Spatial-Temporal Cluster Analysis to Identify Emerging Agglomeration of Texas Wineries, 1973-2014
MetadataShow full metadata
Wine is of interest to geographers for a variety of reasons. To fully understand the geography of wine, one must consider many factors. For example, the geology, biology, climate, culture, economics, and politics of a particular region influence the wine produced there. In Texas, wine production dates back more than 350 years. However, only within the past few decades has the wine industry in Texas grown significantly. This paper has two goals: 1) an examination of the history of the Texas wine industry, and 2) a spatial-temporal cluster analysis to determine emerging patterns of agglomeration of wine production in Texas. Understanding the Texas wine history and identifying these patterns establishes a baseline that will be useful for future study which examines the factors driving growth and development patterns of the Texas wine industry from a geographical perspective.
To conduct the spatial-temporal analysis, I used ArcGIS 10.2.1. I created histograms and decadal snapshots to show change over time. I then created standard deviational ellipses using the decadal snapshots to examine emerging trends. Prior to processing the data with the mapping clusters tools, I used the Incremental Spatial Autocorrelation to determine that there is one statistically significant peak distance, at 198550 meters. I also created a spatial weights matrix to be able to include temporal factors.
I used all three of the mapping cluster tools (Grouping Analysis, Cluster and Outlier Analysis, and Hot Spot Analysis) to examine spatial clustering and, using the spatial weights matrix, spatial-temporal clustering. The Grouping Analysis divided along spatial lines and when time was added as a factor, separated out the early wineries and then along the spatial lines. The Cluster and Outlier Analysis also divided out similarly without time as a factor. When time was added the clustering and outliers were no longer in distinct spatial divisions. The Hot Spot Analysis gave similar results to the Cluster and Outlier Analysis tool, with and without the temporal weighting.
I created another spatial weights file, only using the wineries established after 2003. The spatial results appeared similar, with the exception that almost all of the clusters identified are now shown to have a 99% confidence level.
This study identified provides a solid baseline for further research into the Texas wine industry. It establishes the historical context of the industry and identifies statistically significant emerging agglomeration of wineries in Texas. This information can be used as the basis to study why this agglomeration is occurring.