Sandra Yuter

Northeastern US snowstorms impact large populations in major urban corridors, sometimes up to 100 million people and ~400,000 sq. miles. Major snowstorms have large economic impacts on states, costing as much as $300-700 million per snow-shutdown day. Improving snowstorm prediction, including quantitative precipitation forecasts, can provide significant economic and societal benefit. A 100-200 km error in the forecast of the rain-snow line or locations of snow bands, or inadequate characterization of the microphysical growth regimes, can result in substantial errors in forecasts of precipitation type and quantity Given that no major winter storm science campaign has taken place along the East Coast in 30 years, and no campaign there has been dedicated specifically to snowfall, now is the time to apply NASA������������������s unique precipitation observing systems through The Investigation of Microphysics and Precipitation for Atlantic Coast-Threatening Snowstorms (IMPACTS) mission to improve understanding of snowfall processes, remote sensing of snow, and prediction of band structure and evolution. Multiscale snowbands within winter storms are the primary focal points for intense snowfall accumulations, but are poorly understood and therefore difficult to forecast. IMPACTS comprises a set of aircraft and payloads ideally-suited to the study of eastern US snowstorms. Using measurements from a ����������������satellite-simulating��������������� (top-of-the-atmosphere vantage point with instruments meeting or exceeding the capability of satellite sensors) ER-2 and in-situ measurements from a cloud-penetrating P-3 aircraft, augmented with ground-based radar and rawinsonde observations, data from multiple NASA satellites, convection-permitting regional analyses and short-term forecasts, IMPACTS will provide important observations and research for understanding the mechanisms of snowband formation and evolution within winter storms, provide observations that enable future improvements in measuring snowfall from space, and provide important data to validate and improve numerical models.

Global and regional weather forecast models struggle to consistently reproduce key characteristics of the marine boundary layer, including boundary layer height, cloud thickness, and temperature and moisture profiles. These parameters have direct relevance to Navy applications including ship-based activities, aircraft and drone operations, and electromagnetic and electro-optical propagation. Most model evaluation efforts contrast regional averages for all weather conditions over periods of months or longer when comparing between model output and observations. To complement those efforts, the proposed work will examine errors as a function of weather conditions. Conditioning on specific phenomena permits isolation of subsets of the model output where particular processes are more or less prominent which reduces complexity for diagnosis of error sources. To avoid the conundrum of using high uncertainty reanalysis grid values to evaluate forecast models in conditions where forecast models perform poorly, our approach is to use observations. We will build a relational database of matched observations and model forecasts at locations throughout the continental United States including offshore buoys to facilitate flexible data queries. The goals are 1) to characterize the performance of the Navy������������������s weather models compared to observations and to National Oceanic and Atmospheric Administration operational weather models and 2) to identify processes within the weather models that most need refinement.

During the winter, the northeast coast of the U.S. from Maryland to Maine typically experiences 5 to 10 widespread storms containing a mixture of snow, rain, sleet, and freezing rain. While synoptic scale structures associated with fronts within these storms are reasonably well understood, the interplay among instability, dynamics, and microphysics at the mesoscale is much less clear. Sets of multiple mesoscale snow bands modulate snow accumulation. An incomplete understanding of mesoscale snow band genesis, maintenance, and demise is a key contributor to the large uncertainties (200% or more) in snow accumulation forecasts. This project builds on previous work including a 108 storm climatology of snow bands for the northeast U.S. coast and analysis of millions of photographs of snow particles in free fall that yielded clues to where and how snow forms within the storm. The new project will utilize several new observational facilities centered on Stony Brook, NY including research radars to obtain high spatial and time resolution data, upper air sounding launches to help map the 3D temperature and stability structures within the storms, and an array of pressure sensors to detect atmospheric gravity waves. We will test hypotheses addressing aspects of mesoscale snow band lifecycle related to the role of atmospheric gravity waves, different sources of instability, and amplifying feedbacks among turbulent kinetic energy and release of latent heat during precipitation formation.

Marine stratocumulus cloud variability at multi-day (~3-50 days) timescales constitutes a considerable fraction of the total low-cloud variability, yet most previous studies have focused on longer seasonal time scales. External perturbations such as synoptic advection and gravity waves can strongly influence marine low clouds. We have recently documented extensive, abrupt boundaries in cloudiness >1000 km long in the southeast Atlantic that propagate westward from the coast of Africa. The cloud boundaries appear to be a gravity wave phenomenon that erodes cloud. Our framing hypothesis is that cloud-system resilience to external perturbations and the frequency of these perturbations are key factors in governing low-cloud variability and potential climate sensitivity. The proposed research will lay the groundwork for the study of dynamically-modulated, rapid, large-scale cloud eroding phenomena in marine low clouds.

Weather forecasters make time-sensitive decisions based in large part on weather prediction models that have varying levels of uncertainty for different weather situations and locations. Forecasters have potentially dozens of numerical weather predictions to consult. Different weather prediction models can have substantial disagreements particularly when looking ahead 120 hours. There is not time to look at everything carefully. How to weigh each weather prediction input is a complex task. We propose to build a tool to aid Delta Meteorology forecaster work-flow by distilling and assessing recent trends and historical weather prediction model performance. The deliverable is an adaptive system that will highlight more skillful models for a given situation and generate alerts for weather with large uncertainties that will benefit from focused forecaster attention.

