Ecological Archives A022-121-A2

Julia L. Moore, Song Liang, Adam Akullian, Justin V. Remais. 2012. Cautioning the use of degree-day models for climate change projections in the presence of parametric uncertainty. Ecological Applications 22:2237–2247. http://dx.doi.org/10.1890/12-0127.1

Appendix B. Specifics of parameter estimation methods and results.

Parameter sets were selected for use in the sensitivity analysis when the following four criteria were met: (1) the agreement between simulated presence and historical presence was greater than 50%; (2) the agreement between simulated absence and historical absence was greater than 60%; (3) the total agreement between simulated data and historical data was greater than 60%; and (4) the agreement at selected locations with specific regional characteristics was greater than 80%. These rather flexible criteria allowed parameters to be selected even when fitting the model to the coarse historical data using only temperature as a predictive variable. Indeed, higher agreement between simulations and historical data is not expected. For the fourth criteria, three counties in mountainous regions (n = 31 cells) and three counties in the warmer, low-lying regions of the Province (n = 9 cells) were chosen to ensure that selected parameter sets yield output that is regionally consistent.

From the 2000 simulations, 81 parameter sets met the first three criteria. A further 8 were excluded for failing the fourth criteria. This yielded a total of 73 parameter sets. Of these final parameter sets, agreement between simulated and historical snail presence ranged from 50% to 63%, agreement between simulated and historical snail absence ranged from 61% to 73%, and total agreement between simulations and historical data ranged from 60% to 66%.


[Back to A022-121]