Your Name
Friend's Email Address

Funded Research Projects


Convective precipitating storms (CPSs) and the associated hazards of hail, destructive surface winds, tornadoes, and flash floods pose serious risks to life and property. These hazardous phenomena are governed by the 3D distributions of atmospheric moisture, temperaure, and wind. Simple physical arguments suggest that changes in these state variables resulting from anthropogenic increases in greenhouse gas (GHG) concentrations will in turn affect the frequency and intensity of CPSs. As we have shown in our pilot project, valuable insight about the dynamics of this local response can be gained using climate models in which convective processes are parameterized at subgrid scales. However, the limitations of this “indirect” approach highlight the importance of applying numerical models that explicitly resolve convective storms over large continental areas. In our preliminary work, we also established the basic viability of telescoping modeling strategies that consist of integrations of a convective-cloud-permitting model [the Weather Research and Forecasting (WRF) model] nested within a global model (the G-C strategy) and within a regional model that is itself nested within a global model (the G-R-C strategy). Building on the success of this pilot work, we now propose to utilize these strategies for the purpose of generating climatologies of CPSs and associated hazards over modern and future time periods. Some initial experiments using reanalyses as the global driver will be conducted, primarily to reveal model biases in the CPS statistics. The bulk of the experimentation will involve high-resolution integrations of the WRF model, driven by an ensemble of climate models forced by IPCC SRES emissions scenarios. This ensemble will help to quantify the sensitivity of the CPS dynamics to variations in the large-scale boundary conditions. Accompanying these experiments will be novel techniques to analyze the information-rich WRF model output. First, the automated, object-oriented analysis procedures developed by Baldwin et al. (2005) will be used to identify, characterize, and classify the precipitating systems. This will allow further consideration of attributes such as the areal extent of convective versus stratiform precipitation. Second, the proxy method of Trapp et al. (2007a) will be adapted to provide a quantification of local-scale hazardous weather attendant with the CPSs. For example, tornado occurrence will be estimated using the storm-scale wind field in an empirical parameter. Similar parameters for damaging wind and hail will also be developed. Finally, a powerful resampling technique known as subsampling will be employed to compute statistical characteristics and their confidence intervals from observed and modeled time series, including means, variances, skewnesses, correlations, and parameters of extreme value distributions. This project will provide an estimate of the potential response of convective precipitating storms and associated phenomena to the enhanced global radiative forcing associated with increases in GHG concentrations.



Accurate numerical prediction of high-impact weather events is required in order to allow the NWS to shift to a “warn on forecast” paradigm. Such accurate prediction of high-impact convective weather events (CWEs), such as tornadic thunderstorms and flash floods, remains one of the most important and significant challenges for operational numerical weather prediction. Unfortunately, many high-impact CWEs are considered unpredictable in a classic sense, that is, the exact location, timing, and intensity of individual storms can be extremely sensitive to perturbations in the current or previous state of the forecast modeling system and hence not reliably predicted. On the other hand, the general morphological characteristics in a CWE may be relatively less sensitive to such perturbations, and hence more predictable in a broader sense. Indeed, results from recent Storm Prediction Center/National Severe Storms Laboratory Spring Programs have demonstrated that high-resolution runs of the Weather Research and Forecasting (WRF) model are able to predict various aspects of precipitating weather systems, such as morphology or organizational mode, with some amount of skill. Similar applications have also shown a capability of the WRF model to produce kinematic storm characteristics, such as rotating updrafts, in representative numbers of storms compared to observations. High-resolution numerical weather prediction (NWP) forecasts of high-impact CWEs can exploit such capabilities, reinforced with the fact that the predominant type of high-impact weather (flash flooding, tornadoes, etc.) often relates well to the morphology of the associated convective storms. To date, assessment of these forecasts is done manually and is largely subjective. What is needed is objective, quantitative guidance on how best to use this information to improve forecasts of high-impact CWEs. To enable forecasters to make full use of high-resolution numerical models, ensembles, and the marked increase in the size of the data associated with them, automated objective techniques are needed to extract specific and relevant guidance information about the characteristics of highimpact weather events. We propose to adapt the automated object-oriented analysis procedures developed by Baldwin et al. (2005) to identify, characterize, and classify precipitating weather systems in WRF model output. We will also test and improve upon similar procedures for identification of other convective phenomena, such as supercells. NWS forecasters will be directly involved in the creation, evaluation, and implementation of these procedures and will ensure that the information provided by the forecasting system is relevant to the operational forecasting process. This COMET proposal addresses the following NWS research priorities: (1) Precipitation: Quantitative precipitation forecasts and probabilistic QPF (2) Locally hazardous weather: Severe convection, including tornadoes (3) Aviation hazards: Thunderstorms (4) Interactive Forecast Preparation System: Improve gridded forecast process.



