Konferenzbeitrag
Spatial zoning of precipitation by geo-statistical methods and DEM as auxiliary variable in northeast of Iran
Vorschaubild nicht verfügbar
Volltext URI
Dokumententyp
Text/Conference Paper
Zusatzinformation
Datum
2011
Autor:innen
Zeitschriftentitel
ISSN der Zeitschrift
Bandtitel
Verlag
Shaker Verlag
Zusammenfassung
One of the most important factors in environmental studies is precipitation and its spatial variability. Precipitation data is collected in some individual points in different stations, while these point based data are not often completed in a long term period. In arid and semi-arid area data incompleteness is more serious. There are different methods for reconstruction and zoning of precipitation data, however, it is essential to choose a reliable method that could matches the regional requirements and also provides adequate precision for our applications. Geographical position and topographical factors strongly influence spatial pattern of precipitation. Multivariate geo-statistical method is used to estimate the spatial correlation of two variables that are interdependent in a physical sense. In this paper Ordinary Kriging (OK) as a univariate method and Ordinary CoKriging (OCK) as multivariate method were used for interpolation precipitation. The purpose of this study was to show how the use of a digital elevation model can improve interpolation processes for mapping the mean annual and monthly precipitation amount in the North East of Iran. As most of the study area is mountainous, Digital Elevation Model (DEM) was used as auxiliary variable in the models. Cross validation is used to compare the prediction performances of the geo-statistical interpolation algorithms. In this study Mean Squared Error (MSE) values were used for comparing the models. Cross-validation suggested that OK, which ignores the information on elevation, has lager prediction errors in comparison with prediction errors produced by OCK specially in months with low precipitation. For example MSE values were 0.47, 0.3, 0.25 and 0.5 for March, October, November and December respectively when OCK was applied, while these values were 0.17, 1.8, 1.2 and 1.3 respectively for the same months when OK was applied. Overall, the smallest value of MSE was 0.25 in November when OCK was applied which had the highest correlation coefficient as well. It could be concluded that using of DEM as auxiliary variable improves the results, because of high variability in topography in the study area. The results seem to favor the multivariate geo-statistical method including auxiliary information (related to elevation). We conclude that OCK is a very reliable and robust interpolation method because it can take into account several properties of the landscape; it should therefore be applicable in other mountainous regions, especially where precipitation is an important geomorphological factor.