Through a combination of GIS and remote sensing analysis, our goal was to depict archival data on windmill density, sugarcane production, and enslaved population as these variables relate to historic land use intensity. We utilized windmill points from various sources as a proxy for plantation presence and windmill density as a proxy for land use intensity, with the motivation being that windmills would be more clustered in areas where sugarcane processing demand was highest.
To begin developing an understanding of how windmill density might correlate with land use, we determined (1) the count and (2) the density of windmills within each estate. The spatial analysis team computed kernel density for each of the four windmill sources (Figure 21) to show windmill hotspots across the island, and where windmills were most densely located across the island. We also calculated the mean windmill kernel density of each of the estates.
This figure displays the average windmill density per acre by data source across St. Croix. The source maps mentioned (Beck, Oxholm, USGS) are displayed in the Appendix. The ground-truth windmill locations were taken from William Cleveland of the Windmills of St. Croix website.
To develop the land use intensity index, we first ensured that each layer represented sugar production, enslaved population, and windmill presence. In addition to developing the intensity index, we examined abiotic factors of St. Croix relevant to our research.