You are here:
Semivariogram Clouds and Plots
Citation:
McManus, M. Semivariogram Clouds and Plots. FL Fish and Wildlife Research Institute R Club, Virtual, May 11, 2021.
Impact/Purpose:
Environmental data is inherently spatial. The code in my presentation is used to read and store spatial data using open-sourced software. I then show code used for exploratory spatial data analysis, specifically examining if monitoring stations near each other have similar salinities. If such spatial autocorrelation is detected, then it can be used to make predictions of salinity, in this case, at unsampled locations.
Description:
The R code shows how to read, write, and explore spatial data using several R packages. The exploration of the data includes summarizing the distances between monitoring stations, creating a semivariogram clould to examine the relationship between all pairwise distances between stations and all pairwise salinity measurements between those stations. An empirical, or sample, semivariogram plot is created, and that plot is compared to a randomization of semivariograms as a means of detecting spatial structure in the data. Such exploratory spatial data analysis can lead to more specific modeling to make spatial predictions.