Plot-GEM Example #2
Plot-GEM
Your personal assistant for exploration of sampling alternatives


Example #2: How accurate are our estimates based on the plots we have?

In 2000, Region 1 (R1) conducted a pilot study in the Idaho Panhandle National Forest. The study area contained 59 existing FIA primary sampling units (PSU's). R1 intensified this grid such that there were 400 PSU's on a regular grid in the study area. One of the questions R1 wished to answer via the pilot study was "How much accuracy will be gained by intensifying the FIA grid?"

Plot-GEM was used to generate the graphs in Figure 2. The percent error graph on the left shows that rapid accuracy improvement was made between a sample size of 10 PSU's and 110 PSU's. When the sample size is increased above 110 PSU's accuracy continued to increase, but at a much slower rate. These conclusions are mirrored by the 95% confidence interval graph on the right. The width of the confidence interval shrinks markedly between 10 PSU's and 110 PSU's, but the rate of improvement slows greatly beyond 110 PSU's.

Figure 2 Percent error graph and 95% confidence interval graph for the estimate of mean live basal area per acre (ba) across the entire study area.

From the Plot-GEM output, managers in R1 were able to conclude that the existing FIA grid was sufficient for meeting their accuracy goals: estimating live basal area per acre with an accuracy of ±20%. Further, they were able to see that a single intensification of the FIA grid would decrease the percent error of the estimate from ±7.7% to ±5.4%. Any intensification beyond this would clearly be a questionable investment of time and money.

Thus, Plot-GEM was able to help answer three questions:
  1. How accurate is the current estimate of the mean?
    • The estimate of mean live basal area per acre using the existing FIA grid is accurate to within ±7.7%.
    • The 95% confidence interval for the estimate of mean live basal area per acre using the existing FIA grid is (130.58 ft2, 175.85 ft2).
  2. What estimation accuracy is possible?
    • The estimate of mean live basal area per acre can be estimated to within ±3%.
    • The 95% confidence interval for the estimate of mean live basal area per acre is (142.21 ft2, 160.6 ft2) when the highest sampling intensity is used (400 PSU's).
  3. What estimation accuracy is feasible?
    • The estimate of the mean live basal area per acre using a single intensified FIA grid is accurate to within ± 5.4%.
    • The 95% confidence interval for the estimate of mean live basal area per acre is (136.2 ft2, 169.0 ft2) when the single intensified FIA grid is used (120 PSU's).


|Return to index|