices or cutoff lows. Six events were identified in which an MCS remained nearly stationary for 6-12 hours and produced excessive rainfall that led to significant flash flooding. Examination of individual events and composite analyses reveals that the MCSs formed in thermodynamic environments characterized by very high relative humidity at low levels, moderate convective available potential energy, and very little convective inhibition. In each case, the presence of a strong low-level jet (LLJ) led to a pronounced reversal of the wind shear vector with height. Convection was initiated by lifting associated with the interaction between the LLJ and the midlevel circulation. One of these events was examined in detail using numerical simulations. This MCS, which occurred on 6-7 May 2000 in eastern Missouri, produced in excess of 300 mm of rain in 9 hours and led to destructive flash flooding. Simulations indicate that the MCS was long-lived despite the lack of a cold pool at the surface. Instead, a nearly stationary low-level gravity wave helped to organize the convection into a quasi-linear system that was conducive to extreme local rainfall amounts. Additionally, the convective system acted to reintensify the midlevel MCV and also caused a distinct surface low pressure center to develop in both the observed and simulated system. To further understand the important processes in these MCSs, idealized simulations using a low-level lifting mechanism and a composite thermodynamic profile are employed. These simulations successfully replicate many of the features of the observed systems. The low-level environment is nearly saturated, which is not conducive to the production of a strong surface cold pool; yet the convection quickly organizes into a quasi-linear system that produces very heavy local rainfall. As in the May 2000 case, a low-level gravity wave was responsible for this organization. The upstream development of new convective cells is shown to result from the interaction of the reverse-shear flow with these waves.
  • ItemOpen Access
    Properties of the tropical hydrologic cycle as analyzed through 3-dimensional k-means cluster analysis
    (Colorado State University. Libraries, 2008) Rogers, Matthew Alan, author; Stephens, Graeme, advisor
    As the primary locations of deep convective activity and unrestrained tropical wave dynamics, the tropical West Pacific and East Indian oceans are among the most important regions in the tropics. Given that most of the region consists of unpopulated expanses of ocean, observations of tropical atmospheric properties in this important region is exceptionally difficult. Only with the help of satellite observations are we capable of gleaning valuable data from this region, and our utilization of advanced analysis techniques allows us to gain more from these observations then would otherwise be possible. In that vein, this dissertation reports on the use of a unique statistical technique, long known to other fields of research, as applied to a combined-instrument satellite observation dataset over the warm pool region of the tropical West Pacific ocean. The statistical technique, known as k-means cluster analysis, is used to delineate self-similar populations of cloud type, hereafter referred to as cloud regimes, from frequency-distribution histograms of cloud-top height, cloud optical thickness, and rainfall amount. We will show that four primary cloud regimes exist in the tropical region discussed, that the four regimes vary primarily through differences in convective activity, and that these four cloud regimes exist in a coherent temporal structure that explains the long-observed variability in convective activity seen in the tropics. Combining this regime information with satellite observations, along with reanalysis data, we then examine the individual properties of each cloud regime. These observations give us the means to understand the forcings behind cloud regime change in the region. We confirm the structural properties of these regimes using analysis from a cloud-resolving model, and apply our new understanding of the mechanism behind this large-scale forcing to the governance of the tropical hydrologic cycle as a whole. The insights gained from this analysis have benefits to both the fields of atmospheric remote sensing, and of cloud- and climate modeling of the tropical atmosphere. Applications of this technique are of particular interest to researchers developing retrieval algorithms for latent heat profiles using active sensors such as the cloud-profiling radar aboard CloudSat.
  • ItemOpen Access
    Making real time measurements of ice nuclei concentrations at upper tropospheric temperatures: extending the capabilities of the continuous flow diffusion chamber
    (Colorado State University. Libraries, 2009) Richardson, Mathews, author; Kreidenweis, Sonia, advisor
    Due to their ubiquity, cirrus clouds are important drivers of climate. Researchers have developed a parameterization that predicts the onset of homogeneous freezing for particles of varying chemical composition. This parameterization is widely used to model cold cloud formation, but the applicability of this parameterization to real atmospheric aerosol has yet to be determined. The field-ready version of Colorado State University's continuous-flow diffusion chamber (CFDC-1H) is one of the few instruments capable of measuring atmospheric ice nuclei concentrations in real time. In this study, we examined the operational limits of the CFDC-1H at low temperature through a series of controlled laboratory studies using (NH 4)2SO4 particles at different operating conditions. We found that residence time played a dominant role in the CFDC-1H's ability to detect the onset of freezing at conditions closer to those predicted. Numerical studies confirmed this and indicated that at warmer temperatures the inability of the CFDC-1H to observe freezing onset conditions as predicted was attributable to the inability of particles to dilute rapidly enough while at colder temperatures the limited availability of water vapor in conjunction with limited residence times inhibited cloud particle growth. The final portion of this study focused on measurements of the freezing onset conditions of an ambient aerosol. Using water uptake measurements, we found that the hygroscopicity (κ) of the ambient aerosol (0.1 to 0.2) was significantly lower than that of ammonium sulfate (0.6). However, as predicted by theory, there was no observably significant difference between the onset conditions of size-selected (NH4)2SO4 and size-selected ambient aerosol. Freezing activation curves for the total ambient aerosol indicated that size plays an important role in the fraction freezing and should be considered when making conclusions regarding chemical composition as a function of fraction freezing. The chemical composition of ice crystal residuals was dominated by mineral type elements and carbon containing particles, contrary to expectations. Further work is necessary for any conclusive statement regarding the chemical composition of the freezing nuclei.
