Glacial geomorphology and geographic information systems

Glacial geomorphology and geographic information systems

Available online at Earth-Science Reviews 85 (2007) 1 – 22 Glacial geomorphology and geograp...

2MB Sizes 0 Downloads 38 Views

Available online at

Earth-Science Reviews 85 (2007) 1 – 22

Glacial geomorphology and geographic information systems Jacob Napieralski a,⁎, Jon Harbor b , Yingkui Li c a

University of Michigan-Dearborn, 4901 Evergreen Road, Dearborn, MI 48128, United States University of Colorado at Denver, Campus Box 144, Denver, CO 80217-3364, United States c University of Missouri-Columbia, 8 Stewart Hall, Columbia, MO 65211-6170, United States


Received 22 June 2006; accepted 20 June 2007 Available online 31 July 2007

Abstract Recent developments in the field of glacial geomorphology have dramatically increased the need to acquire, maintain, manipulate, and analyze large amounts of landform, landscape and sediment data. The use of Geographic Information Systems (GIS) has provided new platforms and tools for analysis and visualization of geomorphic data. Glacial geomorphologists have used GIS to integrate multi-source data, manage multi-scale studies, identify previously unrecognized spatial and temporal relationships and patterns in geomorphic data, and to link landform data with numerical models as part of model calibration and verification. GIS-based analyses associated with numerical modeling are improving our understanding of glacial landscape evolution and are allowing new quantitative and systematic examinations of spatial and temporal patterns of glacial landforms and processes. This has allowed for the development of insights and concepts that would be unlikely to arise using traditional methods alone. Key recommendations for future research and applications in glacial geomorphology include enhanced GIS education and dissemination, the development of standards and conventions for glacial geomorphic data, community projects to collect data into readily accessible databases, and enhanced use of linked GIS — model frameworks to address major issues in glacial geomorphology. © 2007 Elsevier B.V. All rights reserved. Keywords: Geographic Information Systems (GIS); glacial geomorphology; ice-sheet reconstructions; glacial landforms; database

1. Introduction A principal goal of glacial geomorphology is to describe and explain the impacts that glaciers and ice sheets have on landform and landscape development. This is achieved by integrating studies of landforms with empirical and theoretical studies of the processes responsible for their development (Harbor, 1993). Assessing and analyzing the spatial distribution and temporal evolution of glacial landforms is an impor⁎ Corresponding author. Department of Natural Sciences, University of Michigan-Dearborn, Dearborn, MI 48128, United States. E-mail address: [email protected] (J. Napieralski). 0012-8252/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.earscirev.2007.06.003

tant approach to better understanding landform genesis, and for revealing patterns and relationships of glacial landforms at various scales. The advent of Geographic Information Systems (GIS), and associated theoretical and technical advances in data handling and analysis, provides an opportunity for more sophisticated and effective analysis and modeling that in turn can have a significant impact on our understanding of glacial landform and landscape evolution (Clark, 1997). The goal of this paper is to provide a review of key issues in the development and application of GIS in glacial geomorphology, and to highlight emerging areas of research and applications (e.g. Fig. 1).


J. Napieralski et al. / Earth-Science Reviews 85 (2007) 1–22

Fig. 1. Emerging areas in GIS and glacial geomorphology include ice-sheet models linked to GIS, with the user able to experience model results through immersive visualization (Napieralski, 2005). In this case output during different time slices from an ice-sheet model is evaluated against a major moraine along the western margin of the Scandinavian Ice Sheet during the last major glaciation (top diagrams). Each time slice (1000 yr) is checked against the moraine using a statistical method developed in GIS (APCA; Napieralski et al., 2007) and a bar graph is used to indicate the time slice in which good correspondence occurs. Other GIS techniques (AFDA; Li et al., 2007) compare ice-flow orientation against suites of glacial lineations (bottom right) to indicate the potential times in which simulated ice-flow orientation best corresponds with field evidence. The visualizations illustrate various glaciological conditions, such as surface or subsurface temperature (bottom left) and ice-flow orientation (bottom right).

Earth scientists sometimes regard GIS as merely mapping tools (Wright et al., 1997), assuming the main purpose is to simply generate attractive maps. The tools associated with most commonly-used GIS packages allow for advanced data integration and sharing, numerical and cartographic modeling, spatial analysis, and advanced analyses of remotely sensed data integrated with other data sources. GIS have already had a significant impact on many areas of the Earth sciences, including hydrology (e.g. Djuokic and Maidment, 2000), watershed and natural resource management (e.g. Bhaduri et al., 2000), and mountain geomorphology (Bishop and Shroder, 2004). However, the use of GIS in glacial geomorphology has progressed relatively slowly, perhaps because of the emphasis placed on field investigations in glacial geomorphology and the lack of focused GIS education and technical training within the discipline. Thus, although some of the geospatial capabilities offered by GIS have been applied to advancing knowledge in glacial geomorphology, there are still many areas in which significant advances are possible. To take advantage of GIS it is important to understand how concepts and common spatial tools can be integrated into glacial geomorphology research, including

areas such as data assimilation and management, spatial analysis, and landform mapping. Many issues that are traditionally important in glacial geomorphology, such as scale, space–time representation, data representation and integration, spatial variation and process, and visualization, are also key components of the larger field of Geographic Information Science (GISc), which forms the basis for GIS. GISc is concerned with the philosophical and scientific foundation of spatial theory, and is a dynamic multidisciplinary field of study that utilizes skills and concepts from a variety of disciplines, such as cartography, geodesy, artificial intelligence, photogrammetry, and computer science (Goodchild, 1992). Advances in GISc, such as improved methods for data extraction and drainage delineation, may have potential uses in glacial geomorphology (O'Sullivan and Unwin, 2003), including improvements in defining alpine valley morphometry and glacial valley networks (e.g. Duncan et al., 1998). Similarly, techniques and theories developed to solve problems in glacial geomorphology have the potential for applications to other fields and to advance more general theories and methods in GISc. Key GISc issues that will be discussed in this review are data integration and sharing, scale and representation,

J. Napieralski et al. / Earth-Science Reviews 85 (2007) 1–22


ontology, numerical modeling, and visualization/animation capabilities. Sample applications will be discussed to illustrate ways in which GIS are being used in glacial geomorphology, including paleo-climate reconstructions using valley glacier extent, correlations between glacial landforms and sediments, verification of numerical icesheet models, and reconstructions of ice extent, flow direction, and subglacial conditions of paleo-ice sheets and glaciers. This review expands upon and updates a previous review by Clark (1997), which focused on the role of Geographic Information Technologies (GIT) in reconstructing dynamics of paleo-ice sheets, and reflects considerable improvements and broader use of GIS in glacial geomorphology in recent years. 2. Recent trends in glacial geomorphology Glacial geomorphology has traditionally focused on recognizing and describing glacial landforms and landscapes, and on advancing our understanding of glacial erosional and depositional processes and their interactions with ice and climate dynamics (Harbor, 1993; Martini et al., 2001). For example, the spatial distribution of glacial landforms such as drumlins (ice-flow direction) and end moraines (ice-marginal position) have been used to infer paleo-ice-sheet behavior (Punkari, 1995a,b; Kleman and Borgström, 1996; Colgan and Principato, 1998; Stokes and Clark, 2003), and to reconstruct ice-sheet extent and chronology (Kleman et al., 1997; Clark et al., 2000; Boulton et al., 2001). Understanding the relationship between patterns of glacial landforms and ice dynamics potentially provides valuable insight into paleo-ice sheets and paleoclimates (Clark, 1997; Ehlers and Gibbard, 2003). Recently, a significant emphasis in glacial geomorphology has been on the development of subglacial regime studies as critical components of understanding icesheet and glacier behavior (Kleman, 1994; Kleman and Hättestrand, 1999; Cuffey et al., 2000). The increasing quality and quantity of data that can be used to infer past spatial patterns of subglacial conditions have allowed icesheet reconstructions to evolve from relying primarily on end moraines to reconstruct stages of ice-sheet retreat (Boulton et al., 1985) to sophisticated integrations of proglacial and subglacial landform and sediment data with ice-marginal data to examine the dynamics of icesheet advance and retreat (Fig. 2; Kleman et al., 2004). The sheer volume of data, the range of data formats, and the need for a wide range of analytical approaches, are important reasons to use GIS in when relying on field data to reconstruct ice-sheet and glacier extent (i.e. inversion modeling; Kleman et al., 1999).

