Fig. 5. Correlogram for Day 11. The near-zero values of Moran's I in the residuals indicate that the spatial components are mainly captured by the 20 filters. Only at short distances (<500 km) does there seem to be a missing filter. However, the oscillation of Moran's I in the response variable (percentage dead plants) is an unusual pattern and reflects alternating similar and dissimilar patches. This graph is similar on all other days.