[ View Thread ] [ Post Response ] [ Return to Index ] [ Read Prev Msg ] [ Read Next Msg ]

RUC20 Evaluation Discussion Board

Re: help on analysis file

Posted By: Stan Benjamin
Date: 30-October-02 2046Z

In Response To: help on analysis file (Mike Ross)

Mike,

The analysis file contains an estimate of the current conditions based on both current observations and a previous forecast (1h, in the case of the RUC) that itself contains information from previous observations.

So, this is an estimation problem. We don't have observations at every grid point, and even if we did, the observations themselves have errors. So, the process of 'data assimilation' produces that gridded analysis that you are looking at. It takes into account the expected error of both the observations and the forecast background field. It also takes into account the expected spatial correlations in forecast error fields, based on previous statistics. It ends being a sophisticated interpolation or minimization problem taking into account all of these and other factors to produce the 'analysis grids'.

There is more information on data assimilation at this web site:

http://meted.ucar.edu/nwp/pcu1/ic6/frameset.htm

So, the short answer is that the analysis file is based on observations but should not be considered equivalent to the observations themselves.

Stan

Messages In This Thread

help on analysis file -- Mike Ross -- 30-October-02 2046Z
Re: help on analysis file -- Stan Benjamin -- 30-October-02 2046Z

[ View Thread ] [ Post Response ] [ Return to Index ] [ Read Prev Msg ] [ Read Next Msg ]

RUC20 Evaluation Discussion Board is maintained by RUC/MAPS Forum Administrators with WebBBS 5.12.