Fig. 2. A conceptual diagram of the differences between four ice-sheet reconstructions, focusing on particular time–space data domains and the data types used in the inversion procedures. The primary data domains: thick black line marks the deglacial landforms; boxes inside glacial margin mark ‘events’ reflected by till lineations pre-dating the final decay phase; triangles schematically illustrate radiocarbon dates (which always reflect ice-free conditions); light gray inside ice margin (in Dyke and Prest, 1987) represents a ‘stretching’ of the deglacial landform record for inferences about older non-deglacial events.

Similarly, the application of new techniques such as cosmogenic nuclide dating (e.g. Lal, 1991; Bierman, 1994; Gosse and Phillips, 2001) in glacial geomorphology (Fabel and Harbor, 1999; Bierman et al., 1999) has improved spatial and temporal constraints for paleo-icesheet reconstruction (Brook et al., 1996; Fabel et al., 2002; Stroeven et al., 2002; Li et al., 2005; Harbor et al., 2006). Thus the spatial distribution of glacial landforms can now be evaluated as the outcome of a sequence of erosional or depositional events (combination of synchronous and asynchronous events) from multiple stages of glaciation. Current research in glacial geomorphology emphasizes process modeling, including linkages between landscape development and ice-sheet dynamics and


J. Napieralski et al. / Earth-Science Reviews 85 (2007) 1–22

chronology, as well as relating erosion processes (and patterns) to smaller scale landforms and landscape elements. Numerical models of glacial processes and landform development continue to evolve, reflecting improvements in computation technology, the ability to model in 3-dimensions (e.g. Hubbard et al., 1998), and recent efforts to verify model output against field observations (Boulton et al., 2001; Näslund et al., 2003; Tarasov and Peltier, 2004; Napieralski et al., 2006; Li et al., 2007). In the future, simulated patterns of landform development will continue to require additional calibration or verification using field evidence, including data observed/measured by remote sensing, Global Positioning Systems (GPS) and dating techniques. 3. Spatial and temporal scale issues 3.1. Time–space representations and glacial geomorphology Although space–time relationships are important in both GISc (Raper and Livingstone, 1995; Peuquet, 2002) and glacial geomorphology, most GIS applications in glacial geomorphology still emphasize space-dominant approaches because of the widespread availability of undated spatial data. Traditional applications of GIS in glacial geomorphology focused primarily on the spatial distribution of glacial data, with much less emphasis on the time dimension. Naturally, it is relatively easier to map landforms compared to dating landforms and their stages of development. Space-dominant approaches have the advantage of being relatively easy to implement, and allow for the production of landform maps that fit well with traditional geologic and geomorphic mapping programs that have existed in many countries since the early twentieth century (Klimaszewski, 1990; Clark et al., 2004; Gustavsson, 2006). Using GIS to produce traditional maps that are useful for applications in areas such as engineering geology, resource extraction and hydrogeology brings new data management and map output capabilities to traditionally-valued products and is a useful first step in demonstrating the value of GIT. However, space-dominant approaches have limitations that ultimately affect inferences made about sequences and processes of landform development. Relying strictly on space-dominant approaches to distinguish and describe the development and distribution of glacial landforms neglects issues related to timing of erosion or deposition. For example, it cannot be assumed that moraines which mark the maximum extent of an ice sheet were formed at the same time. Rather, it is well accepted that many ice sheets experienced non-synchronous growth and decay along their

margins, and thus maximum extents occurred at different times along the length of many ice-sheet margins. In contrast to the space-dominant approach, the development of landforms along a “timeline” are established through dating techniques (e.g. radiocarbon, cosmogenic nuclides, varves), with little consideration of spatial variations or patterns. Although some issues lend themselves to either time- (e.g. occurrence of ice coverage or ice-free conditions at a particular site) or spacedominant (e.g. spatial relationship between landforms) approaches, in most cases it is important to integrate spatial and temporal patterns to provide new insight into glacial geomorphic issues. An integrated space–time approach to glacial landform data provides a more comprehensive perspective on landscape evolution. To achieve this, the spatial distribution of data is linked to time, expressed as either absolute or relative ages for the formation of a glacial landforms or deposits. Relative dating is used to “stack” or order the landform and sediment record in a sequence of development, and is often possible in areas with multiple glaciations and cross-cutting features (e.g. Kleman et al., 1997; Clark et al., 2000; Boulton et al., 2001). Where absolute dating is possible using evidence such as radiocarbon dates, cosmogenic nuclide exposure or burial ages, tephras, or paleo-magnetic data, the space– time relationships of glacial sediments and landforms can be analyzed within an integrated framework (Bishop and Shroder, 2004). 3.2. Spatial and temporal scale dependency In glacial geomorphology, as in many areas of the Earth sciences, both the features of interest and the processes that create them occur over a wide range of spatial and temporal scales. The scale at which any geomorphic study is conducted influences the outcome of the analysis (Walsh et al., 1998), and thus multi-scale analyses are often required to develop a comprehensive understanding of glacial landforms and processes (Levin, 1992). However, glacial landforms can be organized into a spatial hierarchical structure (Ahnert, 1988; Diakau, 1989), ranging from glacial striations (widths measured in millimeters) to drumlins (widths measured in meters) to imprints left by major ice streams (widths measured in kilometers). Techniques, such as fractal analysis, autocorrelation, and semivariance, have proved useful in examining the influence of scale on analyses of fluvial, coastal, and morphotectonic features (e.g. Chase, 1992; Lifton and Chase, 1992; Sung and Chen, 2004), and could equally be used in glacial geomorphology. Scale-dependency analyses have been

J. Napieralski et al. / Earth-Science Reviews 85 (2007) 1–22

used more widely in investigations of the surface characteristics of glaciers, using the fractal responses of ice as seen in remote sensing data (Bishop et al., 1998). Ice sheets and glaciated landscapes occur and develop over varying time scales and it is clear that much of our field evidence for active glacial processes has been collected on short time scales (up to decades at best) and in environmental conditions during the current interglacial that are far different from those involved in the formation of major landforms during glacial epochs (Harbor, 1992). GIS are capable of handling sophisticated analyses across multiple temporal and spatial scales, although the selection of appropriate spatial and temporal scale resolutions for analyses has significant impacts on compu-


tational efficiency (too high a resolution can result in unacceptable compute times) and the scales of landforms that can be examined (Abert, 1996; Walsh et al., 1998). In many studies the data that are available for an analysis have been collected on a variety of spatial and temporal scales (see Fig. 2), and thus it is necessary to merge low-resolution and high-resolution data. This requires that some data to be “smoothed” or “simplified” so that all the data can be treated together (for example, see Clark et al., 2004). Coarse digital elevation models (DEMs) may affect or limit the quality of analysis that can be conducted. For instance, drumlins can be delineated from DEMs, but the cell-size of the DEM will influence estimates of drumlin size (area and height) and calculations of elongation and volume (Fig. 3). As an

Fig. 3. DEMs were used to analyze drumlins in within the Palmyra Drumlin Field in Central New York. Contours were generated on a range of DEM resolutions to determine the morphometric characteristics, including height, length, width, elongation, and volume. Vertical profiles reflected changes in drumlin size and shape determined using various DEM resolutions. Overall, as the DEM resolution increased, estimates of volume decreased and elongation increased (Nalepa and Napieralski, 2007).


