Dr John Taylor

Lecturer
BSc PhD ANU, MAIP, MAGU

Telephone: +61 2 62688813
Fax: +61 2 6268 8786
Email: j.taylor@adfa.edu.au
Location: PEMS Sth, Room G27

Research Interests:

Laboratory investigations of mixing processes in density stratified fluids with application to the ocean and atmosphere. Development of a high frequency acoustic doppler radar for studying the surface layer of the atmospheric boundary layer.

Lower Atmosphere Research - Monitoring the atmospheric boundary layer

Web: www.pems.adfa.edu.au/~jtaylor

Current Research

Monitoring the atmospheric boundary layer (Dr John Taylor)

The development of ground-based systems for acoustic and electromagnetic remote sensing of the atmospheric boundary layer and the utilization of these systems for understanding the mean and turbulent structure of the atmospheric boundary layer. A major focus is on the interaction between the boundary layer and mesoscale meteorological phenomena such as fronts, local wind systems and gravity waves. These problems are being approached in observational and modelling studies. Current projects include:

estimation of uncertainty in of algorithms for extracting wind and turbulence information from sodar and electromagnetic wind profilers (with Dr B. Burns & Dr D. Low).
Measuring boundary layer convergence and its importance for the initiation of convection in the tropical atmosphere (with Dr D. Low & Prof. Michael Reeder) .
Acoustic wind profiling during the passage of ?Morning glory? internal gravity waves (with Dr David Low & Prof. Michael Reeder) .
evelopment of a cwRASS (Radio Acoustic Sounding System) for boundary layer temperature profiling and its application to forecasting fog clearance times (honours projects, with M. Bradshaw in 2007 and J. Moogan in 2008).

Recent Research Projects

Data quality control in atmospheric remote sensing systems: When do you believe the results? (Faculty Research Grant Program 2006) [Chief Investigators: Dr J. R. Taylor & Dr D. J. Low, Research Personnel: Dr B. A. Burns]

Outline

The work performed under this FRG consisted of three tasks:

quantify the retrieval errors as part of the data analysis for radar profiler systems.
implement quality controls that use this error vector and other ?goodness? measures to remove spurious signals.
output error and quality information for future use in multisensor data integration. Although our FRG proposal included treatment of the SODAR system, time did not permit this part of the task to be carried out. However, most of the development and testing that has been carried out on the radar profiler system is directly transferable to the SODAR system.

Summary and future work

A new algorithm has been developed for spectral parameter estimation that operates directly on measured intensity (rather than log intensity) and is therefore insensitive to the noise level. Using a linear combination of Gaussians for the fitting function to the entire spectrum, it also accommodates outliers and secondary peaks without compromising the fit to the primary signal. In future, application of this algorithm to removal of the clutter peak and to cases where more than one signal peak are present (e.g. clear sky and precipitation) will be examined.

An error estimation routine was developed for this algorithm based on the methodology of Rodgers (2000). In this initial implementation, it provides a generalized mechanism for translating radar measurement error into errors in the estimated spectral parameters. With the software now in place, errors in the algorithm itself ? specifically errors in the form of the fitting function and its nonlinearity ? can easily be incorporated into this translation process.

The uncertainty in each parameter is output as part of an extended output vector for each individual retrieval, and carried through (with appropriate mathematical treatment) subsequent averaging and quality control steps in the processing. The end product is then a set of retrieved parameters (mean wind vector, turbulence parameters) representative of the atmospheric state and tagged with an error vector that can be used as weights when data are combined into composite atmospheric profiles from our integrated profiling system.

Implementation of a quality flag matrix proved immediately useful in processing control and in judging the relative robustness of the various retrieval algorithms. Along with the average RMS errors, this matrix will provide additional information on the output products either for combination with those from other sensors into an integrated product or for assimilation into regional weather prediction models.

As per the original goals of this project, the next step is the application of the processing system to data from the SODAR system. It is anticipated that several modifications will be required including the characterization of the measurement error. The multiple Gaussian algorithm is expected to be particularly useful for treating non-atmospheric noise signals in these data.

References

Rodgers, C. D., 2000, Inverse Methods for Atmospheric Sounding , Series on Atmospheric, Oceanic and Planetary Physics, 2, World Scientific, Singapore , 238 pp.

Recent Publications

In press

Burns, B. , Taylor, J. , Low, D., Retrieval errors in spectral parameters estimated by gaussian fitting, submitted to IEEE Transactions on Geoscience and Remote Sensing .

Conference proceedings

Burns, B.A., Davis, C., Kiss, A., & Taylor, J.R., (eds) 2010, 'IOP Conference Series: Earth and Environmental Sciences', 17th National Conference of the Australian Meteorological and Oceanographic Society, Canberra, ACT, Australia, 27-29 January 2010.

Conference papers

Burns, B.A., Taylor, J.R., & Sidhu, H., 2010, Uncertainties in bathymetric retrievals, IOP Conf. Series: Earth and Environmental Science, 11, 012032, doi:10.1088/1755-1315/11/1/012032, 17th National Conference of the Australian Meteorological and Oceanographic Society, Canberra, ACT, Australia, 27-29 January 2010.
Taylor, J.R., & Moogan, J.C., 2010, Determination of visual range during fog and mist using digital camera images, IOP Conf. Series: Earth and Environmental Science, 11, 012012, doi:10.1088/1755-1315/11/1/012012, 17th National Conference of the Australian Meteorological and Oceanographic Society, Canberra, ACT, Australia, 27-29 January 2010.
Taylor, J. R. , Low, D. J. & Reeder, M. J., 2007, Sodar observations of organized flows in tropical northern Australia , 14th National Conference of the Australian Meteorological and Oceanographic Society , Adelaide , 5-8 February 2007.
Burns, B. A. , Taylor, J. R. & Low, D. J. , 2006, Retrieval errors in spectral parameters estimated by gaussian fitting, 13th National Conference of the Australian Meteorological and Oceanographic Society , Newcastle, 6-8 February 2006.
Burns, B. A. , Taylor, J. R. & Low, D. J. , 2006, Retrieval errors in spectral parameters estimated by gaussian fitting, in Climate, Water and Sustainability , eds Howard Bridgman, Newcastle, NSW, Melbourne, Vic, p. 17.
Taylor, J. , Jovanovich, M. & Low, D. , 2006, Measuring turbulence with doppler sodars, 13th National Conference of the Australian Meteorological and Oceanographic Society , Newcastle , 6-8 February 2006.
Taylor, J. R. , Jovanovich, M. & Low, D. J. , 2006, Sodar spectra at low levels in the stable nocturnal boundary layer, 13th International Symposium for the Advancement of Boundary Layer Remote Sensing , Garmisch-Partenkirchen, Germany, 18-20 July 2006.
Taylor, J. R. , Low, D. J. & Reeder, M. J., 2006, Sodar diagnosis of mesoscale boundary layer convergence, 13th International Symposium for the Advancement of Boundary Layer Remote Sensing , Garmisch-Partenkirchen, Germany, 18-20 July 2006.

Recent Grants

External Grants

J. Taylor , H. Sidhu, Investigation of uncertainties in bathymetric retrievals using a semi-analytic method, Defence-Related Research Funding Scheme Grants 2007: $18,627.

UNSW Grants

J. Taylor, & D. Low , Is it possible to remotely sense the turbulent heat flux in the atmospheric boundary layer?, Special Research Grant, 2006: $5,000.

J. Taylor, D. Low & B. Burns , Data quality control in an atmospheric remote sensing system: when do you believe the results?, Faculty Research Grants Program, 2006: $17,167.