The Principal Investigators (PIs) will conduct a multi-institution, integrated study encompassing several core science foci of the second Verification of the Origin of Rotation in Tornadoes Experiment (VORTEX2) to be held in 2009. In support of this research, and in support of the overall VORTEX2 program, the PIs will field several core VORTEX2 instrumentation systems. The field-project and analysis effort will permit the coordinated collection of multi-platform measurements and their integrated analysis. The diverse data sets will result from fixed and mobile radars, mobile mesonets, and in situ instrumentation with the goal of developing a more comprehensive scientific understanding of the genesis, maintenance, and structure of tornadoes and supercells and their relationship to the environment. This study will combine integrated analyses with those of less comprehensive data from over 130 tornadoes obtained during the Radar Observations of Thunderstorms And Tornadoes Experiment (ROTATE) program (1996-2007).

Scientific Motivation and Intellectual Merit: Current level of understanding in several areas of tornado science fails to provide answers to fundamental questions including the true range of sizes, intensities, and structures of tornadoes. The factors governing the occurrence and timing of tornado genesis, growth, intensification, maintenance, and dissipation, as well as the modes of genesis, are hypothesized from previous work, but remain largely unvalidated observationally, as do details concerning vertical wind field profiles, and the evolution and structure of damaging winds near the ground. The specific role of downdrafts in tornadogenesis and the sensitivity of the tornadogenesis process to microphysical and thermodynamic characteristics are not known. Thus, while significant advances have been made in addressing many of these questions using data from the original VORTEX and from ROTATE; these results have been limited. Specifically, VORTEX2 will provide an integration of multiple-Doppler mobile radar data with comprehensive in situ thermodynamic data that has not been achieved previously.

Broader Impacts: Improved understanding of tornado genesis, maintenance, structure, and environmental dependencies will have a broad impact on the ability to detect, warn, and forecast these severe events and to reduce subsequent casualties. The fielding of mobile Doppler radars, mobile mesonets, and in situ thermodynamic probes will be led by the PIs. These datasets will be available to the scientific community for both scientific and educational purposes and are likely to be extensively used by a large number of investigators and educators.



Weather has one of the largest direct and indirect impacts on human life around the globe, and there is an increasing demand for professionals who can serve as authorities on weather and climate phenomena and interpret associated data. However, the typical education of atmospheric scientists at the undergraduate level results in little understanding of how atmospheric science research is performed, or the evolutionary nature of its results. Thus, graduates of such programs fail to appreciate the strengths and limitations of the current scientific information, and thereby fail to communicate the uncertainty of this information to each other, policymakers, the media, and the general public.

In the proposed work, an established, successful model of incorporating research into chemistry laboratory courses (from the Center for Authentic Science Practice in Education, “CASPiE”) will be adapted to create a new atmospheric science laboratory course, using data and numerical models readily available, for students in their sophomore year. The proposed work will investigate if a research-based learning experience in the early stages of an atmospheric science curriculum can enhance the students’ scientific understanding and the ability to communicate science. With the assistance of personnel at Purdue University that developed the CASPiE model, three new research modules, aiding the research of three different faculty members, will be developed: forecast verification, potential influences of regional climate change on precipitation, and thunderstorm identification and geospatial distribution. The proposed modules are designed specifically for the depth of understanding that the students can have at this early stage in their coursework, and yet challenge them to learn to use state-of-the-art data analysis software and numerical modeling necessary to perform novel research in these areas. Each module allows the students to work in teams to analyze preliminary results, use such results to revise the experimental design to conduct further inquiry, and collectively summarize the conclusions of the research and present final results both in oral and poster formats. The modules can be used in multiple years and shared with other institutions. Rubrics for course grading will be adapted from those already developed for the CASPiE model, as will CASPiE instruments for assessment of the educational outcomes of the new laboratory course.

