Session 9 : Spatial-temporal Models for Forest Inventory Data

The Forest Inventory and Analysis (FIA) program is tasked with monitoring forest resources in the United States. The program maintains a network of over 300,000 permanent field plots randomly distributed across the country. The plots are cyclically revisited and important forest attributes are measured. To estimate area-averages of attributes, FIA uses design-based inference, relying on the random selection of the plot locations. While random over space, the plots are not random over space-time. making time-specific estimates and change estimates difficult with design-based inference. I will discuss model-based solutions to this problem, addressing the complex nature of the data and the spatial-temporal relationships within.