Improved forecasting of precipitation within extratropical cyclones requires continued integration of observations and models to understand the evolution and processes associated with precipitation bands. Mesoscale banding within the comma-head portion of Northeast U.S. cyclones has been documented to cause localized intense precipitation rates and amounts. Our group and others have documented the life cycle and microphysics of the primary band within the comma head in the presence of mid-level frontogenesis and reduced stability, but there has been less attention investigating storms with multiple bands. The underlying hypothesis is that many of the quantitative precipitation forecast (QPF) errors associated with winter storms are the result of difficulties in models predicting these multi-bands given their smaller spatial scales, more convective characteristics, and their upscale growth to a single band. We propose an integrated modeling and observational study to understand the spectrum of precipitation features within extratropical cyclones over the Northeast United States. We will use ~20 years of radar data from the WSR-88Ds, several years of Terminal Doppler Weather Radar (TDWR) at the major airports, upper air soundings, observations from a vertically pointing radar, a disdrometer, and Multi-Angle Snowflake Camera (MASC) at Stony Brook, NY on Long Island, as well as high resolution gridded analyses and Weather Research and Forecasting (WRF) predictions to characterize banded precipitation and ice microphysics and to evaluate model output. By the end of the study we expect to be able to characterize band lifecycle and tracks of cells and bands around the cyclone and to quantify the importance to storm precipitation accumulation of multi-banded structures relative to other precipitation features. We will also identify what set of parameters are needed to realistically model multi-banded precipitation. Our proposed work will address the parameterization of ice microphysics and mesoscale predictability of banded precipitation using an ensemble of historical runs and sensitivity studies.

Nowcasting convective initiation (CI) and the intensification of convection to severe criteria (i.e., tornadoes, large hail, high winds) pose significant challenges to forecasters in the United States (U.S.). Much of this difficulty stems from the lack of available data (surface and remote sensing) on spatial and temporal scales suitable to observe rapid changes in storm structure or a lack of understanding of the processes. In particular, tornado detections and warning lead times for the northeast U.S. are significantly worse than across the rest of the U.S. In this northeast region, there are additional and unique challenges to forecasters, including the lack of continuous and reliable total lightning flash data, the impact of surface boundaries, maritime air, and variable terrain on storm development and evolution, and the ubiquity of shallow severe convection. Because of these difficulties, little previous work has been done on nowcasting storm evolution in the region, nor on determining the preferred areas where CI and storm intensification occur. GOES-R data, including high spatial and temporal resolution visual and infrared imagery from the Advanced Baseline Imager (ABI), and total lightning flash data from the Geostationary Lightning Mapper (GLM), should enhance our ability to identify developing convection and introduce several new tools that may serve as near-real-time proxies for the strength of convective updrafts in the Northeast. As a result, we see an opportunity to significantly advance convective nowcasting in this region. Specifically, using data from the 2016-17 convective seasons in the Northeast, we aim to (i) test the regional effectiveness of GOES-R CI algorithms, (ii) employ a novel polarimetric radar algorithm to determine how well various ABI and GLM products can indicate changes in storm updraft strength, and, therefore, storm severity, and (iii) use cell tracking techniques to link the two processes and identify preferred areas of CI and severe intensification in the region. The ABI data also will allow for the effects that terrain and coastal boundaries have on CI and storm intensification to be studied in great detail. In all of the above areas of study, there will be a focus on the similarities and differences between tornadic, nontornadic severe (large hail and damaging winds), and non-severe storms. In accordance with the solicitation, the proposal meets both of the broad severe storm objectives listed. If the ABI and GLM data can be demonstrated to be useful proxies for updraft intensity in the northeast convective environment, the nowcasting of storm intensification to severe criteria, including the transition of some storms from nontornadic to tornadic, is likely to be enhanced. Second, a demonstration of GOES-R utility to forecasters, both through the aforementioned intensification aspect and better identification of where and why CI and severe transitions occur, promises future forecaster use, improved short-term forecasts and nowcasts of these processes, and, therefore, more accurate warnings.

