Forest ecologists commonly quantify the growth of trees by measuring their 'diameter at breast height' (135cm off the ground). To this end, many types of equipment are used, each having different historical time scales, scalability of deployment and monitoring, selection criteria, and sources of error. Such disparate measurement data can be fused into a single linear model via Hidden Markov Models (HMM), allowing for the identification of factors influencing tree growth. Furthermore, HMM's also allow for the partitioning of different sources of error, including those not of direct interest, 'one-and-done' measurement error, and errors that propagate when forecasting outcomes across time. In this presentation, I will present a preliminary HMM for forecasting tree growth and apply it to data from the ForestGEO Network's Smithsonian Conservation Biology Institute, a research center devoted to the long-term study of the world's forests in an era of rapidly changing landscapes and climate.