  • ItemOpen Access
    On the role of warm rain clouds in the tropics
    (Colorado State University. Libraries, 2008) Rapp, Anita Denise, author; Kummerow, Christian D., advisor
    A combined optimal estimation retrieval algorithm has been developed for warm rain clouds using Tropical Rainfall Measuring Mission (TRMM) satellite measurements. The algorithm uses TRMM Microwave Imager (TMI) brightness temperatures that have been deconvolved to the 19-GHz field-of-view (FOV) to retrieve cloud liquid water path (LWP), total precipitable water, and wind speed. Resampling the TMI measurements to a common FOV is found to decrease retrieved LWP by 30%. These deconvolved brightness temperatures are combined with cloud fraction from the Visible Infrared Scanner (VIRS) to overcome the beam-filling effects and with rainwater estimates from the TRMM Precipitation Radar (PR). This algorithm is novel in that it takes into account the water in the rain and retrieves the LWP due to only the cloud water in a raining cloud, thus allowing the investigation of the effects of precipitation on cloud properties. The uncertainties due to forward model parameters and assumptions are computed and range from 1.7 K at 10 GHz to about 6K at the 37 and 85 GHz TMI channels. Examination of the sensitivities in the LWP retrieval shows that the cloud fraction information increases the retrieved LWP with decreasing cloud fraction and that the PR rainwater reduces retrieved LWP. Retrieval algorithm results from December 2005 to January 2006 show that warm rain cloud LWP and the ratio of warm rain cloud LWP to rainwater both decrease by 50% over sea surface temperatures (SST) ranging from 292 to 302 K in the tropical western Pacific due to increased precipitation efficiency depleting more of the cloud water at higher SSTs. The LWP retrieval developed in this study is also applied to study the influence of warm rain clouds on atmospheric preconditioning for deep convection associated with tropical depression-type disturbances (TDs). Results show that positive warm rain cloud LWP anomalies occur with positive low-level moistening and heating anomalies prior to TD events, but that there is little variation in the properties of non-raining clouds. The moistening by these clouds is also shown to influence the generation of convective available potential energy (CAPE) prior to deep convection.
  • ItemOpen Access
    The optical, chemical, and physical properties of aerosols and gases emitted by the laboratory combustion of wildland fuels
    (Colorado State University. Libraries, 2008) McMeeking, Gavin R., author; Kreidenweis, Sonia M., advisor
    Biomass burning is a major source of trace gases and particles that have a profound impact on the atmosphere. Trace gases emitted by fires include the greenhouse gases CO2 and CH4, as well as CO and volatile organic compounds that affect air quality. Particle emissions affect climate, visibility, the hydrologic cycle, and human health. This work presents measurements of trace gas and aerosol emissions from a series of controlled laboratory burns for various plant species common to North America. Over 30 fuels were tested through ~250 individual burns during the Fire Laboratory at Missoula Experiment. Emission factors are presented as a function of modified combustion efficiency (MCE), a measure of the fire combustion conditions. The emissions of many trace gas and aerosol species depended strongly on MCE: smoldering-phase combustion dominated fires (low MCE) emitted roughly four times as much gas-phase hydrocarbon species and organic aerosols than flaming-phase dominated fires (high MCE). Inorganic aerosol emissions depended more strongly on plant species and components than on MCE. Flaming-phase dominated fires tended to produce aerosol with high mass fractions of strongly light-absorbing elemental carbon. Smoldering-phase fires produced aerosol with large mass fractions of more weakly light absorbing organic carbon, but this material was found to have a strong wavelength dependence of absorption, greater than the inverse wavelength relationship typically assumed for light absorbing aerosol. A two component model-featuring elemental carbon with a weak wavelength dependence but high mass-normalized absorption efficiency and organic carbon with a strong wavelength dependence but low mass-normalized absorption efficiency-is shown to represent the bulk absorption spectra of biomass burning aerosol. The results show that at wavelengths below ~450 nm, organic carbon light absorption could rival that of elemental carbon for aerosol dominated by organic carbon. If ignored, the light absorption by organic carbon can cause errors in predicted surface ultraviolet and visible radiation fluxes and photochemical photolysis rates in regions affected by biomass burning emissions. The dependence of spectral aerosol optical properties on combustion conditions means that fire behavior must be accurately assessed and predicted to ensure accurate emissions inventories and estimates of biomass burning atmospheric impacts.
