GIS 3015 - Module 5 - Spatial Statistics
In Module 5 we were provided the opportunity to take an ESRI virtual training course "Exploring Spatial Patterns in Your Data Using ArcGIS". Through this online course we learned how to visually examine our data as well as perform an in-depth analysis. First we visually looked at our map data, looking for higher concentrations and lower concentrations of weather station locations across Western Europe. We also looked for absences of weather stations. Next we used the "mean center" tool to determine the "average" location of a weather station. The "median center" tool was used to determine the "middle" of the ordered station location data set. The "directional distribution" tool created a standard deviational ellipse showing the orientation of the data set (east-west). All of these tools created a new layer in our map enabling us to change the symbology of each location.
My map shows weather station locations throughout Western Europe. I have denoted these stations with a red symbol. The mean center (green circle) and median center (yellow cross) are also displayed. Because the mean and median are similarly located that indicates our data is symmetric and has very little skewness. A purple ellipse shows the directional distribution of the data set (east west).
In the second part of the training, we were able to evaluate if our data was normally distributed by creating and viewing a histogram and Normal QQ Plot. First we viewed statistics of the temperature field in the attribute table. To create our histogram and Normal QQ Plot we used the tools found under the Geostatistical Analyst Menu.
We continued to analyze our data set by exploring more tools under this menu such as Voroni Map, Semivariogram, and Trend Analysis. All of these tools enabled us to further explore our data set, both the locations and the temperature values collected at each station. We were able to determine that our data set was normally distributed and we identified an outlier weather station location in La Fretaz, France. After thoroughly examining our data we can now select the correct analysis tool to aid in making a more accurate predication of locations where freeze warnings are needed. The geostatistical tools led us to select geostatistical interpolation as the proper analysis technique.
No comments:
Post a Comment