Bias estimation of historic in situ upper air data

Principal Investigator

Dr. Leopold Haimberger
Institute for Meteorology and Geophysics,
University of Vienna
Althanstrasse 14,
A-1090 Wien, Austria

leopold.haimberger@univie.ac.at

Project description

In situ upper air data, mostly radiosonde and tracked balloon data, are available over the northern hemisphere from the 1930s onwards and global coverage is given from the International Geophysical Year 1958 onwards. To exploit their full potential it is essential (i) to remove spurious biases and shifts from the station records and (ii) to assimilate the data with proper dynamic data assimilation systems in the framework of reanalyses. The task of removing temporal changes in observation biases is referred to as homogenization.

Within previous special projects and with financial support from the EC and the Austrian science funds, the PI has developed a homogenization method (RAdiosonde Observation COrrection using REanalyses - RAOBCORE) for radiosonde temperature and wind data. It works with time series of departures between observed temperatures/winds and background forecasts used in the atmospheric data assimilation. The calculated homogeneity adjustments are used for bias correction in the ongoing ERA-Interim reanalysis. Descriptions of RAOBCORE and related methods can be found in Haimberger (2007), Haimberger et al. (2008), Gruber and Haimberger (2008) and Santer et al. (2008).

While it is encouraging that RAOBCORE adjustments are used as input in ERA-Interims, they need to be improved and substantially extended before they can meet the needs of reanalyses in the years to come. The proposed new project has two major aims: (i) the extension of homogeneity adjustments of radiosonde data back to the 1930s, (ii) implementation and testing of ''online bias estimation'' methods for radiosonde data during the data assimilation process.

Aim (i) should be feasible with the increasing number of digitized upper air data together with background departure time series from assimilation runs over this early period. These could come either from a surface observation only reanalysis of the 20th century as provided by the University of Colorado at Boulder (UoC) or could be generated at ECMWF with low resolution pilot assimilations. The homogenization of the post-1958 period shall be improved as well utilizing improved reference series such as ERA-Interim background forecasts and calibrated radiance data.

Aim (ii) is the implementation of methods which estimate observation biases together with the atmospheric state vector without the need for externally specified bias correction tables as provided by RAOBCORE. Such methods work well for satellite radiances but have not yet been applied to conventional upper air data. If these could be used for radiosonde data, at least from 1979 onwards, it would substantially facilitate the reanalysis process.

In the proposed project, it is planned to pursue both goals - homogenization and online bias estimation. While it is hoped that online bias estimation will work also for radiosonde data until the next reanalysis, this cannot be guaranteed and at least needs to be carefully tested. The improved homogenized dataset will provide comprehensive test data about existing shifts in observed radiosonde records.

The requested computer time will be needed for assimilation experiments with the ERA-Interim data assimilation system. In the assimilation experiments it will be tried to ingest early upper air data from the 1930s onwards. Background departure information from these experiments will help in bias assessment and homogenization of the early data. It will also be tried to develop a statistical model for radiosonde temperature and wind biases, whose parameters can be estimated with the already existing variational bias correction technique.

In the ideal case, the special project will be supported by the Austrian science funds (proposal submitted) and will profit from a 7FP project led by ECMWF, which will be devoted to a reanalysis of an extended period (1938-).

References:

Gruber, C., and L. Haimberger, 2008: On the homogeneity of radiosonde wind time series. Meteorol. Z. in press.

Haimberger, L., 2007b: Homogenization of radiosonde temperature time series using innovation statistics. J. Climate, 20, 1377–1403.

Haimberger, L., C. Tavolato, and S. Sperka, 2008: Towards elimination of the warm bias in historic radiosonde tem-perature records—some new results from a comprehensive intercomparison of upper air data. J. Climate 21. in press.

Santer, B., P. Thorne, L. Haimberger, K. Taylor, T.Wigley, J. Lanzante, S. Solomon,M. Free, P. Gleckler, P. Jones,T. Karl, S. Klein, C.Mears, G. Schmidt, D. Seidel, S. Sherwood, and F.Wentz, 2008: Consistency of modelled and observed temperature trends in the tropical troposphere. Int. J. Climatol., 28. submitted.

Additional information

New Project for 2009

Would accept support for 1 year only, if necessary.