  • ItemOpen Access
    A study of the relationship between thunderstorm processes and cloud-top ice crystal size
    (Colorado State University. Libraries, 2008) Lindsey, Daniel T., author; Johnson, Richard H., advisor
    Satellite observations and numerical models are used to understand the physical mechanisms responsible for thunderstorms with varying cloud-top ice crystal sizes. Geostationary Operational Environmental Satellite (GOES) data are used to create a three-year climatology of cloud-top 3.9 µm reflectivity, a quantity which is closely correlated with particle size. Maximum mean values are found over the High Plains and Rocky Mountain regions of the U.S., suggesting that convection over that region tends to generate smaller anvil ice crystals than areas throughout much of the eastern U.S. To correct for preferred forward scattering by the cloud-top ice crystals, an effective radius retrieval using GOES is developed. Forward radiative transfer simulations are run for a wide range of cloud-top ice crystal sizes and sun-cloud-satellite scattering angles. The output is used to generate a lookup table, so that GOES-measured radiances may be used along with sun-satellite geometry to obtain an estimate for particle size. Validation of the retrieval shows that the assumed scattering properties perform quite well. To help explain the geographical variation in cloud-top ice crystal size, a composite analysis is performed in the High Plains region by averaging environmental conditions for days which produced both small and large ice crystal storms. Small ice is found to occur with relatively high based storms and steep mid-level lapse rates. Additionally, observational evidence from a pyrocumulonimbus event is presented to show the effect of low-level cloud condensation nuclei (CCN) on cloud-top ice crystal size. Model simulations using the Colorado State University Regional Atmospheric Modeling System (RAMS) are performed to help understand the physical mechanisms responsible for cloud-top ice crystal size. Through a series of sensitivity tests, it is found that larger low-level CCN concentrations lead to smaller anvil ice. In addition, as cloud-base temperature decreases (and cloud-base height increases), storm-top ice crystals get smaller. A weaker updraft strength is found to have very little effect on ice crystal size.
  • ItemOpen Access
    Improving numerical weather prediction: error growth at the convective scale and speed
    (Colorado State University. Libraries, 2009) Leoncini, Giovanni, author; Pielke, Roger A., Sr., advisor
    Despite the continuous growth of the available computational power, it is undoubtedly beneficial, for both the research and operational communities, to increase the efficiency of Numerical Weather Prediction. Because parameterizations often occupy a significant portion of the total execution time the first focus of this work is to provide a methodology to transform parameterizations into algorithms that provide the same output at a fraction of the computational cost (i.e., transfer schemes). Several transfer schemes are developed for the Harrington radiation parameterization, in the clear sky case and implemented in the Regional Atmospheric Modeling System. The best one requires roughly 5% of the computational expense of the parent scheme. Accuracy is generally preserved and an analysis of the main meteorological fields after two days of simulations does not show significant differences. The differences for the 2 m temperature are larger than for the other fields, but still smaller than the differences introduced by a second common parameterization. A second area where NWP is in need of improvements is convective-scale forecasting. The advantages of more accurate forecasting derive from the high societal impact of convective events, which can be severe and lead to loss of life and property. Ensemble forecasting is an ideal tool to handle uncertainties in forecasts and the second aim of this study is to identify the processes that lead to error growth at the convective scale, for a case study over the United Kingdom using the Met Office Unified Model. The perturbation was applied to the potential temperature at a specific model level within the boundary layer, either sequentially (every 30 minutes) or at specific times. It was determined that acoustic waves are generated and can affect the background state. Vertical stability is also altered and occasionally lids can be set or removed. The unique boundary-layer scheme also contributes to error growth, by triggering different parameterizations as a response to the perturbation. Finally there are qualitative differences between high amplitude perturbations (1 K) and the smaller ones (0.01 and 0.1 K), but the root mean square error reaches similar values at saturation.
  • ItemOpen Access
    The impact of natural dust aerosol on warm and cold cloud formation
    (Colorado State University. Libraries, 2008) Koehler, Kirsten, author; Kreidenweis, Sonia M., advisor
    Dust particles' ability to scatter and absorb radiation and their potential to interact with water vapor may lead to important direct and indirect radiative impacts on the climate. Due to differences in solubility, hygroscopicity, chemical and surface properties, dust particles from different regions may interact with cloud development in a variety of ways that are not fully understood. In this work four types of dust from the Southwestern U.S. and Northern Africa were studied. The characteristics of the samples used cover a range of soluble contents, geographic locations of origin, and collection methods. Hygroscopic growth of the particles was determined using a humidified tandem differential mobility analyzer (HTDMA) at relative humidities (RH) from 5-95% and cloud condensation nuclei (CCN) activity was measured using a continuous flow CCN counter at supersaturations up to 2%. At cold temperatures (-60
  • ItemOpen Access
    Tropical cyclone evolution via internal asymmetric dynamics
    (Colorado State University. Libraries, 2008) Hendricks, Eric A., author; Schubert, Wayne, advisor
    This dissertation advances our understanding by which tropical cyclones (TCs) evolve solely due to internal dynamics, in the absence of large-scale environmental factors and surface fluxes, using a hierarchy of numerical model simulations, diagnostics and observations. In the first part, the role of inner-core (eye and eyewall) transport and mixing processes in TC structure and evolution is examined, and in the second part, some asymmetric dynamics of tropical cyclone evolution are studied: spontaneous inertia-gravity wave radiation from active TC cores and an observational case study of the role of vortical hot towers in tropical transition. The role of two-dimensional transport and mixing in TC structure and intensity change is quantified. First, the mixing properties of idealized hurricane-like vortices are assessed using the effective diffusivity diagnostic. Both monotonic and dynamically unstable vortices are considered. For generic deformations to monotonic vortices, axisymmetrization induces potential vorticity (PV) wave breaking outside the radius of maximum wind, forming a finite radial length surf zone characterized by chaotic mixing. Although on a much smaller scale, this surf zone is analogous to the surf zone outside the wintertime stratospheric polar vortex. For unstable rings, during barotropic instability both the inner and outer breaking PV waves create horizontal mixing regions. For thin ring breakdowns, the entire inner-core becomes a strong mixing region and passive tracers can be transported quickly over large horizontal distances. For thick ring breakdowns, an asymmetric partial barrier region may remain intact at the hurricane tangential jet, with mixing regions on each side where the waves break. The inner, breaking PV wave is quite effective at mixing passive tracers between the eye and eyewall; with a monotonic low-level equivalent potential temperature radial profile, these results support the hurricane super-intensity mechanism. Next, a systematic study of inner-core PV mixing resulting from unstable vortex breakdowns is conducted. After verifying linear theory, the instabilities are followed into their nonlinear regime and the resultant end states are assessed for 170 different PV rings, covering a wide spectrum of real hurricanes.