J. Napieralski et al. / Earth-Science Reviews 85 (2007) 1–22

example, DEMs of 80, 30, 10, 5, 3, and 1 m were used to extract drumlins (north of Palmyra, NY) from the New York drumlin field. In this simple case study, drumlin size and shape were affected by the resolution, such that the features became more streamlined (elongation ratio: length/width) and volume generally decreased as resolution increased from 80 m to 1 m. When switching from coarse resolutions (80–30 m) to a fine resolution (10–1 m), drumlin orientation was slightly affected, as the spine of the drumlin rotated several degrees (Nalepa and Napieralski, 2007). To further illustrate the importance of spatial resolution, the type of analysis used to extract the location of the equilibrium line altitudes of former glaciers is sensitive the resolution of the DEM and this, in turn, impacts the level of accuracy reconstructions of past glacier extent (Duncan et al., 1998). Topographic data resolution is of particular importance in glacial geomorphology, and in many areas even the highest resolution elevation models traditionally available may not be sufficient to allow for the identification of subtle landforms (Abert, 1996). However the increasing availability of high-resolution data from comprehensive satellite and shuttle-based programs is solving the traditional prob-

lem of gaps that were common in data for more remote areas (Gao and Liu, 2001), including the recent launches of “polar” satellites such as ICESat in 2003 (http:// 3.3. Spatial analysis and geostatistics One of the most compelling reasons to use GIS in glacial geomorphology research is the power and ease of use of advanced geostatistical and geospatial analysis tools that are integrated into most software packages (see Dunlop and Clark, 2006; Napieralski et al., 2007). A geospatial analysis can include objective studies of spatial correlations between sediments, landforms, and topography. In addition, descriptive statistics can be used to partition and transform data into new combinations or classifications (i.e. empirical GIS spatial modeling; Bishop and Shroder, 2004). Landform data can be queried and separated according to common attributes and viewed as new datasets, generating additional perspectives of patterns and correlations that may have otherwise gone unrecognized (Fig. 4). Advanced geostatistical tools, such as kriging and autocorrelation, can be used for spatial predictions and for

Fig. 4. Using a paper map, it is often difficult to distinguish spatial or temporal landform patterns. GIS permits landform assemblages to be “stripped” away into separate layers, which can then be arranged into various themes.

J. Napieralski et al. / Earth-Science Reviews 85 (2007) 1–22

investigating variance within or between glacial geomorphic studies (e.g. interpolation, prediction, and certainty of glacial drift characteristics). For example, interpolation algorithms and methods, such as kriging, splines, and bicubic functions, have been developed and tested (Evans, 1972; Zevenbergen and Thorne, 1987; Desmet, 1997; Wise, 1998; Hutchinson and Gallant, 1999), so that software users can focus more on finding the appropriate digital terrain representation than understanding the underlying mathematics of the geostatistical tool (Raseman et al., 2004). Of particular importance are geostatistical tools that enhance our understanding of scale dependencies and self organization, such as variograms and semivariance analysis (Hallet, 1990; Allen et al., 2004). Spatial analyses of digital elevation data have been used to establish altitudinal zonation in geomorphic features, allowing for the reconstruction of paleoclimate conditions (such as atmospheric circulation patterns determined from erosional and periglacial features, e.g. Humlum, 1997), and have been used to generate topographic profiles used to identify and delineate landforms (e.g. Dunlop and Clark, 2006). Most GIS provide tools for exploring and displaying elevation data, such as hillshading, slope angle, aspect, fractal, and curvature and many also have tools that automatically delineate topographic features, such as peaks or ridges and concave or convex hillside allowing rapid exploratory analyses of DEMs from glaciated terrain (Montgomery, 2004). In sum, the exploration of patterns and relationships using geospatial and geostatistical tools in GIS is dramatically enhancing our understanding of glaciated landscapes of glacial geomorphic processes, and relationships and feedbacks between process controls over a range of spatial and temporal scales. However, because of equifinality issues, it is always important to evaluate conclusions drawn using this approach in terms of what is already well known about the behavior and processes of glaciers and ice sheets. 4. Data representation and GIS in glacial geomorphology 4.1. Data representation choices In any geomorphic mapping project the features that are identified can be represented in databases in a wide variety of ways. For example, drumlins can be represented by a single line along the long axis, and classified in terms of orientation, spacing, or density. Alternatively, the boundaries of each drumlin can be digitized to


form a polygon, which provides an opportunity to calculate the area and to examine spatial parameters such as area and elongation of the drumlin. Because each representation allows for different analyses it is critical to choose a suitable data representation approach based on the needs/goals of the analysis. The representation of topography is often critical in glacial geomorphology and is typically addressed using a combination of cartographic principles and GIT. Digital elevation data, including measures of uncertainty, are typically derived from existing contour maps, photogrammetric interpretation of stereographic aerial images, laser scanning, terrestrial surveying, or aerial and space measurements (e.g. Boulton and Clark 1990; Aber et al., 1993; Stokes and Clark, 2001). Because many surface processes are constrained by landscape geometry and relief (Raseman et al., 2004), the integration of digital elevation data with glacial geomorphic data has provided new insight into these relationships (Clark, 1997). For example, an analysis of elevation, topography, and cosmogenic nuclide apparent exposure age for glacial erosional landforms in the northern Swedish mountains revealed a pattern of landscape protection and modification resulting from repeated glaciation by the Scandinavian Ice Sheet. This suggests cold-based (non-erosional) basal conditions dominated at intermediate elevations, whereas warm-based (erosional) basal conditions dominated at both low elevations and in high elevation valleys (Fabel et al., 2002). Digital elevation data thus provide the basis for linking topographic influences with geomorphic processes, and also providing opportunities to view the landscape in three dimensions with an overlay of satellite or aerial imagery, or glacial modeling results (e.g., Dunlop and Clark, 2006; Napieralski et al., 2007). 4.2. Ontology and taxonomy Ontology is a philosophical discipline that examines the use and implications of formal specifications in describing different types of object domains (e.g., Smith and Mark, 2001; Mark et al., 2004). This is important in glacial geomorphology because of the need to define and describe landforms using formal specifications for landforms (landform taxonomies) as part of landform process and development research (Rhoads, 1999). While some geomorphic features are clearly defined and have boundaries that are pronounced and wellestablished, the boundaries of many glacial landforms are frequently vague or subjective (Mark and Smith, 2004; Harbor et al., 2006). For example, attempts to categorize drumlins according to height, length, size


