Date

27-4-2016 12:00 AM

Major

Meteorology

Department

Geological and Atmospheric Sciences

College

College of Liberal Arts and Sciences

Project Advisor

Bradley Miller

Project Advisor's Department

Agronomy

Description

The difference between simple kriging (SK) and ordinary kriging (OK) is the reliance on the assumption of stationarity. This study tested the importance of this assumption for creating accurate, continuous maps of Iowa precipitation for 2015. Spatial interpolation is a process in which unobserved locations in a particular geographic region are estimated by using the observed points nearby. Stationarity is defined by the mean and distribution of the data remaining constant. SK relies on this assumption while OK does not. In this study, these predictive tools estimated annual precipitation for unobserved locations across the state from a limited number of observed locations. The inability to measure a variable at all geographic locations is a common problem in the natural sciences. Precipitation maps were produced in ArcMap using SK and OK. In addition, a difference map was calculated to quantitatively compare the estimations made by the two interpolation methods across the study area. Although the difference map does highlight differences in the precipitation patterns estimated by SK and OK, the deviation proves to be minor for the sum of the entire 2015 year. Therefore, the assumption of stationarity had only a slight influence on the spatial interpolation of 2015 Iowa precipitation.

File Format

application/pdf

Included in

Meteorology Commons

Share

COinS
 
Apr 27th, 12:00 AM

Comparing Simple and Ordinary Kriging Methods For 2015 Iowa Precipitation

The difference between simple kriging (SK) and ordinary kriging (OK) is the reliance on the assumption of stationarity. This study tested the importance of this assumption for creating accurate, continuous maps of Iowa precipitation for 2015. Spatial interpolation is a process in which unobserved locations in a particular geographic region are estimated by using the observed points nearby. Stationarity is defined by the mean and distribution of the data remaining constant. SK relies on this assumption while OK does not. In this study, these predictive tools estimated annual precipitation for unobserved locations across the state from a limited number of observed locations. The inability to measure a variable at all geographic locations is a common problem in the natural sciences. Precipitation maps were produced in ArcMap using SK and OK. In addition, a difference map was calculated to quantitatively compare the estimations made by the two interpolation methods across the study area. Although the difference map does highlight differences in the precipitation patterns estimated by SK and OK, the deviation proves to be minor for the sum of the entire 2015 year. Therefore, the assumption of stationarity had only a slight influence on the spatial interpolation of 2015 Iowa precipitation.