  • ItemOpen Access
    The near-global distribution of light precipitation from CloudSat
    (Colorado State University. Libraries, 2008) Haynes, John M., author; Stephens, Graeme L., advisor
    The W-band (94 GHz) Cloud Profiling Radar (CPR) on CloudSat is sensitive to both clouds and precipitation. A precipitation retrieval applicable to space-borne, millimeter wavelength radars is introduced. Measurements of the attenuated backscatter of the surface are used to derive the path integrated attenuation (PIA) through precipitating columns, which follows from the clear-sky scattering characteristics of the surface. Over ocean, this can be estimated as a function of near-surface wind speed and sea surface temperature. Assuming an exponential rain drop size distribution, the relationship between PIA and rain rate is derived from Mie theory for homogeneous columns of warm rain. Multiple scattering is found to be significant for rainfall rates exceeding 3 to 5 mm h-1. To correct for this effect, Monte Carlo modeling is used to simulate the relationship between rainfall and PIA for various vertical precipitation profiles. Multiple scattering is found to increase return power to the radar, acting opposite attenuation. A model of the melting layer is also incorporated to better represent attenuating characteristics near the bright band, where snow aggregates melt into rain. It is found that failure to account for extra attenuation caused by melting particles results in overestimation of precipitation rate. The retrieval algorithm is applied to near-global CloudSat observations. Precipitation in the tropics is found to prefer clouds with lowest-layer cloud tops near 2 and 15 km. A third mode, likely associated with congestus, is found to be common in the tropical western Pacific, Indian, and Atlantic basins. There are vast regions of the globe where nearly all precipitation falls from cloud with lowest-layer tops below 4.75 km. Over the tropical oceans as a whole, precipitation falls twice as often from these clouds as any other cloud type. Furthermore, multiple layered cloud systems are found to be ubiquitous globally. In the tropics, it is estimated that half the accumulated precipitation comes from multiple layered systems rather than the classic "deep convective" model. Outside the tropics, the CPR observes precipitation more often than the passive microwave AMSR-E, with greater resulting seasonal accumulations.
  • ItemOpen Access
    Characteristics of precipitation: CloudSat observations and model predictions of the current and future climate
    (Colorado State University. Libraries, 2008) Ellis, Todd Douglas, author; Stephens, Graeme L., advisor
    The overall purpose of this study is to examine the characteristics of precipitation as they are predicted to change in a typical climate change scenario and as they exist now and how well model reproduces those observations. The first part of this study examines the controls on global precipitation evident in a transient carbon dioxide doubling experiment conducted using coupled climate models collected for the Intergovernmental Panel on Climate Change Fourth Assessment Report (IPCC AR4). As noted in other studies, the ensemble mean changes in water vapor occur at a rate more than three times that of precipitation. A simple ratio of these changes is introduced as a type of measure of the efficiency of the atmospheric hydrologic cycle in responding to changes in moisture, and varies between about 0.09 and 0.25 for the models studied. It is shown that the change in precipitation sensitivity is primarily governed by how emission of radiation from the clear-sky atmosphere increases as water vapor increases. This relationship closely matches one derived from simple energy balance arguments involving changes to water vapor emission alone. The study also presents the precipitation incidence over the global oceans as calculated from the CloudSat satellite, showing precipitation into the high latitudes and calculating that precipitation occurs 11% of the time over the oceans. These data are verified using ship-based (ICOADS) and island-based (GSOD) data. This study then extends the use of these data to an analysis of the observed cloud structures that are associated with rainfall over the oceans and then comparing them to special runs of the ECMWF weather forecast and HadGAM1 climate prediction models. These comparisons show that the models don't predict shallow precipitation nor layered precipitation structures as often as they are observed, and predict incorrect global precipitation incidences.
