Radar-derived Refractivity and its Impact on Model Analysis and Forecasting using Advanced Data Assimilation Methods

PI: R. Palmer. Co-PI: M. Xue, T. Yu, J. Brotzge, S. Torres

NSF, ATM-0750790. $710K (4/2008 - 3/2011).

Project Summary

The analysis and prediction of convective-scale weather and quantitative precipitation for casting are known to be very sensitive to ﬁne-scale structures in boundary layer moisture, which can affect the exact timing and location of convective initiation. The prediction and understanding of such processes, as well as the subsequent intensity, areal coverage, and distribution of convective rainfall, are, however, seriously hindered by inaccurate and incomplete water vapor measurements. Recently, Doppler radars have been used to measure the near-surface refractivity ﬁeld using ground target echoes. These developments are exciting given that refractivity is strongly dependent on atmospheric moisture. As a result, it is possible to use refractivity as a proxy for moisture, providing a radar-based method of estimating near-surface water vapor ﬁelds with unprecedented spatial and temporal resolutions. Signiﬁcant technical challenges exist with this promising radar technique, however, including limited spatial coverage, lack of error quantiﬁcation, phase wrapping, and a cumbersome experimental procedure, for example. It is also important to note that many of these problems are exacerbated at shorter radar wavelengths, which would be relevant for any future operational network of low-cost, X-band weather radars. The overarching goal of this proposed project is a thorough and systematic study of the impact of radar-based estimates of refractivity on the prediction of convective initiation and storm evolution through the development and use of advanced signal processing and data assimilation techniques.

As part of the proposed research, the investigators will develop and implement a platform-independent refractivity engine for estimating refractivity from ground clutter signals. Extensive experiments are planned for the 2008–2010 storm seasons using seven strategically located Doppler radars in Oklahoma. In addition to these data, advanced radar simulations will be used to conduct the critical error analyses and determine the effective resolution associated with refractivity measurements. In order to assess the meteorological impact of refractivity, the investigators will develop the framework necessary for optimal assimilation of refractivity, which will be used to study the feasibility of assimilating short-term refractivity change using a cycled ensemble Kalman ﬁlter. From both simulated as well as real observations, a systematic study will be performed on the impact of assimilating refractivity on the model initialization of pre-existing storms, the forecasting of convective initiation, and subsequent storm evolution and quantitative precipitation forecast. Intellectual Merit of Proposed Activity: The scientists and engineers assembled for this project bring a wealth of experience in radar engineering, numerical weather prediction, data assimilation, observational meteorology, and signal processing. In recent years, radar-derived refractivity has emerged as an exciting technique given the possibility for high-resolution remote measurement of moisture. It is now time to systematically study the true impact of this quantity on the prediction of convective initiation and storm evolution. As an initial step, an in-depth understanding of radar signal processing is needed to transform the original experimental algorihtm into a robust, efﬁcient, operational code. Finally, new data assimilation methodologies will be developed and brought to bear to quantitatively assess the statistical signiﬁcance of refractivity. Broader Impacts Resulting From Proposed Activity: The PI has taken the lead in the development

of a new inter-disciplinary curriculum at the University of Oklahoma on weather radar and instrumentation, which has brought engineering and meteorology students together to learn the important aspects of weather radar, from fundamental theory to advanced applications. With the impetus of the proposed project, the investigative team will strive to incorporate modeling and data assimilation concepts into this successful program, further strengthening its diversity and student impact. The investigators are dedicated educators and are committed to making research accessible to undergraduate students. Therefore, active participation in the existing REU program in Norman is planned. As a role model of inspiring women in science, our team includes an early-career women meteorologist who is deeply committed to diversity and, along with the rest of the team, will use every opportunity to encourage participation of underrepresented minorities. Research and education results will be broadly disseminated through internet-based mechanisms, research publications, and presentations at professional meetings. The project will also provide much needed graduate education in advanced data assimilation and radar data applications, where trained practitioners are needed in government, universities, and the private sector.