|
Article information
2008 , Volume 13, Special issue, p.117-123
Nogoykov I.V., Chavro A.I., Dmitriev Y.V.
On the stability of statistical downscaling of regional extreme temperature fields
This paper addresses the problems related to the stability of statistical methods for solving the downscaling problems for the regional fields of various meteorological parameters such as surface temperature, pressure, wind etc. The statistical model, which allows reconstructing more than 75 % of the natural variability of temperature extremes at meteorological stations in Moscow is suggested. In our opinion, there are some interesting difficulties, which are related to the inhomogeneity of data, which establishes a priori statistical information. It is demonstrated that even a little amount of missing data could cause an instability of the solution and reduce its accuracy. In this connection, some frequently used methods of spatial interpolation for reconstruction of the missing values of meteorological parameters are compared. We propose a combined method for filling the missed values, which allows substantial improvement of stability for solutions of the inverse problems.
[full text] Author(s):
Nogoykov I.V.
Office: Institute of Numerical Mathematics RAS
Address: 119991, Russia, Moscow, ul. Gubkina, 8
Phone Office: (095) 938 39 17
Chavro Anatoly Ivanovich
Dr. , Professor
Position: Leading research officer
Office: Institute of Numerical Mathematics RAS
Address: 119991, Russia, Moscow, ul. Gubkina, 8
Phone Office: (095) 938 39 17
E-mail: chavro@inm.ras.ru
Dmitriev Yegor Vladimirovich
PhD.
Position: Research Scientist
Office: Institute of Numerical Mathematics RAS
Address: 119991, Russia, Moscow, ul. Gubkina, 8.
Phone Office: (095) 938 37 72
E-mail: yegor@inm.ras.ru
Bibliography link:
Nogoykov I.V., Chavro A.I., Dmitriev Y.V. On the stability of statistical downscaling of regional extreme temperature fields // Computational technologies. 2008. V. 13. Special issue 3. P. 117-123
|