  • ItemOpen Access
    Analysis and application of the CASA IP1 X-band polarimetric radar network
    (Colorado State University. Libraries, 2009) Dolan, Brenda, author; Rutledge, Steven A., advisor
    The Collaborative Adaptive Sensing of the Atmosphere's Integrated Project 1 (CASA IP1) network of four X-band, polarimetric, Doppler, adaptively scanning radars is investigated for studying storm microphysics and kinematics. The complications of non-Rayleigh scattering and attenuation at X-band are explored for impact on microphysical interpretation. The rapid and adaptive scanning strategy is evaluated for application of dual-Doppler techniques to retrieve the 3-D wind field, and general understanding of storm interactions. Several rain rate algorithms are invoked to estimate surface rainfall. A case study from 10 June 2007 illustrates the capabilities and limitations of using the IP1 network for studies of storm interactions, and lightning data are analyzed to relate these interactions to storm electrification. The nearby S-band, polarimetric KOUN radar is studied for comparison. Scattering simulations using the T-matrix model are performed on seven hydrometeor types (excluding hail) to understand the non-Rayleigh effects at X-band compared with S-band. The simulations show the greatest non-linearities in Zdr and Kadp of rain and graupel. Results of the simulations are used to develop a specific X-band fuzzy logic hydrometeor identification algorithm (HID) for diagnosing bulk regions of hydrometeors. Attenuation and non-Rayleigh scattering are present in the IP1 data, but with mitigation techniques these have minimal impact on the analysis. The high temporal resolution is integral in resolving up- and downdrafts, as well as hydrometeor evolution, but the inconsistent and lack of upper-level coverage are significant limitations for quantitative analysis of kinematic and microphysical relationships. Observations using IP1 data of a storm on 10 June 2007 show the development of the updraft, subsequent graupel echo volume evolution, and onset of lightning. Development of the downdraft is preceded by large volumes of graupel in the mid-levels. A second peak in intra-cloud lightning is observed to be associated with an increase in height of the upper positive charge, resulting from a kinematic intensification. Many of these trends are corroborated by KOUN. Rain rate estimation comparisons show that the X-band blended algorithm performs better compared with ground-based sensors than the simple Z-R relationship and employs polarimetric estimators more often than S-band blended methods.
  • ItemOpen Access
    Investigating causes of regional variations in atmospheric carbon dioxide concentrations
    (Colorado State University. Libraries, 2008) Corbin, Katherine D., author; Denning, A. Scott, advisor
    Atmospheric CO2 concentrations are rapidly increasing due to anthropogenic activities; however, only about half of the emissions have accumulated in the atmosphere, and the fate of the remaining half remains uncertain. Since atmospheric CO2 concentrations contain information regarding carbon sources and sinks, it is important to understand CO2 variability. This study investigated causes of atmospheric CO2 variability, focusing on the relationship between CO2 concentrations and clouds, the impact of heterogeneous land cover and agricultural production, and the effect of redistributing fossil fuel emissions. Due to global coverage and sheer data volume, satellite CO2 concentrations will be used in inverse models to improve carbon source and sink estimates. Satellite concentrations will only retrieve CO2 measurements in clear conditions, and it is important to understand how CO2 concentrations vary with cloud cover in order to optimally utilize these data. This study evaluated differences between clear-sky and mean concentrations on local, regional, and global scales. Analyses of in situ data, regional model simulations, and global model output all revealed clear-sky differences that were regionally coherent on sub-continental scales and that varied both with time and location. In the mid-latitudes, clear-sky CO2 concentrations were systematically lower than on average, and these differences were not due to biology, but rather to frontal convergence of large-scale gradients that were covered by clouds. Instead of using satellite data to represent temporal averages, inverse models and data assimilation systems that use satellite data to calculate carbon sources and sinks must be sampled consistently with the observations, including precise modeling of winds, clouds, fronts, and frontal timing. Just as CO2 concentrations vary with cloud cover, variability in atmospheric CO2 concentrations is also caused by heterogeneity in land cover and surface fluxes. This study focused on the impacts of land-cover heterogeneity and the effects of agricultural production on regional variations of atmospheric CO2 concentrations. Including sub-grid scale land cover heterogeneity improved simulated atmospheric CO2 concentrations by ~ 1 ppm. Implementing a crop-phenology model that explicitly simulated corn and soybeans into a coupled ecosystem-atmosphere model dramatically improved CO2 fluxes and concentrations over the mid-continent, with reductions in CO2 concentration root mean square errors of nearly 50% (over 10 ppm at some locations). Both the model and observations showed concentrations as low as 340 ppm over central Iowa, and a regional gradient of over 30 ppm in ~ 200 km occurred due to a combination of fluxes and meteorology. Since corn and soybeans have such a significant impact on both carbon fluxes and atmospheric concentrations, it is essential to model these crops accurately. In addition to biological surface fluxes, surface emissions due to fossil fuel combustion also cause variability in regional atmospheric CO2 concentrations. Using high-resolution fossil fuel emissions caused differences of over 10 ppm near the surface; and including temporal variability in the emissions impacted regional CO2 concentrations on monthly timescales, causing seasonal differences of more than 20 ppm in some locations. Using coarse spatial distributions and unaccounting for temporal variability in fossil fuel emissions created biases in the atmospheric CO2 concentrations and thus may cause significant errors in source and sink estimates from atmospheric inversions.