J. Napieralski et al. / Earth-Science Reviews 85 (2007) 1–22

ratio, orientation and other physical characteristics require decisions to be made about the definition of a drumlin and the capability to separate the drumlin from the surrounding landscape. Does the drumlin “begin” where it protrudes from the landscape or surrounding sediment? Or, are there subsurface characteristics that should be considered? Consequently, different analyses of a drumlin field may produce different results based on the lack of a widely-accepted nomenclature and set of definitions. GIS cannot directly answer these questions, but their use forces a more objective approach to addressing these issues; the user is obligated to address the issue of how landforms such as drumlins are defined, and once specifications and rules have been created automated feature extraction will impartially delineate all drumlins using these criteria. Thus, with the improving quality and quantity of the glacial landform databases, opportunities arise to scrutinize previous landform classifications (see Abert, 1996 for an example) and rectify “theory-laden evidence” (Rhoads and Thorn, 1993), or interpretations that have been promulgated through the decades but are no longer legitimate (e.g. Dunlop and Clark, 2006). Updating landform definitions is particularly important for the verification of numerical ice-sheet models, as sets of glacial lineations, frozen bed zones, moraines, and other glacial landforms are critical “building blocks” used in ice-sheet and glacier reconstructions. The taxonomy of glacial landforms needs to be clearly defined to avoid errors that may occur when digitized or delineated landform data based on different criteria are used in geomorphic analyses (Clark, 1997; Colgan et al., 2003). Landform terms have different meanings in various parts of the world, for example streamlined features such as drumlins (Smalley and Unwin, 1968; Boulton, 1987; Shaw and Sharpe, 1987) and ribbed moraines (Lundqvist, 1969; Markgren and Lassila, 1980; Hättestrand, 1997; Dunlop and Clark, 2006) are defined using different criteria in different parts of the world (Colgan et al., 2003), leading to potential confusion and alternate outcomes of large-scale or comparative analyses. Objectively defining glacial features according to geomorphometric parameters has become a significant and productive challenge for geomorphologists. For example, distinguishing between cirque classes forced geomorphologists to scrutinize the factors that control the spatial occurrence and morphology of the features (Unwin, 1973; King, 1974; Evans and Cox, 1974; Trenhaile, 1976; Evans, 1977; Gordon, 1977; Embleton and Hamann, 1988; Federici and Spagnolo, 2004). Parameters such as cirque volume, concavity, gradient, altitude, and orientation were combined as part of identifying the controlling

factors in cirque development (Gordon, 1977), however comparisons of cirques between mountain ranges have shown wide variations in the significance of different potential controlling variables, illustrating the difficulty of defining cirques based on a limited set of geomorphic characteristics that happen to be significant in one region (Embleton and Hamann, 1988). Embleton and Hamann suggested that other variables, such as duration or timing of ice coverage and subglacial thermal regime, need to be considered along with morphological characteristics to better explain variability in cirque forms. Thus it is essential that a wide range of morphometric characteristics be explicitly described in geomorphic studies of glacial landforms, and that these data be included in GIS databases, so that subsequent research aimed at linking landforms with the mechanics of possible land-forming processes over a variety of spatial and temporal scales are not limited by a reliance on a highly restricted range of parameters. 4.3. Data integration and sharing Databases used in studies of glacial geomorphology can include data from hundreds of sources (e.g. Clark et al., 2004) and include results that range from 19th century glaciological studies to recent aerial and satellite images and surveying (Evans et al., 2005). In addition, a variety of data formats are likely to be included, such as points (e.g. radiocarbon dates), lines (e.g. ice-flow indicators, till fabric analysis) and polygons (e.g. frozen bed zones), all then assimilated into a common theme for statistical analyses or visualizations (Napieralski, 2005; Dunlop and Clark, 2006). Finally, data used for an analysis can be merged from many scientific disciplines, thus producing a more comprehensive, multidisciplinary approach to a problem. For example, combining soil properties (Sharpe et al., 1999; Paulen and McClenaghan, 2000; Fenton et al., 2003), botanical evidence (e.g. Barnett and Singhry, 2000), geophysical data (e.g. Fenton et al., 2003), groundwater aquifer characteristics (e.g. Sharpe et al., 2003), and lithology data with landform data may reveal previously unrecognized relationships and correlations or produce new insight into the processes that produce specific landforms and landscapes. Efforts to accomplish large-scale data integration with traditional cartographic methods were relatively inefficient, as paper maps are generally rigid illustrations of spatial data that do not provide the ability to easily query data according to specific attributes. GIS have provided a platform that makes data integration much easier to accomplish in ways that also allow the user to account for different levels of data accuracy and precision.

J. Napieralski et al. / Earth-Science Reviews 85 (2007) 1–22

The ability to share data through advanced information technology continues to improve, and it is very likely that Web-based Geographic Information Systems (WebGIS), which are becoming increasingly routine (e.g. Pandey et al., 2000; Clark et al., 2004), will become a standard means of data communication. WebGIS provide the basis for allocating high-end GIS and mapping services through the Internet, so that users can integrate local data sources with internet data sources for display, query, and analysis in an easy-to-use Web browser. For example, the International Work Group on Geospatial Analysis of Glaciated Environments (GAGE), a subgroup of INQUA that was interested in investigating modern and ancient glaciated terrains using GIS, compiled a glacio-tectonic database as a dataset that is accessible on the internet at earthsci/gage/gage.htm, and specific data can be downloaded for use in other studies. Many software packages have been designed for WebGIS, such as ArcIMS, Internet Mapper, and Map Xtreme, and these will increase data sharing as viewing and display capabilities improve. In addition, web-based approaches (e.g. GoogleEarth) provide opportunities for geomorphologists who lack high-end computing capabilities to view and display data. This reduces the tendency for advanced analyses to be limited to those who have the funding to be able to purchase highly sophisticated computer clusters and supercomputers, increasing the range of individuals and groups who can participate in the science, and thus the diversity of approaches and insights that are generated. 4.4. Visualization and animation GIS have the functionality to produce a wide variety of maps and diagrams to display data and the results of


analyses. Traditionally, maps were labor-intensive, final steps in a research project, often created by a cartographer who was not involved in other steps of the research. The ease with which layouts and maps can be produced in a GIS has allowed a wide range of alternate visual displays of data to become an integral part of the analysis. As alternate displays reveal interesting patterns and relationships, new analyses and displays can be completed on-the-fly. This multifaceted, adaptable approach continues until a satisfactory result is attained, and research questions can be pursued more effectively and new insights may arise, increasing the breadth or depth of research (Fig. 5). Thus, output from a spatial analysis may actually function as a precursor to generating new questions and formulating new hypotheses. Many GIS now include photorealistic display capabilities. High-resolution imagery can be draped over DEMs to display complex glacial landscapes that the user can explore using virtual fly-through utilities. However, these approaches have limitations, such as the influence that azimuth-biasing (i.e. an illumination source or direction that influences the visualization) can have on the identification and delineation of subtle landforms (Smith et al., 2001; Smith and Clark, 2005). Recent work to evaluate alternate approaches for visualizing DEMs for landform mapping suggests that there was no single method that provides complete, unbiased visualizations; rather, alternate approaches such as relief-shading, combined viewing, surface derivatives, and spatial enhancements each have shortcomings that a user must take into consideration (Smith and Clark, 2005), In addition to standard 2- and 3-D displays (e.g., Sharpe et al., 1999; Atkinson et al., 2000; Fenton et al., 2003), it is becoming increasingly common to view glacial geomorphic data in 4-D (Buckley et al., 2004).