The growing popularity of online educational resources such as those of the Khan Academy and free online academies such as Udacity are testaments to the increasing integration of concise, instructional videos into college educational resources. We propose to use similar technology (computer, microphone, graphics tablet, video editing software) to develop, implement and assess targeted, online instructional videos that will support students in a series of large introductory geoscience courses (physical geology, oceanography, atmospheric science) at North Carolina State University and City College of San Francisco. Video-based educational resources are suited to explaining the dynamic nature of geosciences that challenge students to think both temporally and spatially. We predict that students who have access to these videos in addition to other course resources will develop a more comprehensive knowledge of important geoscience concepts than students who have access to traditional resources (e.g., textbook, lecture notes). Specifically, we predict that; 1) More higher-order thinking problems and activities can take place in the classroom; 2) Students will get more time to ask questions of each other and the instructor; 3) The instructor will get more assessment information on the student experience and mastery; and 4) Students will take more responsibility for their classroom preparation and review, thereby improving their performance in their future courses. This project seeks to develop and assess video-based resources that provide additional time-on-task opportunities for students in ways that are both pedagogically sound and discipline-appropriate. In addition, this project seeks to identify the styles of video most appealing to this media-friendly generation of students. We expect that these interventions will improve students? ability to apply geologic concepts to successfully answer higher order scientific questions. Assuming their success, we aim to create an efficient workflow for the development and sharing of these resources for a variety of geoscience courses. Video topics will be based on content identified by the literacy principles and subsequently by the appropriate course learning goals. These learning goals will be directly tied to both formative and summative assessments. Goal 1: Compare students? geoscience literacy in to standard (non-video) teaching formats. Goal 2: Identify which characteristics of video resources most appeal to students. Goal 3: Develop and share video resources with the geoscience community. Goal 4: Create professional development opportunities to demonstrate the process for creating short instructional videos.

Conceptual model diagrams are frequently used in undergraduate and graduate atmospheric science courses because they allow students to visualize the dynamic interwoven pieces of complex storm systems. These models are visual simplifications of the relationships that govern the system and the use of them in the classroom has been shown to significantly improve student learning. Despite their recognized importance, traditional conceptual models are most often static representations of dynamic systems and often struggle to reproduce the complex and noisy reality of authentic storms. Studies have shown that the educational value of conceptual models is maximized when students are able to compare and contrast the model with real data and identify weaknesses in the simplified model. This proposal seeks to develop a series of case study examples of commonly observed storm structures using observational data collected during the FRONT-PORCH experiment. High resolution dual-Doppler reconstructions of the three-dimensional wind field, dual-polarization hydrometeor identification, and surface and upper-air meteorology observations will be combined and displayed side by side to give a complete view of a real example of storm structure that can be compared and contrasted with commonly used conceptual models. These examples will be housed on the web in a case study library where students across the country can choose among different storm types and view the evolution of the storms in a movie whose playback they can control. Additionally, a series of laboratory exercises using the case study library will be written to accelerate bringing this data into classrooms and thus enhance the effectiveness of meteorology courses without significant additional effort on the part of individual instructors.

Much of the uncertainty in projections of climate change in global climate models is related to the representation of low clouds over the tropics and subtropics. Five large areas of low clouds (stratus and stratocumulus) off the western subtropical coasts of the Chile, California, Namibia, Australia, and the Canary Islands region profoundly affect the earth's radiation budget. Our proposal seeks to better understand cloud system processes in underexplored but climatologically important maritime regions and applies to the ASR goal of improving the representation of clouds and aerosol processes in climate models. Our proposal addresses specific areas of interest to the ASR program in exploring cloud life cycle and aerosol?cloud?precipitation interactions, both from an observational perspective using data from the CAP?MBL mobile facility campaign and process modeling. The proposed work will improve physical understanding of low marine clouds on temporal scales of hours to days by combining data sets obtained from the DOE ARM mobile facility deployment at CAP?MBL (May 2009 to Dec 2010) in the northeast Atlantic with data from a series of research cruises over the southeast Pacific that conducted nine ship transects over a seven year period between 2001 and 2008. We will determine the relative roles of the thermodynamic, aerosol, and synoptic environment factors on low cloud drizzle formation and lifetime by determining the joint frequency of occurrence of observed characteristics and exploring these joint frequency distributions via sensitivity studies using an LES model. Our approach is to document the joint variability of observed environmental factors and associated cloud characteristics. We will use LES modeling to test the sensitivity of cloud system properties to each of these environmental factors deemed important by the observational evidence. CAP?MBL and the VOCALS cruise in 2008 had similar sets of instruments to measure aerosol, cloud, and thermodynamic properties. We will place these measurements into an areal context using satellite data sets?particularly geosynchronous VIS/IR, MODIS products, and a new method to identify drizzle cells within liquid phase clouds based on 89-GHz AMSR-E passive microwave data. The combination of measurements at a single site with satellite data is particularly important for addressing aerosol?cloud?precipitation interactions, because drizzle cells in marine low clouds are infrequent and small, and vertically pointing cloud radar has difficulty obtaining a representative sample. Numerical simulations using a ?near?LES? framework will allow us to evaluate the mechanisms underlying the establishment of mesoscale organization and to test the primary sensitivity of cloud properties to different environmental factors such as boundary layer depth, SST, large-scale vertical motion, and aerosol concentration. We envision employing a combination of simulations using explicit microphysics, particularly where we seek detailed comparison with observations from the cloud radar, while longer-period simulations focusing on mesoscale organization and boundary layer structure will employ a computationally less expensive bulk microphysical parameterization. This research will yield information on the joint probability of occurrence of drizzle/no drizzle and steady state/transition of marine low clouds in terms of the different environmental factors as well as on the relative sensitivity of cloud properties to key environmental factors.