  • ItemOpen Access
    Mechanisms of observed sea surface temperature variability in the extratropical southern hemisphere
    (Colorado State University. Libraries, 2008) Ciasto, Laura M., author; Thompson, David, advisor
    The physical mechanisms that drive sea surface temperature (SST) variability in the extratropical Southern Hemisphere (SH) are examined using multiple ocean temperature datasets. The first part of the study provides a detailed analysis of the relationships between variability in SH SST anomalies, the Southern Annular Mode (SAM) and the El-Niño/Southern Oscillation (ENSO) during the warm (November-April) and cold (May-October) seasons. It is shown that the signatures of the SAM and ENSO in the SST field vary as a function of season, both in terms of their amplitudes and structures. SAM-related SST anomalies are primarily driven by surface turbulent heat fluxes with a smaller contribution from heat advection by Ekman currents. The role of turbulent heat fluxes in generating ENSO-related SST anomalies is less clear. Analyses of the temporal evolution of the relationships between the SAM and the SST field demonstrate that SST anomalies are largest when SSTs lag by ~1 week and persist for up to 8 weeks. In the absence of ENSO, cold season SAM-related SST anomalies persist longer than their warm season counterparts, consistent with seasonal variations in the depth of the mixed layer. The second part of the study uses observations of subsurface temperatures to examine the winter-to-winter "reemergence" of SST anomalies in the extratropical South Pacific. Reemergence is the mechanism whereby SST anomalies formed in the late winter are sequestered beneath the shallow summer mixed layer and then re-entrained into the deepening mixed layer during the following fall/winter. The results exhibit a pronounced reemergence signal in which surface temperature anomalies during the late winter season are strongly correlated with surface temperature anomalies during the subsequent early winter months, but are only significantly correlated with temperature anomalies beneath the mixed layer during the intervening summer months. The results are robust to small changes in the period of analysis and are qualitatively similar to existing evidence of reemergence in the Northern Hemisphere. The signal of reemergence evident in the subsurface data is readily apparent in SST data in the western South Pacific. Reemergence is less evident in SST data in the eastern South Pacific.
  • ItemOpen Access
    The atmospheric circulation response to climate change-like thermal forcings in a simple general circulation model
    (Colorado State University. Libraries, 2009) Butler, Amy Hawes, author; Thompson, David W. J., advisor
    Temperature changes due to increased greenhouse gases and depleted stratospheric ozone are associated with robust changes in the large-scale atmospheric circulation. In this thesis we explore how these anthropogenically-driven temperature changes affect the atmospheric circulation. Our approach is to force a simple dry dynamical general circulation model (GCM) with idealized thermal forcings that resemble three key effects of greenhouse gas increases and stratospheric ozone depletion: warming at the polar surface, warming of the tropical upper troposphere, and cooling of the polar stratosphere.
  • ItemOpen Access
    Model evaluation using space-borne lidar observations
    (Colorado State University. Libraries, 2008) Ahlgrimm, Maike, author; Randall, David A., advisor
    In this study, the use of space-borne lidar observations for the comparison with, and evaluation of modeled clouds is explored. Four version of the ECMWF Integrated Forecast System and two versions of the Goddard Earth Observing System (GEOS-5) model are assessed for their ability to produce marine boundary layer clouds. The cause of some of the model deficiencies is investigated, and specific suggestions for improvements are made and tested. In order to do so, two cloud types are defined: a stratocumulus type (Sc), and a trade cumulus or transitional cumulus type (TCu). Samples in four oceanic regions are classified into those categories, and the frequency of occurrence, location, and properties of the samples compared between models and observations.
  • ItemOpen Access
    Comparing precipitation estimates, model forecasts, and random forest based predictions for excessive rainfall
    (Colorado State University. Libraries, 2023) James, Eric, author; Schumacher, Russ, advisor; Bell, Michael, committee member; Van Leeuwen, Peter Jan, committee member; Morrison, Ryan, committee member
    Flash flooding is an important societal challenge, and improved tools are needed for both real-time analysis and short-range forecasts. We present an evaluation of threshold exceedances of quantitative precipitation estimate (QPE) and forecast (QPF) datasets in terms of their degree of correspondence with observed flash flood events over a seven-year period. We find that major uncertainties persist in QPE for heavy rainfall. In general, comparison with flash flood guidance (FFG) thresholds provides the best correspondence, but fixed thresholds and average recurrence interval thresholds provide the best correspondence in certain regions of the contiguous US (CONUS). QPF threshold exceedances from the High-Resolution Rapid Refresh (HRRR) generally do not correspond as well as QPE exceedances with observed flash floods, except for the 1-h duration in the southwestern CONUS; this suggests that high-resolution model QPF may be a better indicator of flash flooding than QPE in some poorly observed regions. Subsequently, we describe a new random forest (RF) based excessive rainfall forecast system using predictor information from the 3-km operational HRRR. Experiments exploring the use of spatial predictor information reveal the importance of averaging HRRR predictor fields across a spatial radius rather than using only information from sparse input grid points for regimes with small-scale excessive rain events. Tree interpreter results indicate that the forecast benefits of spatial aggregation stem from greater contributions provided by storm attribute predictors. Forecasts are slightly degraded when there is a mismatch between the trained RF model and the daily HRRR forecasts to which the model is applied, both in terms of initialization time and HRRR model version. Use of FFG as an additional predictor leads to forecast improvements, highlighting the potential of hydrologic information to contribute to forecast skill. In addition, averaging predictor information across several HRRR initializations leads to a statistically significant improvement in forecasts relative to using predictor fields from a single HRRR initialization. The HRRR-based RF has been evaluated at the annual Flash Flood and Intense Rainfall Experiment (FFaIR) over the past three years, with year-over-year improvements stemming from the results of sensitivity experiments. The HRRR-based RF represents an important baseline for future machine learning based excessive rainfall forecasts based on convection-allowing models.