Fig. 5. The manner in which spatial problems are approached has drastically changed within the previous two decades, essentially a result of the manner in which spatial data are acquired, analyzed, and displayed. Geomorphologists frequently used aerial photos prior to field work to focus field efforts, which were then verified in the field. Now, visualizations and maps produced by GIS are now a part of the analytical process (O'Sullivan and Unwin, 2003).


J. Napieralski et al. / Earth-Science Reviews 85 (2007) 1–22

For example, in Napieralski (2005; Fig. 1) ArcGIS was used to integrate numerically derived ice-sheet data and field evidence in a simulation of the Scandinavian icesheet and high-resolution visualizations were created and displayed using additional software. 5. Recent advances in GIS and glacial geomorphology 5.1. Data integration and database management One of the most critical basic uses of GIS in glacial geomorphology has been management and integration of geomorphic and glaciological data. Efforts to create and maintain landform databases have been conducted on global, continental, and regional or localized scales. The INQUA Commission on Glaciation Work Group has undertaken one of the most comprehensive mapping efforts of glacial extent around the world (Ehlers and Gibbard, 2003). They used a digital-based technique that would allow both users and compilers the flexibility to query, select and view data using various projections and scales (Ehlers and Gibbard, 2003). Universal compilation guidelines were provided to facilitate data integration and management, and the end results included data layers containing glacial limits, end moraines, ice-dammed lakes, and glacierinduced drainage diversions for several Pleistocene glaciations (Ehlers and Gibbard, 2004a,b,c). At the XVI INQUA Congress in Reno Nevada, maps derived from GIS databases were presented for Central Europe (Ber and Aber, 2003; Ber et al., 2003) and the Baltic region (Zelčs, 2003). The INQUA datasets provide global data, and were intended to support large-scale studies (e.g. 1:1,000,000) such as examinations of the extent and chronology of the Pleistocene glaciations, and regional assessments of glacio-tectonic landforms. Other databases have been created to support small-scale, more localized studies, including specific databases for Britain (Clark et al., 2004), the northern Midwest region of the United States (Aber, 1999), the northern United States (Colgan et al., 2003), North America (Aber et al., 1993) and Europe (Aber and Bluemle, 1991; Croot and Michalak, 1993). For example, Colgan and Principato (1998) collected unpublished and published data from topographic maps, aerial photographs, and previous work, and used this to reconstruct the Green Bay and Lake Michigan lobes of the Laurentide Ice Sheet. Additional databases have been compiled to study Pre-Illinoian glacial geomorphology and dynamics in the central United States (Aber, 1999), to reconstruct the Laurentide Ice Sheet based on landform records (Clark et al., 2000), and to reconstruct ice-surface

geometry from erosional and periglacial features in the Alps (Kelly et al., 2004). There are also numerous national mapping organizations as well as university, local and municipal agencies that contribute landform and sediment data for glacial geomorphology studies. Recent satellite missions, such as NASA's Shuttle Radar Topography Mission (SRTM), have provided topographic data with almost global coverage, although at a relatively coarse spatial resolution (90–30 m) and excluding high latitude regions. The British Geological Society developed digitized glacial landform data for Britain (Clark et al., 2004), with the specific aim of providing glacial geomorphologists with data coverage that would stimulate new insights into the glacial history of the British Isles, as well as reveal gaps in the geomorphic record that are a high priority for future research. Databases have also been developed for local areas using intensive surface and subsurface mapping, including use of ground penetrating radar (Klempe, 2004). 5.2. Spatial analysis and spatial statistics Spatial analysis tools and spatial statistics are used to analyze relationships in data and to examine the statistical significance of patterns and associations. For example, spatial analysis tools have proved useful in the identification of anomalies in glacial landform and sediment data, and to examine the distribution of erosional, depositional, and relict landforms in glaciated landscapes (e.g., Fabel et al., 2002). Spatial statistical analyses provide additional capabilities, and in particular have been used to analyze distributions of landforms and to relate these to controlling processes such as ice-flow patterns and subglacial conditions in ways that would have been difficult using traditional cartographic methods. Recent examples include work relating subglacial bedforms and sediment characteristics (Colgan and Principato, 1998), the influence of meltwater on the development of subglacial landforms (Fisher et al., 2005), and the use of bedform attributes to determine flow sets (Fig. 6; Clark, 1997; Stokes and Clark, 2003). Particular emphasis has been placed on examining bedform density, packing, and parallel conformity of lineation sets (e.g. Clark and Wilson, 1994; Clark and Meehan, 2001; Stokes and Clark, 2001, 2003) to provide insight into subglacial regimes (Clark and Wilson, 1994) and to delineate ice divides, interlobate areas and zones of streaming ice (Punkari, 1993, 1995a,b; Stokes and Clark, 2001, 2003). For example, Punkari (1993, 1995a,b) used these techniques to conclude that zones of ice streams were dominated by fan-

J. Napieralski et al. / Earth-Science Reviews 85 (2007) 1–22


Fig. 6. A spatial distribution of glacial bedforms reveals a distinct difference that can then be used to discriminate between flow events: (A) hypothetical lineation pattern, (B) an interpretation that assumes all of the flow evidence is of the same age, (C) an alternative that takes account of the cross-cutting lineations, and (D) shows how the spatial pattern and morphometry can assist in discriminating flow events (from Clark, 1997).

shaped drumlin fields and intensive scouring, while interlobate areas were generally devoid of drumlins (Fig. 7). Glacial sediments and landforms (e.g. moraines) have also been combined to infer relationships between moraine type and ice margin positions (Maclachlan et al., 2003) and to extract possible sequences of ice margin positions during stages of retreat (Engel and Pair, 2001). In addition, the compilation of glacial features relative to bedrock geology and surface-sediment types has resulted in thematic maps that illustrate the relationships between glacial features, such as drumlins, and ice-marginal positions in the Midwest United States (Colgan and Principato, 1998). This work showed no spatial correlation between drumlin occurrence and bedrock lithology, but revealed that drumlins were grouped into individual fields associated with specific ice-marginal positions, and that eskers and drumlins were spatially correlated with the occurrence of sandy till. More detailed examinations of the characteristics of specific landforms, using morphometric analyses, have provided new insights into landform development. Dunlop and Clark (2006) conducted a morphometric analysis of ribbed moraines in North America and Northern Europe using remote sensing imagery and DEMs. This included an assessment of basic physical characteristics (e.g. length, width, height, slope, symmetry, “connectability”; Fig. 8) as well as correlations between moraine characteristics, location, topographic

characteristics, and proximity to other glacial landforms. The results highlighted the complexity of these moraine systems, and showed how an objective analysis using GIS tools can produce glaciological interpretations contrary to long-held views, such as those regarding ribbed moraine genesis (Dunlop and Clark, 2006) and drumlin formation (Boulton, 1987; Shaw and Sharpe, 1987). Any set of morphological observations must be constrained by hypotheses of landform genesis, so that the

Fig. 7. An illustration of the orientation and distribution of glacial landforms on the southern side of the Tampere interlobate zone (Finland). The orientations of different ice-flow indicators (plotted along x-axis) and their frequency (y-axis) was used to indicate the locations of ice lobes, streams, and melting beds (Punkari, 1993).


