Share:


Spatial analysis and prediction of Curonian lagoon data with Gstat

    R. Garška Affiliation
    ; I. Krūminiene Affiliation

Abstract


The typical goal of geostatistical analysis is to interpolate values of variable under consideration at unobserved locations using data on observed locations because it is not feasible to gather all data of the observations in the study area. The second goal is to know how they represent the study area on the basis of the sample points. Kriging is one of geostatistical methods for spatial interpolation. This method relies on the spatial correlation reflected in the available data and so represents a global view of all the data as well as the nearest neighbor influence. Before spatial prediction using kriging can be executed, the semivariogram has to be computed and modelled.


The objective of our work is to create maps of the Curonian lagoon using kriging and cokriging methods. Our spatial data consist of observations on sounding and bed sediments of different Curonian lagoon locations. For computation and simulation of semivariograms, as well as for application kriging and cokriging methods and visualization of results on maps Gstat and PCRaster are used.


Apie Kuršių marių duomenų erdvinę analizę ir prognozavimą Gstat programos pagalba


Santrauka



Šio darbo pagrindinis tikslas‐Gstat bei PCRaster programu pagalba sukurti prognozuojamu duomenu ir ju dispersiju žemelapius. Žemelapiams sudaryti pritaikyti krigingo ir kokrigingo metodai. Krigingas yra vienas iš geostatistikos metodu, kuris atsižvelgdamas i erdvini dvieju kintamuju ryši ir kaimyniniu tašku reikšmes atlieka erdvine interpoliacija. Tuo tarpu kok‐rigingas atlieka pirminio kintamojo duomenu prognoze naudojant antriniu kintamuju duomenis. Pagrindinis geostatistines analizes tikslas yra interpoliuoti duomenis nežinomuose srities taškuose, nes dažniausiai atliekant geostatistinius tyrimus naudojami daliniai stebejimai, kurie apima tik visumos dali; arba nera žinoma, ar imties duomenys pakankamai gerai atspindi visa studijuojama sriti. Rezultatu analize parode, kad tikslesne prognoze gaunama taikant kokrigingo metoda.

Keyword : variogram, semivariogram, cross semivariogram, kriging, cokriging

How to Cite
Garška, R., & Krūminiene, I. (2004). Spatial analysis and prediction of Curonian lagoon data with Gstat. Mathematical Modelling and Analysis, 9(1), 39-50. https://doi.org/10.3846/13926292.2004.9637240
Published in Issue
Mar 31, 2004
Abstract Views
207
PDF Downloads
79