Procedures Used To Produce Second Approximation SIBEC Estimates

The provincial SIBEC database was initially compiled in 1994-5 (Thrower et al. 1995), and the first approximation estimates of site index by site unit were completed and published in 1997 (B.C. Ministry of Forests 1997). At the time these estimates were released, they were considered the best available for the Biogeoclimatic Ecosystem Classification (BEC) site series and species combinations. At that time, the model developers realized that the accuracy and precision of these estimates would improve over time through a process of continued data acquisition and refinements to the analysis tools. The following points describe these refinements and the procedures used to generate the updated second approximation estimates.

• SIBEC data and sampling standards were developed in 1997. These standards have received subsequent revisions in the past few years. The most recent set of standards (version 5.1) came out in June 2001 (B.C. Ministry of Forests Site Productivity Working Group 2001).
• After 1995, new data were collected for the provincial SIBEC data warehouse, a central repository for ecological and mensuration data. This data came from work carried out on Tree Farm Licenses, Innovative Forest Practices Agreement lands, Permanent Sample Plots, and other projects. In 2000-01, the SIBEC data warehouse was re-designed and modelled to efficiently store and retrieve all data fields. A modified VENUS data model was used to develop data entry and reporting tools. Approximately 2200 new records have been added to the data warehouse.
• In 2001-02, a SIBEC administrator analysis tool was designed and developed to create a SIBEC Reporting Data Mart (RDM). The RDM obtains specific data from the SIBEC data warehouse to complete mean site index calculations for each BEC site series/tree species combination. This process allows the generation of the site index site series (SISU) reports by region and biogeoclimatic unit.
• The following methodology is used for analysis and reporting of second approximation estimates:
  • New data are added to the provincial SIBEC data warehouse. Site index is calculated for all plots with available data. Mean plot site index and its standard error are calculated for each BEC site series/species combination that have a minimum sample size of 7.
  • This new estimate replaces the first approximation estimate. However, BEC site series/species combinations that do not have the minimum number of samples will retain the first approximation site index class estimate in the reports.

• Improved accuracy: The second approximation site index estimates are obtained through the addition of new data collected following the revised SIBEC data and sampling standards for ecological and site tree data. In addition, previously collected data were reviewed, and those data inadequately meeting the current data and sampling standards were removed from the database.
• Added Precision Estimates: First approximation site index reliability ratings are replaced by the mean site index and standard error of the mean.
• Expanded coverage: Site index estimates for some hardwood tree species are also included.