J. Napieralski et al. / Earth-Science Reviews 85 (2007) 1–22

Fig. 8. (A) Generalized overview of the Lac Naococane ribbed moraine ridge in Central Quèbec and estimated ice-flow direction. (B) The connectivity of moraines is determined by sliding moraines along the axis of ice-flow direction (solid line) and rotating the moraines to achieve a better fit (shaded moraines were rotated). (C) Summary of ribbed moraine morphology from Northern and Central Quèbec, Sweden, and Ireland. Shown are the means, ranges and “typical dimensions” of ribbed moraine parameters forms by various ice sheets (modified from Dunlop and Clark, 2006).

observations allow for rejection of one or more of the hypotheses (Martini et al., 2001; Dunlop and Clark, 2006). Similarly, GIS-based morphometric analyses have been used very productively to reveal trends in the shape, orientation, and altitude of cirques in the Alps (Federici and Spagnolo, 2004), to relate elongation, areal extent, spacing, and orientation of drumlins to subglacial conditions (Smalley and Unwin, 1968; Boulton, 1987; Shaw and Sharpe, 1987; Lanier and Norton, 2003;

Maclachlan and Eyles, 2005; Cook and Regis, 2005), and to analyze the morphologies of glaciated valleys (Duncan et al., 1998). In addition to expanding the types of analyses that can be performed, GIS have dramatically reduced the time involved in analysis; Clark and Wilson (1994) analyzed 4800 drumlins, and estimated that with traditional methods they would have spent more time on the analysis and would have been forced to work with a sample size of 50–1600 drumlins. The orientation of glacial bedforms has been an area of particular focus in recent studies because of the potential to use position and orientation of landforms to reconstruct ice-sheet behavior (e.g., see Fig. 7; Kleman, 1992; Kleman and Borgström, 1994; Clark, 1999). Orientation/rose diagrams, common tools in directional analyses (e.g. for drumlins, eskers, and till fabric) have been used to suggest possible ice-flow directions and patterns, illustrate the presence of a shifting center of mass, and explain the erosion–transportation–deposition process of ice sheets (Punkari, 1993; Clark et al., 2000; Näslund et al., 2003). Orientation/rose diagrams have also been used to distinguish primary and secondary ice-flow directions, which are critical in regions with complex ice-flow histories that have resulted in palimpsest landscapes. GIS have also been used to calculate ice-flow direction based on till fabric analyses (Treague and Syverson, 2002; Li et al., 2007) and to determine the source areas of erratics by combining remote sensing based flow directions with geologic data (Knight, 1996; Clark, 1997; Clark et al., 2000). Spatial analyses have also prompted the reassessment and revision of glacial landform classifications. For example, previously unidentified landscape features in central Illinois were recognized using highresolution DEMs and identified as end moraines (Abert, 1996). The relatively low relief of these ridges was ignored during earlier classifications but the close alignment of the ridges to previously mapped end moraines became apparent using gridding tools and cross-section profile analysis (Fig. 9). As a result, more mapped moraines are available for future reconstructions of the southern portion of the Laurentide Ice Sheet. The accessibility of high-resolution DEMs has also created opportunities to conduct more rigorous morphometric analyses of glacial landforms (e.g. ribbed moraines, drumlins). 5.3. Glacial model calibration and verification GIS-based modeling of geomorphic processes is a relatively new approach (Goodchild et al., 1993; Bishop and Shroder, 2004) that has great potential (Raper and

J. Napieralski et al. / Earth-Science Reviews 85 (2007) 1–22


Fig. 9. GIS can be used to examine landform classifications suggested by prior work. Topographic profiles were used to indicate ridges that may be interpreted as moraines (modified from Abert, 1996).

Livingstone, 1995). Recent work has focused on developing GIS-based approaches that allows for the verification of output from stand-alone ice-sheet models using geomorphic data (Napieralski, 2005; Napieralski et al., 2006, 2007; Li et al., 2007). In this work GIS data has been integrated with a stand-alone ice-sheet model to calibrate the model to maximize the agreement between ice-sheet simulations and field evidence (e.g. icesheet extent, ice core record, and isostatic rebound rates). To achieve this required the development of novel ways to statistically assess output from process models against a heterogeneous collection of field data that are spatially and temporally distributed (Napieralski et al., 2006; Li et al., 2007). Although ice-sheet modelers have long been interested in the reliability of model outcome, until recently there has been little emphasis on developing objective

ways to calibrate and validate model output using field data. With greatly improved computing capabilities and GIS software, it is now possible to assimilate output from numerical models with field observations. Ice-sheet models are used to examine ice-sheet evolution and so there is much interest in linking landforms to glacier dynamics, such as ice-marginal landform assemblages (Sharp, 1988; Hart, 1995; Evans et al., 1999) and subglacial landform assemblages (Kleman, 1992;Kleman and Hättestrand, 1999). The enormous amounts of field data pose a computing problem that can be resolved but, more importantly, the task of reducing these data to coherent sets, whether flow patterns from one glacial stage or a previous glacial stage (Kleman et al., 2004), can be accomplished by analyzing the physical characteristics of landforms (e.g. orientation, correlation with sediment data).


J. Napieralski et al. / Earth-Science Reviews 85 (2007) 1–22

In Napieralski (2005) a numerical ice-sheet model (from Hubbard, 1999) was used to simulate the growth and decay of the Scandinavian ice sheet and the output was compared to selected end moraines and glacial lineations. Key climatic input variables were altered to generate various ice-sheet configurations and topographies. Two GIS techniques were designed to statistically analyze the level of correspondence between model output and sets of major moraines and streamlines glacial landforms (i.e. glacial lineations). Automated Proximity and Conformity Analysis (APCA; Napieralski et al., 2006) and Automated Flow Direction Analysis (AFDA; Li et al., 2007) compare simulated ice-marginal extent against sets of moraines (Last Glacial Maximum and Younger Dryas) and simulated ice-flow direction against sets of glacial lineations (i.e. flow sets), respectively. APCA uses a system of buffers and overlays, common

GIS spatial analysis tools, to quantify the level of correspondence between model output and field evidence (Fig. 10). AFDA calculates the disparity (i.e. residual) between predicted basal flow direction and observed lineations; this residual is then plotted against their corresponding time slices to assess temporal patterns of correspondence between model output and field evidence (Fig. 11). Time series of one or more numerical models were then ranked according to their level of agreement with suites of field data, indicating periods of predicted marginal stability and fluctuating correspondence between ice-flow orientation data during ice-sheet growth, maxima, and decay, and to support rigorous sensitivity analyses of model input parameters (see Napieralski et al., 2007). More generally, an integrated GIS-based approach also allows measures or controls of glacial processes to be

Fig. 10. Automated Proximity and Conformity Analysis (APCA) uses a system of GIS-based buffering and overlays to quantify the level of agreement between simulated and observed ice extent by calculating the area-percent between predicted ice extent and major moraines (1). The cumulative plot of offset between features (determined by the area under the curve) is used to quantify the level of correspondence, based on the distance and angle between simulated and observed features. Three ice-sheet configurations are compared against the same two moraines to illustrate the application of APCA. Clearly, simulation 3 appears to match the moraines better, and APCA confirms this (large area under the curve). However, visual comparisons can be subjective and APCA is used to determine which moraines are reasonably well-matched, including distance (e.g. simulation 3, moraine B) and angle (e.g. simulation 1, moraine B are parallel, as reflected in the steep slope of the curve) (modified from Napieralski, 2007).

