Naslov
Combined use of Multiple Linear Regression, Optimal Interpolation and GIS in producing temperature maps

Autori
Perčec Tadić, Melita ; Pandžić, Krešo

Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, sažetak, znanstveni

Izvornik
Abstract book of the XXVII EGS General Assembly ; u: Geophysical Research Abstracts 4 (2002) / - : European Geosciences Union (EGU), 2002

Skup
EGS General Assembly (27 ; 2002)

Mjesto i datum
Nica, Francuska, 21.04.2002. - 26.04.2002

Vrsta sudjelovanja
Poster

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
air temperature; multiple linear regression; optimal interpolation; GIS

Sažetak
In intention to produce climate maps of monthly and annual average air temperature for the CLINO period 1961-1990 the data from the meteorological stations are interpolated to regular grid. Combined use of multiple linear regression method and optimal interpolation was introduced. For every grid point, the temperature anomaly from the corresponding average is assumed in the form of linear combination of anomalies from the average on the meteorological stations. Weighting factors for every grid point are unique and calculated from the demand that the mean square difference between grid point temperature anomaly and suggested linear combination is minimal. An interpolation technique using these weighting factors is optimal interpolation and is formulated in details by L. S. Gandin. The obtained anomaly field is then added to average temperature field evaluated on regular grid using the regression equation to provide the final grided field. Multiple linear regression with measured temperature as predicted and stations altitude and geographic coordinates as predictors is used. Data for independent predictor variables are elaborated from 1 km resolution digital elevation model (DEM). GIS is used for calculation of average temperature field and for visualisation and producing final maps.

Izvorni jezik
Engleski

Znanstvena područja
Geologija

POVEZANOST RADA