Monitoring land/forest cover using the Kalman filter: A proposal

Abstract

Although growth and yield models have been used to update forest inventories for large regions, such models poorly predict cover changes from land use conversions, regeneration, and harvest. These changes could be monitored directly for large areas using remote sensing, which can be expensive, or estimates made by agricultural agencies, which are not detailed for condition of timberlands. The Kalman filter, which is a flexible statistical estimator, might increase statistical efficiency and produce annual estimates of cover by combining such direct monitoring with past knowledge (i.e., previous forest inventory) and expected change (i.e., model for annual change in land cover). This paper describes a potentially useful application of this topical estimator and presents specific proposals for practical methods.

Parent Publication

• Forest growth modeling and prediction (Volumes 1 & 2).

Citation

Czaplewski, Raymond L.; Alig, Ralph J.; Cost, Noel D. 1988. Monitoring land/forest cover using the Kalman filter: A proposal. In: Ek, Alan R.; Shifley, Stephen R.; Burk, Thomas E. Forest growth modelling and prediction: Volume 2. Gen. Tech. Report NC-120. St. Paul, MN: U.S. Dept. of Agriculture, Forest Service, North Central Forest Experiment Station. p. 1089-1096.