  • ItemEmbargo
    Marine ice nucleating particles: sources, composition, emissions, and model parameterizations
    (Colorado State University. Libraries, 2023) Moore, Kathryn A., author; Kreidenweis, Sonia M., advisor; DeMott, Paul J., advisor; Farmer, Delphine K., committee member; Pierce, Jeffrey R., committee member; van den Heever, Susan C., committee member
    Sea spray aerosol has received increasing attention over the last decade as a source of ice nucleating particles (INPs) to the atmosphere. Sparse measurements in remote marine regions indicate both marine INP concentrations and ice nucleating efficiency are several orders of magnitude lower than those of mineral or soil dusts, which dominate the INP budget on a global scale. The Southern Ocean (SO) surrounding Antarctica is thought to be the only region where marine INPs are the predominant INP type due to its remoteness from continental and anthropogenic aerosol sources and persistent strong westerlies, although several recent studies have suggested this may also be true of the high Arctic seasonally or intermittently. INPs are critical for initiating cloud glaciation at temperatures warmer than ~-36 °C and can thus have an outsize effect on cloud phase and related climate feedbacks due to their relative scarcity. This is particularly true over the polar oceans, where low and mid-level mixed phase and supercooled clouds are ubiquitous and especially sensitive to aerosols due to the generally low background particle concentrations. The research presented here aimed to improve our understanding of the factors influencing marine INP emissions and the sources and composition of INPs in remote marine regions, as well as to evaluate and improve current INP model parameterizations. This was accomplished using observations made in the Southern Ocean, one of the few remaining pristine aerosol environments, during the Southern Ocean Cloud Radiation Aerosol Transport Experimental Study (SOCRATES) aircraft campaign on the NSF/NCAR G-V, and the second Clouds, Aerosols, Precipitation, Radiation and atmospherIc Composition Over the southeRN ocean (CAPRICORN-2) ship campaign on the R/V Investigator in 2018. Ambient observations were supplemented by measurements from the CHaracterizing Atmosphere-Ocean parameters in SOARS (CHAOS) mesocosm experiment in the new Scripps Ocean-Atmosphere Research Simulator (SOARS) wind-wave channel. CHAOS measurements allowed for isolation of the role of wind speed in marine INP production, which had not previously been characterized through controlled experiments. SOCRATES and CAPRICORN-2 are notable for collecting the first vertically resolved INP measurements over the Southern Ocean, including the first in situ observations in and above cloud in the region. Both aerosol and INP concentrations showed excellent agreement between G-V and R/V Investigator observations during overflights of the ship, supporting the use of such a multi-platform measurement approach for future campaigns interested in aerosol and INP vertical profiles. New techniques for estimating marine aerosol surface area and the number of particles >0.5 μm, key quantities often used in INP parameterizations, were developed based on lidar and nephelometer measurements. An additional parameterization for marine INPs is proposed, which uses both wind speed and activation temperature, and reduces bias compared to the existing parameterization based solely on temperature. Marine boundary layer (MBL) and above cloud INP concentrations from the same SOCRATES flight support the hypothesis suggested by several modeling studies that marine INPs dominate at low altitudes, and mineral dust becomes increasingly important with height. Unexpectedly, enhanced INP and aerosol iron concentrations, but low iron solubilities, were observed for samples collected south of 60 °S during CAPRICORN-2. Antarctica is suggested as a potential source of both biological and inorganic INPs to the Southern Ocean marine boundary layer through the emission of mineral and soil dusts from ice-free areas. Similar high latitude dust sources in Iceland and Svalbard have been observed to contribute to INPs in the Arctic atmosphere, and are anticipated to increase in importance as the climate warms.
