Sovremennye problemy distantsionnogo zondirovaniya Zemli iz kosmosa, 2018, Vol. 15, No. 7, pp. 9-14
Using a variational filtering algorithm to fill in surface temperature data gaps
V.S. Kochergin
1 , S.V. Kochergin
1 , S.V. Stanichny
1 1 Marine Hydrophysical Institute RAS, Sevastopol, Russia
Accepted: 01.11.2018
DOI: 10.21046/2070-7401-2018-15-7-9-14
To restore the missing surface temperature measurement data, a filtration algorithm is considered. It is based on the variational approach, the search for the extremum of a quadratic functional, the representation of a function that approximates the measurement data as an expansion in terms of orthogonal functions. The procedure is based on a flat rotation of the original system by maximizing the quadratic functional characterizing the ratio of the coefficients of the system and its right-hand sides. In this case, a priori information is used about the possible maximum values of the decomposition coefficients. As a result of the algorithm, an equivalent system is obtained, in which the equations are arranged according to the rank of informativity. Approbation of the method on test examples and real data was carried out. As a source of data, a consistent long-term range of surface temperatures was taken for the same area in the Black Sea for thirty-five years. The measurement data were selected from the existing series randomly. The resulting gaps in the data had a duration from one to seven days. For two randomly selected years, experiments were conducted to restore the missing data. The procedure was implemented in two ways for the temperature values themselves and for deviations from the mean values. The surface temperature data were restored for the time points when the gaps in the measurement data were observed. As a result of numerical experiments, the effectiveness of the variational filtering algorithm is shown in the implementation of the procedure for recovering gaps in surface temperature measurement data.
Keywords: satellite data, surface temperature, variation algorithm
Full textReferences:
- Eremeev V. N., Kochergin V. P., Kochergin S. V., Sklyar S. N., Matematicheskoe modelirovanie gidrodinamiki glubokovodnykh basseinov (Mathematical modeling of deep-water basin hydrodynamics), Sevastopol: Ekosi-Gidrofizika, 2002, 238 p.
- Kochergin S. V., Yankovsky A. E., Primenenie algoritma fil’tratsii dannykh pri vosstanovlenii struktury zakhvachennykh voln na shel’fe Kryma (The application of the algorithm filtering data if you restore the structure of the trapped waves on the shelf of the Crimea), Morskoi gidrofizicheskii zhurnal, 1995, No. 5. pp. 62–65.
- Pukhtyar L. D., Stanichny S. V., Timchenko I. E., Optimal’naya interpolyatsiya dannykh distantsionnogo zondirovaniya morskoi poverkhnosti (Optimal interpolation of sea surface remote sensing data), Morskoi gidrofizicheskii zhurnal, 2009, No. 4, pp. 34–50.
- Strakhov V. N., Metod fil’tratsii sistem lineinykh algebraicheskikh uravnenii ― osnova dlya resheniya lineinykh zadach gravimetrii i magnitometrii (The method of filtering systems of linear algebraic equations — the basis for solving linear problems of gravimetry and magnetometry), Doklady AN SSSR, 1991, No. 3, pp. 595–599.
- Reynolds R. W., Smith T. M., Liu C., Chelton D. B., Casey K. S., Schlax M. G., Daily High-Resolution-Blended Analyses for Sea Surface Temperature, J. Climate, 2007, No. 20, pp. 5473–5496.