Office preparation prior to field work is a critical phase in the implementation of the OASIS method. The quality of the site series proportions derived by the OASIS method are only as good as the data collected in the field and the statistical procedures used to analyze the data. Ensuring that the sampling strategy will yield appropriate data and that the analytical approach will yield reliable results are direct results of the program planning phase.

The following procedures describe the survey design and pre-screening phase of the program.

The SIBEC data is reviewed, with the assistance of the SIBEC project leader, to identify gaps in the SIBEC database. The SIBEC project is documented in the Site Productivity Working Group’s web site:

Contacts for this project can be obtained through the chairperson of the Site Productivity Working Group. The SIBEC site series/species combination that require more data can be targeted for supplemental sampling in the OASIS sampling plan.

Plot locations are selected using simple random sampling with replacement. Plots that fall within water bodies, alpine areas, or other non-productive areas must not be replaced by alternate plot locations. Discussion of the simple random sampling method is provided in Linnell Nemec (1997), attached as Appendix 1. We recommend establishing 500 plots. This number can be increased if better precision is desired. A statistician should be consulted to determine a suitable sample size if increased precision is needed.

GIS code is provided to establish random points in a TSA or TFL. A GIS specialist should be contacted to provide guidance in selecting the random points. You should specify the number of points that you want. The GIS code generates 4 times as many points as you specify in case you need to establish more plots than you anticipated to achieve a specified precision. Although not necessary, it would be a good idea to plot the random points on a map of the TSA or TFL to detect errors. The points should appear to be randomly spread over the TSA/TFL. If they aren’t, there may be an error in the randomization procedure, or perhaps, strictly by chance, the points are not randomized very well. In the latter situation, the randomization procedure should be repeated. After the random points are selected, they should be imported into an Excel spreadsheet which, by the end of the project, will become the data file and will contain the data analysis (Exhibit 3.2.1).

An alternative method of selecting plot locations is to choose them using a system of simple random sampling on a uniform grid. Sampling locations are chosen randomly with replacement from a 100 m by 100 m provincial grid, which is aligned with the 20 km by 20 km National Inventory Grid established by Natural Resources Canada, overlaid on the TSA or TFL. This alternative has the advantage of potentially aligning the plots with other inventory programs, which may allow the sharing of data across programs.

The next task of the survey design and office pre-screening is to locate and mark the plot locations on the forest cover maps and air photos. Accuracy at this stage of the program is important to facilitate the field program, particularly with planning logistics and assessing the degree of difficulty and time required to reach a plot location. The outcome of this task becomes the basis for developing a sampling plan and preparing the materials required to facilitate the field sampling and data collection phase of the program.

The following steps describe how to locate plots on the maps and air photos.

1. Mark the plot locations (i.e., latitude and longitude) clearly on the forest cover map (Exhibit 3.3.1) using mapping software (e.g., a GIS) and on the air photo (Exhibit 3.3.2) using a red grease pencil.
2. Locate the nearest suitable or accessible tie point using the air photo and forest cover map for each plot. The tie point should be a readily identifiable feature such as a helicopter landing zone, road junction, or distinctive water body. Where there is a discrepancy between the logistical information presented on the map and the photo, the document with the most current date is to be considered the most accurate. As an example, a cut block exists on a map but not on the corresponding photo. The map is more current than the air photo and should be used for locating the tie line and any other information relevant to the plot location.
3. Identify the tie point on the forest cover map (Exhibit 3.3.1) using mapping software and on the air photo (Exhibit 3.3.2) with a red grease pencil.
4. Draw a tie line from the tie point to the plot centre on the forest cover map (Exhibit 3.3.1) using mapping software and a red grease pencil for the air photos (Exhibit 3.3.2).
5. Determine the bearing and distance from the tie point to the plot centre.
6. Write the bearing and distance of the tie line near the tie line on the forest cover map (Exhibit 3.3.1) using mapping software and on the air photo (Exhibit 3.3.2) using a red grease pencil.
7. Estimate the degree of difficulty and time required to reach the plot and enter these data into the OASIS Reconnaissance Form (Exhibit 3.3.3).

A sampling plan is developed after the plot locations are identified on the forest cover maps and air photos. The sampling plan provides information regarding each potential plot, including its location, air photo coverage, access information, and difficulty rating.

The following is a guideline for preparing the sampling plan:

1. For each plot location provide a file containing:
1. An 8.5 x 11 copy of the forest cover map produced by mapping software with the required identifying features labeled as described in step 2, section 3.3 (Exhibit 3.3.1).
2. Photos that provide stereo coverage for the plot, tie point, and tie line (Exhibit 3.3.2).
3. A reconnaissance form containing access information, including mode of access (four-wheel drive, boat, helicopter, etc.), and difficulty rating (Exhibit 3.3.3).
2. Some plots will be located in areas where it is possible to classify the plot from the aerial photograph. For example, if the plot is located on a water body, rock, glacier, farm, urban or industrial area, etc., then the plot can be assessed from the photograph. Identify the plots that can be classified from aerial photographs in the OASIS Reconnaissance Form (Exhibit 3.3.3). Note: You must be positive that the plot can be classified from the photograph. If the plot is located near the edge of a glacier, water body, etc., then it must be field inspected to confirm its location.
3. Pre-screening will identify some plots that have difficult or dangerous access, or that require going on private land. These plots should be identified as air call plots. Identify the air call plots in the OASIS Reconnaissance Form (Exhibit 3.3.3).
4. Identify the plots that are to be field inspected and the mode of access that they require in the OASIS Reconnaissance Form (Exhibit 3.3.3).
5. Develop a strategy to establish plots to make the most efficient use of helicopter and crew time. The plots do not need to be established in the random order, but all plot locations in the sampling plan must be sampled. For example, if 500 samples are to be established then you must use the first 500 samples from the random list. You cannot replace any of the first 500 locations with other locations in the sampling plan (remember, the randomization procedure produces a list that contains 4 times as many sample locations as you specify). However, you do not need to sample plot 1 first, plot 2 second, and so on.

The sampling plan must be signed off by the project coordinator before plot establishment commences.

Continue to Chapter 4.
Return to Table of Contents.