The intellectual merit of the proposed work is multifold. From a science education perspective, it is an investigation of: (i) how readily a model for learning developed in one discipline can be adapted to a very different discipline; (ii) whether a research-based learning experience of this style can enhance students’ scientific understanding and the ability to communicate science; and (iii) and if research-based learning experiences early in an undergraduate program can aid in retaining students in the major. From an atmospheric science perspective, the proposed research modules used in the laboratory course will provide data for evaluating and improving forecast verification algorithms, numerical modeling results to begin to address potential precipitation sensitivity to regional climate change, and case studies to develop data mining techniques for geospatial classification of thunderstorms.

The broader impacts of the proposed work are also multifold. The proposed work will provide experience in adapting a successful model of a research-based course to another science discipline, so that such a model may be used in a wide variety of disciplines, or adapted in the future to upper-level laboratory courses in the same discipline to extend the research experience gained by the students. Dissemination of the educational outcomes from the proposed work, and even the modules themselves, will allow other atmospheric science programs to enhance the educational experience of their own undergraduates. The research results acquired through the use of the proposed modules themselves will provide valuable feedback for the atmospheric science community in improving numerical weather prediction models (with obvious societal benefit), will provide information to agriculture and government regarding the regions where precipitation in the U.S. may be highly susceptible to regional climate change, and will provide initial data on the geospatial frequency of thunderstorms which ultimately will be used to improve climate model predictions of weather extremes.



The bow echo is well regarded as a prolific producer of severe, though primarily nontornadic, wind damage at the ground. The generally accepted paradigm for such damaging wind production involves a strong surface cold pool and associated rear-inflow jet (RIJ) that descends to the ground behind the apex of the bow. Recent idealized numerical simulations and limited observations are suggesting an additional paradigm, in which damaging surface winds are also induced tens of kilometers north of the apex by low-level, meso-gamma scale (order 10 km diameter) vertical vortices. The simulations in particular suggest that under certain environmental conditions, such mesovortex winds can be stronger, have longer durations, and produce a larger area of significant damage than can the RIJ winds at the apex. To date, reliable and sufficient numbers of observations to validate this new paradigm have yet to be collected. Further, the generality of this damaging-wind producing mechanism (as well as of generality of the mesovortexgenesis mechanism itself) has yet to be established in instances when the mesoscale environment of the bow echo is significantly inhomogeneous. Underscoring the importance of this are recent observational studies that indicate a relation between reports of damaging straight-line (and/or tornadic) winds and the interaction of a bow echo and mutually perpendicular “external” outflow boundary. The primary objectives of this proposal, therefore, are to: (i) conduct highly detailed aerial and ground surveys of wind damage following bow echo events, relating the severity and scale of damage to radar-observed convective system location and structural characteristics; and (ii) perform analyses and realistic mesoscale model simulations of bow echo events in order to investigate dependencies of convective system-relative location of damaging winds, mesovortexgenesis, and convective system structure on the inhomogeneous mesoscale environment, which may consist of external boundaries, a locally stable planetary boundary layer, etc. Special, high-resolution observations will be assimilated into the numerical model to provide this mesoscale environment. The bow echo and MCV experiment (BAMEX) will provide the framework for the proposed work. BAMEX seeks to understand and improve prediction of: the meso- and cellular-scale processes, within bow-shaped mesoconvective systems, that lead to damaging winds at the ground; and MCVs and the deep cumulus convection they often trigger. BAMEX is currently scheduled for 20 May-6 July 2003, will be conducted over a large experimental domain centered on St. Louis, Missouri, and will involve unprecedented data collection via specialized airborne and ground-based observing platforms. Post-event damage surveys, a critical component of the BAMEX dataset, will be used to meet numerous BAMEX objectives besides that of the proposed work. The results of the proposed study will significantly clarify the understanding of damaging wind production in bow echoes, and in particular, will illuminate where (with respect to convective system-relative location) the most damaging winds are most likely to occur, with what radar-observable attributes, and under what mesoscale environmental conditions. Ultimately, the results of the study will be applied by operational forecasters to issue more timely and accurate forecasts and warnings of damaging nontornadic surface winds.


Feedback | E-mail Webmaster

Purdue University, West Lafayette, IN 47907
(765) 496-6661, E-mail:
© 2010 Purdue University | An equal access/equal opportunity university