J. Napieralski et al. / Earth-Science Reviews 85 (2007) 1–22


Fig. 11. Steps of Automated Flow Direction Analysis (AFDA). (A) Field-based glacial lineations and model outputs used in the analysis. (B) Overlay model outputs and field evidence to produce series residual datasets (offset between simulated and observed) for different time slices. (C) Plot of mean residuals against their corresponding time slices to identify temporal patterns of agreement between predicted orientations and field observations. (D) Frequency analysis (rose diagrams) of selected time slices (e.g. d and f) provides detailed information on the distribution of residuals across the area (from Li et al., 2007).

merged with other fields to reconstruct current or previous glacial environments, including previous glacier extents and subglacial regimes. For example, combining DEMs and reconstructed ice-surface topographies of ice sheets substantiated the occurrence of glacio-hydraulic supercooling and basal freeze-on processes (Ensminger et al., 2002). In this work GIS were critical in providing a more realistic spatially-distributed reconstruction, compared to simple flow line calculations which, in turn, allowed the researchers to identify new matches/mismatches with field data in two and three dimensions. 6. Recommendations for the future of GIS and glacial geomorphology This review has focused on several key issues in applications of GIS in glacial geomorphology. The applications highlighted in this review have highlighted the importance of managing databases, data sharing, spatial analyses, and linking spatial and temporal field data with process models. Many of these utilities and applications are critical to productive multidisciplinary and international collaborations, producing new knowl-

edge and data that have allowed for reassessment of established theories and assumptions. However, the full benefits of GIS thinking and expertise have not yet been fully utilized in glacial geomorphology. The following recommendations are intended to stimulate more advanced use and appreciation of GIS in glacial geomorphology. 6.1. Education Although GISc and GIT have continuously improved, current applications in glacial geomorphology rely heavily on the experiences and skills of the user. It will take considerable time to develop this experience in the glacial geomorphology research community. Glacial geomorphologists need to be able to make appropriate procedural decisions, customize GIS to handle the data input and analysis, and to generate and interpret results (e.g. Plummer and Phillips, 2003). Traditional training in glacial geomorphology has emphasized the development of content knowledge and field skills; as digital data become more readily available, it is important that students of glacial geomorphology learn GIS and GIT as


J. Napieralski et al. / Earth-Science Reviews 85 (2007) 1–22

a routine part of their coursework. In addition, glacial geomorphology teaching that now commonly involves aerial photos or remote sensing should be enhanced to include mapping and spatial analyses as a way to provide new insights into glacial geomorphology and to expose students to GISc issues and skills. 6.2. Standards and benchmarks The wide range of data formats and GIS software is evidence of a rapidly expanding technology and discipline. Current efforts should focus on developing standards and benchmarks to improve data sharing and stimulate more comparative analyses. The methodology by which GIS are “customized” for glacial geomorphic studies need to be emphasized and made explicit in conference presentations, publications (e.g. Evans et al., 2005; Dunlop and Clark, 2006), technical reports and online documents. Enough detail should be provided to allow for published work to be reproduced, including data acquisition and description, software, steps in analysis (e.g. tools or scripts), and the process of generating output and visualizations. This allows others to scrutinize previous studies and improve techniques or approaches. While the design of GIS is typically not the focus of glacial geomorphology, it does warrant attention for several reasons. First, it provides a basis for which knowledge, approaches, and methodologies can be developed for future research, as current methodologies can be used as a foundation for designing new techniques. Second, it provides a means by which previous and current GIS approaches and results can be scrutinized. If data manipulation or methodology is flawed, the final results may likewise be erroneous. Open source software, which is freely distributed along with source code, alleviates some of the high costs associated with site licenses for software programs, and allows for the development of specific tools that may not be readily available in the commonly-used software packages (e.g. 3D display). Increasing use of open source software would help to stimulate international collaboration and to broaden the range of scientists who can afford to use GIS applications in geomorphology. 6.3. Data quality and error reporting Sources of error that occur during data acquisition and analysis should be routinely assessed and reported. In current studies errors and error propagation are generally neglected, yet they may have significant impacts on outcomes and conclusions. Methods to quantify errors, for example the root mean square (RMS) error, should

be used to control the quality of the analysis and understand the impacts of errors on the results and the error propagation (e.g. Clark, 1997; Colgan and Principato, 1998; Smith and Clark, 2005). Therefore, it is recommended that glacial geomorphology studies using GIS should include information on data quality, error levels, and error propagation. For example, recent efforts to build a database of glacial features of the British Ice Sheet included source information for each digitized feature (Clark et al., 2004). Errors associated with DEMs have also been reported, as this may influence a scale dependent spatial analysis of landforms (Dunlop and Clark, 2006). This effort provides the possibility to trace the data to its original source and resolve contradicting versions of landform type or location that occurred during the digitizing process. 6.4. Research and applications There are many directions to be recommended for future applications and research in glacial geomorphology, and a few examples are provided here to suggest possible new research avenues, such as spatial analysis, modeling, data integration, and artificial intelligence. 6.4.1. Spatial analysis Most GIS allow for analyses of the spatial attributes of geographic data, such as landforms or sediment deposits. An exploratory analysis of glacial landform data may reveal trends or abnormalities in the data and uncover special characteristics that are of interest to glacial geomorphologists. Therefore, glacial data should be more fully explored before interpretation or presentation. For example, ArcGIS and IDRISI include tools for analyzing spatial autocorrelations, which can yield new perspectives on landform distributions; these new perspectives can then be better related to geomorphic processes. In addition, kriging and other geostatistical tools can be used to interpolate subsurface distributions of glacial materials. Many geostatistical and spatial analysis tools developed for other disciplines (e.g.hydrology, soil science, and ecology) have great potential for use in glacial geomorphology (e.g. Maidment and Djokic, 2000; Maidment, 2002; Clarke et al., 2002). Spatial analyses can provide new insight in patterns and distributions of landforms relative to each other and to controlling variables. Cirques, drumlins, and ribbed moraines vary in spacing, orientation and elevation, and how these characteristics relate to variations to glaciological parameters of former glaciers and ice sheets can provide new insights into the development of landforms. Similarly, linking the pattern of frozen bed areas predicted by