  • ItemOpen Access
    Using laboratory and airborne measurements to investigate the role of ice nucleating particles in ice and mixed-phase clouds
    (Colorado State University. Libraries, 2023) Patnaude, Ryan John, author; Kreidenweis, Sonia M., advisor; DeMott, Paul J., advisor; van den Heever, Susan C., committee member; Chui, J. Christine, committee member; Willis, Megan D., committee member
    Ice may be present in the atmosphere either in cirrus or mixed-phase cloud regions, each with their own distinctly different characteristics and formation mechanisms. The former is characterized by the presence of only ice crystals at temperatures < -38 °C, while the latter includes the coexistence of both supercooled liquid cloud droplets and ice crystals between temperatures of 0 °C and -38 °C. Cirrus clouds represent an important cloud type as they are ubiquitous in the atmosphere and their radiative effects depend upon their microphysical properties. Their formation mechanisms may proceed via homogeneous or heterogeneous nucleation, and whether one or the other or both occur determines the size and number of ice crystals. The ocean represents one of the largest sources of aerosols into the atmosphere, and sea spray aerosols (SSA), if they are lofted to the upper troposphere, may act as ice nucleating particles (INPs) to initiate heterogeneous nucleation under cirrus conditions. Although a number of previous studies have investigated the ice nucleating behavior of SSA proxies such as sodium chloride (NaCl), or SSA generated from commercially-available artificial seawater products, ice nucleation under cirrus conditions of SSA generated from natural seawater had not been examined at the inception of this research program. Additionally, whether secondary marine aerosols (SMA), which form via the gas-to-particle conversion of ocean-emitted gas-phase species, may act as an INP in cirrus clouds is currently unknown. The first half of this dissertation highlights two laboratory studies that investigated the role and characteristics of SSA and SMA to act as INPs at cirrus cloud temperatures. The first study compared ice nucleation results for submicron SSA and NaCl particles and examined whether particle size affected the low temperature ice nucleation. Results showed that both SSA and NaCl initiated heterogeneous nucleation strongly at temperatures below 220 K, and that the size of the particles did not affect the ice nucleating ability of SSA. The similarities between the freezing behaviors of SSA and NaCl particles suggested the salt components were controlling heterogeneous ice nucleation. The second study used a more realistic aerosol generation method, utilizing a Marine Aerosol Reference Tank (MART) that was filled with natural seawater, and investigated the effects of atmospheric oxidation on SSA using an oxidation flow reactor (OFR), which was also used to generate SMA from gaseous emissions released in the MART. SMA alone were also examined for their ice nucleation behavior at cirrus temperatures. Results from this study indicated that atmospheric oxidation did not hinder low temperature ice nucleation of SSA, and that SMA are not efficient ice nucleating particles at cirrus temperatures, but could participate in homogeneous nucleation. Finally, the similarities between the findings from the two studies indicated that the generation method of SSA, and any impacts on SSA organic aerosol content, did not affect the ice nucleating behavior of SSA at cirrus temperatures. Ice in mixed-phase clouds (MPCs), on the other hand, forms initially via heterogeneous nucleation at a wide range of temperatures and relative humidity conditions, depending on the abundance and characteristics of available INPs. Secondary ice production (SIP) may follow heterogeneous nucleation in MPCs, where new ice crystals form either during the heterogeneous freezing event, or through subsequent interactions between the pre-existing liquid cloud droplets and ice crystals. SIP may lead to enhanced ice crystal number concentrations via a number of proposed mechanisms, especially in convective environments. Despite decades of study toward developing better understanding of ice formation in MPCs, the freezing pathways of ice crystals over the course of cloud lifetimes, and the conditions that favor the various proposed SIP pathways, are not fully resolved. The third study in this dissertation reports and interprets observations of INPs during an airborne campaign over the U.S. Central Great Plains during the Secondary Production of Ice in Cumulus Experiment (SPICULE) campaign that primarily sampled cumulus congestus clouds. Coincident measurements of INP and ice crystal number concentrations in cumulus congestus clouds were used to infer the ice formation pathway, either through heterogeneous nucleation or SIP. Warmer cloud base temperatures and faster updrafts were found to facilitate environmental conditions favorable for SIP. Further, the fragmentation of freezing droplets (FFD) SIP mechanism was found to be critical in the enhancement of observed ice crystal number concentrations during the earliest stages of the cloud lifetime. Numerical model simulations of an idealized, single congestus cloud, designed to mimic the clouds sampled during SPICULE, were conducted with newly-implemented SIP mechanisms, added to the existing Hallet-Mossop (HM) rime-splintering mechanism. The model results indicated that HM dominated the production of ice crystals, but without the FFD and ice-ice collisional breakup (BR) SIP mechanisms, the model could not accurately resolve ice crystal number concentrations compared to observations. Competing results in the dominant SIP mechanisms underscore the need for improved mechanistic understanding of these SIP processes, either through laboratory or observational studies, in order to close this gap between model prediction and observations. The final portion of this dissertation describes airborne observations of INPs during a field campaign along the U.S. Gulf Coast, also aimed at investigating the impacts of various aerosol-cloud interaction mechanisms on development of convective clouds. During this campaign, a widespread and prolonged Saharan Air Layer (SAL) event took place and INP characteristics during this event are reported and contrasted with INP characteristics prior to the arrival of the SAL. The INP concentrations at temperatures below -20 °C were enhanced by 1–2 orders of magnitude compared to the flights prior to the dust intrusion, and showed good agreement with one previous study of Saharan dust near Barbados, but lower INP concentrations than another study off the coast of western Africa. The INP concentrations in the SAL also generally overlapped with or exceeded INP concentrations during SPICULE, but only for INPs active temperatures < -25 °C. These observations were the first airborne measurements in nearly two decades tagging INP concentrations to North African dust that had been transported all the way to the United States. Further, they provide the most comprehensive description of these INPs yet recorded, and suggest a common natural INP perturbation in the southeastern U.S. and Gulf regions in early summer, with implications for cloud processes that warrant further study.