The main objective of this manuscript is to present and discuss the application of SLAMM to the New York coast. Although the base analysis considers a range of different possible SLR scenarios, the effects of various sources of uncertainties such as input parameters and driving data are not accounted for. In addition, refined and site-specific data are often not available requiring the use of regional data collected from literature and professional judgement in order to run the model. To ignore the effects of these uncertainties on predictions may make interpretation of the results and subsequent decision making misleading since the likelihood and probabilities of predicted outcomes would be unknown. A unique capability of the current version of SLAMM is the ability to aggregate multiple types of input-data uncertainty to create outputs accompanied by probability statements and confidence intervals. Uncertainty in elevation data layers have been considered by several modeling groups to various extents (Gesch, 2009; Gilmer and Ferdaña, 2012; Schmid et al., 2014). However, to the best of our knowledge, no other marsh migration model simultaneously accounts for the combined effects of uncertainty in spatial inputs (DEM, VDATUM, etc.) and parameter choices (accretion rates, tide ranges, etc.) on landcover projections. This added feature of SLAMM allows results to be evaluated in terms of their likelihood of occurrence with respect to input-data and parameter uncertainties. Further, by assigning wide ranges of uncertainty when appropriate, it permits the production of meaningful projections in areas where high-quality local data are not available.
