A Bayesian Network Modeling Approach to Forecasting the 21st Century Worldwide Status of Polar Bears - Arctic Sea Ice Decline: Observations, Projections, Mechanisms, and Implications - page 221

Plate 4. Projected polar bear population outcomes of Bayesian network model for four ecoregions at five time periods
relative to present. Present and prior decade (years 0 and −10) sea ice conditions were from observed record. Future ice
conditions were based on the ensemble mean of 10 GCMs and the two GCMs that forecasted maximum and minimum
ice extent in each ecoregion at each time period. Note that strength of dominant outcomes (tallest bars) is inversely pro-
portional to heights of competing outcomes. Outcome definitions are as follows: larger, more abundant than present (year
0) plus distribution at least the same as at present; same, numerical and distribution responses similar to present; smaller,
reduced in numbers and distribution; rare, numerically rare but occupying similar distribution or reduced numerically but
spatially represented as transient visitors; and extinct, numerically absent or distributionally extirpated.