Figure 5
From: Reconstructing the historical expansion of industrial swine production from Landsat imagery
Examples of limitations to the changepoint detection algorithm. In some cases, an insufficient number of high-quality Landsat 5 images were available to capture the year of construction of an individual swine waste lagoon (a), resulting in errors of + /− 1 year. In other cases, the changepoint algorithms detected the start of the construction of the swine waste lagoon but the swine waste lagoon was not fully operational until later years due to prolonged construction timelines (b).
