The Southern Annual Forest Inventory System

The Southern Annual Forest Inventory System (SAFIS) is in various stages of implementation in 7 of the 13 southern states serviced by the Southern Research Station. The SAFIS design is an interpenetrating design where the n units (1/6 acre plots) are divided into k = 5 panels, each panel containing m = n/k units. Panel 1 plots are measured in year 1, panel 2 in year 2, etc., such that all plots have been visited by the end of year 5. The panel cycle is repeated into perpetuity. Each panel, in effect, is a 5-year periodic survey with complete overlap of sample units. Numerous estimation schemes are possible, and we explore five possible options. The five options are (1) use existing periodic inventory programs to produce 5-year survey estimates by adjusting all five panels to a common year, (2) analyze each annual panel independently, (3) produce 5-year estimates by combining the five panel estimates by varying the weight given to each panel, (4) base inventory estimates on mixed estimation where actual and predicted values are combined, and (5) use imputation techniques such that unmeasured plots are filled in with imputed plots. A two-phase method for forest area estimation that uses the known map marginals from a thematic map is presented as an alternative to photo interpretation-based estimates.

[1]  W. Deming.  On Simplifications of Sampling Design Through Replication with Equal Probabilities and without Stages , 1956 .

[2]  Andrew P. Robinson,et al.  Development and testing of regeneration imputation models for forests in Minnesota , 1997 .

[3]  P. C. Mahalanobis,et al.  Experiments in statistical sampling in the Indian Statistical Institute , 1961 .

[4]  Howard L. Jones,et al.  Investigating the Properties of a Sample Mean by Employing Random Subsample Means , 1956 .

[5]  A. Winsor.  Sampling techniques. , 2000, Nursing times.

[6]  P. Vandeusen.  Correcting bias in change estimates from thematic maps. , 1994 .

[7]  Innis G. Sande,et al.  Imputation in Surveys: Coping with Reality , 1982 .

[8]  Charles T. Scott,et al.  Sampling with partial replacement and stratification , 1994 .

[9]  P. Deusen,et al.  Multiple-occasion partial replacement sampling for growth components , 1989 .

[10]  P. Deusen,et al.  Annual forest inventory statistical concepts with emphasis on multiple imputation , 1997 .

[11]  D. H. Card.  Using known map category marginal frequencies to improve estimates of thematic map accuracy , 1982 .

[12]  Raymond L. Czaplewski,et al.  Calibration of Remotely Sensed Proportion or Area Estimates for Misclassification Error , 1992 .

[13]  T. Başar,et al.  A New Approach to Linear Filtering and Prediction Problems , 2001 .

[14]  P. Deusen.  Incorporating predictions into an annual forest inventory , 1996 .

[15]  P. Mahalanobis,et al.  Experiments in Statistical Sampling in the Indian Statistical Institute. , 1962 .

[16]  Gary J. Brand,et al.  A description of STEMS-- the stand and tree evaluation and modeling system. , 1982 .

[17]  H. Theil.  Principles of econometrics , 1971 .

[18]  Paul C. Van Deusen,et al.  Unbiased estimates of class proportions from thematic maps , 1996 .

[19]  P. G. D. Vries,et al.  Continuous Forest Inventory with Partial Replacement of Sample Plots , 1986 .