The data assimilation challenge of bounded, non-Gaussian, non-linear and multi-scale variables

Abstract: 

Current state estimation or data assimilation techniques assume Gaussian uncertainties for both forecasts and observations. However, unbiased observations of bounded variables can be shown to have highly non-Gaussian uncertainties and observation error standard deviations that depend on the value of the unknown true state. In particular, the observation error variance of such observations must tend to zero as the unknown true state tends to zero. Furthermore, observed bounded variables such as wind-speed, clouds, precipitation, fire and ice are highly non-linear functions of mass, momentum, and moisture variables. To simultaneously address both issues, we extend the previously developed Gamma, Inverse-Gamma and Gaussian (GIGG) variation on the Ensemble Kalman Filter (EnKF) to better account for non-linearity. Specifically, we augment the linear regression used by EnKFs to update model state variables from observed variables with a local, low-dimensional, non-linear variational step. In a synthetic Tropical Cyclone wind energy assimilation problem, the approach is shown to profoundly reduce the analysis errors associated with Tropical Cyclone wind-vector fields. The talk also summarizes some other recent improvements to the GIGG filter. In addition, we discuss challenges associated with incorporating features of the non-linear GIGG filter into global variational data assimilation frameworks such as 4DVar.

Speaker Biography:

Professor Craig Bishop was born in Melbourne and was awarded a bachelor’s degree with honours and a PhD in Applied Mathematics from Monash University. Prof Bishop’s ensemble-based data assimilation and ensemble-forecasting techniques are now used by leading environmental forecasting agencies such as the European Center for Medium Range Weather Forecasting, the UK Met Office, the German weather service, the Swiss weather service, the US National Weather Service, the US Navy and the Japanese, Korean and Brazilian Meteorological agencies. After completing his PhD, he was a post-doc at the University of Reading where he was awarded the Royal Meteorological Society’s L.F. Richardson prize for his PhD work on the dynamics of baroclinic waves in deformation fields. He then worked as a visiting scientist at the NASA Goddard space flight center where he received the Universities Space Research Association, 1994 Excellence in Scientific Research Award. This was followed by an appointment to the faculty of the Pennsylvania State University’s prestigious Department of Meteorology – then the largest atmospheric science department in the United States. There he was granted early tenure and promotion. However, to obtain a better

understanding of the operational weather prediction problem, he left Penn State for the Marine Meteorology Division of the Naval Research Laboratory in Monterey, California. There he was awarded six outstanding contribution awards, three NRL Alan Berman publication awards, and one NRL Edison patent award. He returned to Australia as Professor of Weather Prediction at the University of Melbourne in June 2018. He is a founding co-chair of the World Meteorological Organization’s Working Group on Predictability, Dynamics and Ensemble Forecasting. He is an associate editor of the Quarterly Journal of the Royal Meteorological Society. He served as chair of the Science Steering Committee of the Joint (NASA, NOAA, US Navy, US Air-Force, National Science Foundation) Center for Satellite Data Assimilation from 2007 to 2010. He was elected to the International Commission on Dynamical Meteorology in 2010 and as a Fellow of the American Meteorological Society in 2012. In 2015, he served as the PhD student elected Distinguished Visiting Scientist of the University of Reading’s internationally renowned department of meteorology. His current research mainly focusses on the data assimilation science of using models, observations and advanced estimation theory to initialise ensemble forecasts and to identify and account for systematic and stochastic aspects of model error in ensemble forecasting.