Detecting recent changes in the areal extent of North Cascades glaciers, USA Michael A. O'Neal a, ⁎, Brian Hanson b , Sebastian Carisio b , Ashley Satinsky b a Department of Geological Sciences, University of Delaware, Newark, DE, USA b Department of Geography, University of Delaware, Newark, DE, USA abstract article info Article history: Received 4 March 2015 Available online 14 July 2015 Keywords: Glacier Cascades Climate change GIS Inventory Global warming We present an exhaustive spatial analysis using the geographic, geometric, and hypsometric characteristics of 742 North Cascades glaciers to evaluate changes in their areal extents over a half-century period. Our results in- dicate that, contrary to our initial expectations, glacier change throughout the study region cannot be explained readily by correlations in glacier location, size, or shape. Because of the large error attributable to annual varia- tions in glacier area due to snowpack, no statistically reliable change could be detected for 444 glaciers in our study (a slight majority). Of the North Cascades glaciers that do exhibit detectable change, a majority decreased in area, but nevertheless, some were detectably growing. These findings suggest that the integration of weather patterns over time does not neatly translate into correlations with natural variations in the geometry of glaciers. Our statistical analyses of the changes observed indicate that geometric data from a large number of glaciers, as well as a surprisingly large amount of spatial change, are required for a credible statistical detection of glacier- length and area changes over a short (multidecadal) period of time. © 2015 University of Washington. Published by Elsevier Inc. All rights reserved. Introduction The North Cascades Range of Washington State represents the most heavily glacierized area in the contiguous United States, where many individual glaciers have become high-profile indicators of regional and global climate change (Dyurgerov and Meier, 2000; Kovanen, 2003; Oerlemans, 2005; Pelto, 2008). Much of the currently available litera- ture discusses North Cascades glaciers in the context of retreat since the recent advances of the 16th to 19th centuries, or annual-to- decadal climate variations that result in substantial shorter-term fluctu- ations (e.g., Long, 1955; Harrison, 1970; Harper, 1993; Thomas, 1997; O'Neal, 2005). The majority of these analyses rely on a small number of large glaciers with substantial geochronological and historical records (e.g., South Cascade and Easton glaciers). Previous research of North Cascade glaciers that relied on field-based techniques was limited by the practical sample size of glaciers that could be studied. Alternatively, those studies that are able to combine remote- ly sensed imagery (multi-temporal airborne and spaceborne imagery) with traditional field-based methods are able to evaluate large popula- tions of glaciers at once, yielding much larger sample sizes, and regional inventories of glacier characteristics (e.g., Meier, 1961; Post et al., 1971). The few available studies that use GIS-based techniques to evaluate North Cascades glaciers have identified changes in termini at both the local and regional scale (Granshaw, 2002; O'Neal, 2005; Granshaw and Fountain, 2006; Satinsky, 2009). A notable effort in this regard is the North Cascade Glacier Climate Project (NCGCP), a comprehensive mon- itoring project that documents the geometric changes of 47 glaciers since 1984 (Pelto, 1993; Pelto and Riedel, 2001; Pelto, 2008). Other data- bases, especially GLIMS (Pfeffer et al., 2014), now contain many of these glacier outlines. However, our study specifically attempted to develop data with a one-to-one match for the ice features identified by Post et al. (1971). It is possible that evaluation and matching of different glacier boundaries in other inventories could be used to complete similar analyses with a longer temporal range or a greater spatial scale. In this study, we evaluate multidecadal changes in the areal ex- tents of glaciers from North Cascades, USA with a threefold purpose. First, we present the geographic, geometric, and hypsometric char- acteristics of 742 North Cascades glaciers, previously identified in AD 1958 maps and imagery by Post et al. (1971). With these data, we are able to 1) assess the areal changes over a nearly half-century, a period of time during which a global warming signal went from unde- tectable to unmistakable, and 2) evaluate spatial correlations between the geographic, geometric, and hypsometric variables of individual glaciers and changes in glacial area. Second, we evaluate the sensitiv- ity of our geometric data to important errors inherent to the process of remotely monitoring glacier populations. This includes the spatial error inherent to the imagery used, digitization-operator error, and temporal variations in ice-marginal snow cover, to estimate the threshold required to detect true areal difference—above noise—in glacier geometry change over time. Lastly, we evaluate the sample Quaternary Research 84 (2015) 151–158 ⁎ Corresponding author at: Department of Geological Sciences, University of Delaware, Newark, DE 19716. E-mail addresses: oneal@udel.edu; oneal.michael@gmail.com (M.A. O'Neal). http://dx.doi.org/10.1016/j.yqres.2015.05.007 0033-5894/© 2015 University of Washington. Published by Elsevier Inc. All rights reserved. Contents lists available at ScienceDirect Quaternary Research journal homepage: www.elsevier.com/locate/yqres