J. Napieralski et al. / Earth-Science Reviews 85 (2007) 1–22

ice-sheet models with mapped sets of glacial lineations can provide a way to constrain simulated subglacial thermal conditions for alternate ice-sheet reconstructions. 6.4.2. Modeling GIS-based modeling is currently limited by software capabilities, and so advanced applications take the approach of linking data from GIS to stand-alone models, either as a calibration or verification step (see Goodchild et al., 1993 for examples of linking GIS and environmental models). In glacial geomorphology such work has focused on climate–glacier interaction (e.g. Kaser, 2001) and subglacial conditions (Harbor, 1992; Copland and Sharp, 2001; Truffer et al., 2001), though much of the contemporary interest in glacial geomorphology emphasizes the linkage between landform morphometry/ patterns and glaciological processes (and perhaps postglacial processes of modification). Future work will continue to advance numerical models that can be evaluated against field observations in more quantitative manners using GIS (e.g. Napieralski, 2005). Larger-scale Earth system models, such as the Grid Enabled Integrated Earth System Model (GENIE) (see Gulamali et al., 2003 or, will advance the study of paleo-ice-sheet characteristics (ice-flow direction, isostatic depression and ice topography, subglacial thermal conditions) and provide geomorphologists with unique opportunities to link output from large-scale models with mapped flow sets of glacial lineations, the distribution of frozen beds, and evidence from isostatic rebound to better constrain the glaciological conditions for landscape evolution. 6.4.3. Remote sensing and in situ data Many of the analyses presented in this review made use of data on external morphometry or land form patterns based solely on remote sensing data, without use of data acquired in the field. Additional insights can be provided by linking remotely sensed data with information on the internal structure of landforms. For example, many drumlin and ribbed moraine fields exhibit forms that are similar in orientation or morphometry, but which have wide variations in sedimentological characteristics (e.g. Meehan et al., 1997; Dunlop and Clark, 2006). When internal structure, morphometry, and larger-scale patterns are analyzed together it is likely that additional insights will be gained in glaciological processes and landscape evolution. 6.4.4. Artificial intelligence Artificial intelligence (AI) offers significant potential to assess impacts of particular forcing parameters (e.g.


climate, topography) on complex, previously glaciated terrain that interacted with processes over large spatial and temporal scales (Moody and Katz, 2004). Applications of AI in geomorphology and subsurface geology have included artificial neural networks (Rizzo and Dougherty, 1994; Bishop et al., 1999; Campolo et al., 1999; Gautam et al., 2000), cellular automota (Chase, 1992; Luo, 2001), fuzzy sets (Burrough, 1989; McBratney and Odeh, 1997; Oberthur et al., 2000), and genetic algorithms (Seibert, 2000; Seibert et al., 2000). However this approach has not yet been widely exploited in glacial geomorphology. For example, cellular automata, which can be used to model processes in 3D, may provide a new tool for re-evaluating the processes responsible for the formation and modification of glacial landforms. Cellular automota models rely on prior knowledge or sets of rules about processes, such as those that exist at the ice-land interface. As a result, landforms can be partitioned into new data that illustrate spatial relationships that would otherwise be difficult to recognize with visual or qualitative techniques. Fuzzy set theory, which focuses on vague or undefined boundaries between objects or the inclusion of objects within several classes (Moody and Katz, 2004), can potentially distinguish and help classify landforms that are somewhat ambiguous (e.g. boundaries of a drumlin). Overall, AI provides glacial geomorphologists with new opportunities to evaluate and develop theories of landscape evolution and consider both spatial and temporal dynamics. 7. Conclusions Deciphering glacial landform genesis and paleoglaciation, the core of glacial geomorphology, relies on describing and analyzing the morphological and spatial characteristics of glacial landforms and deposits, and linking these to process domains, all of which can be advanced through the use of GIS and GIT. However, to take advantage of this potential requires an understanding of issues such as scale, data representation, ontology, time–space representation, and modeling and visualization limitations. There has been noteworthy progress to date, as many of the past and current applications of GIS in glacial geomorphology have yielded new information on the relationships and patterns of landforms, as well as dispelling or confirming assumptions or conjectures on landform morphometry or genesis. However, future successful applications in glacial geomorphology will hinge on our ability to link scientific thinking and skills within both disciplines. Glacial geomorphologists must identify and


J. Napieralski et al. / Earth-Science Reviews 85 (2007) 1–22

appreciate the potential and limitations of GIS and embrace the concepts and theories that are the underpinnings of GISc. In this way, spatial information related to glacial landforms will be better understood, and various GIT will enhance geomorphic field-based projects and numerical modeling efforts. The full potential of applications is also limited by user expertise, so it is critical to enhance GIS education within the discipline of glacial geomorphology. Once the full potential of GISc is recognized and appreciated, there will not only be more contributions of new knowledge to glacial geomorphology, but also the potential to contribute to the ongoing development of GISc. Acknowledgements The authors would like to thank three anonymous reviewers for their insightful comments and suggestions which significantly improved the flow and content of this paper. A portion of this paper was completed while Napieralski was supported as a US Department of Education GAANN fellow at Purdue University and through the Rackham Fellowship at the University of Michigan, and this material was based upon work supported by the National Science Foundation under Grant No OPP-0138486 to Harbor. Li's work was also supported in part by the Research Council of the University of Missouri (grant no. URC-07-042). References Aber, J.S., 1999. Pre-Illinoian glacial geomorphology and dynamics in the central United States, west of the Mississippi. Geological Society of America, Special Paper 337, 113–1339. Aber, J.S. and Bluemle, J.P., 1991. Great Plains Geotectonics. Miscellaneous Map 31, North Dakota Geological Survey. Aber, J.S., Spellman, E.E., Webster, M.P., 1993. Landsat remote sensing of glaciated terrain. In: Aber, J.S. (Ed.), Glaciotectonics and Mapping Glacial Deposits, Proceedings of the INQUA Commission on Formation and Properties of Glacial Deposits, pp. 215–225. Abert, C., 1996. Modeling glaciated terrains. Proceedings of the Sixteenth Annual ESRI User Conference, California, p. 61. Ahnert, F., 1988. Modelling landform change. In: Anderson, M.G. (Ed.), Modelling Geomorphological Systems. John Wiley & Sons, pp. 375–400. Allen, T.R., Walsh, S.J., Cairns, D.M., Messina, J.P., Butler, D.R., Malanson, G.P., 2004. Geostatistics and spatial analysis: characterizing form and pattern at the alpine treelin. In: Bishop, M.P., Shroder, J.F. (Eds.), Geographic Information Science and Mountain Geomorphology. Praxis Publishing, Chichester, UK, pp. 189–214. Atkinson, D.M., Deadman, P.J., Traynor, S., 2000. A digital terrain and GIS model of an Arctic esker near Lac du Gras in the Northwest Territories of Canada. Abstr. Programs — Sixth Circumpolar Symposium on Remote Sensing of Polar Environments, Yellowknife, Canada.

Barnett, P.J., Singhry, V.H., 2000. Satellite-based engineering geology terrain studies for Ontario's far-north. Summary of field work and other activities 2000. Ontario Ministry of North Development and Mines, pp. 38-1–38-5. Ber, A. and Aber, J.S., 2003. Glaciotectonic map of Central Europe in scale 1: 1 500 000. Abstract with Programs — XVI INQUA Congress, p 39–1. Ber, A., Zelčs, V., Jakobsen, P.R., Kalm, V., Matoshko, A.V., Karakanov, A., Zagorodnykh, V.A., Bitinas, A., Stephan, H.J., 2003. The national glaciotectonic maps of Central Europe. Abstract with Programs — XVI INQUA Congress, Paper No. 39-2. Bhaduri, B., Harbor, J., Engel, B., Grove, M., 2000. Assessing watershed-scale, log-term hydrologic impacts of land use change using a GIS NPS model. Environmental Management 26, 643–658. Bierman, P.R., 1994. Using in situ produced cosmogenic isotopes to estimate rates of landscape evolution: a review from the geomorphic perspective. Journal of Geophysical Research 99, 13885–13896. Bierman, P.R., Marsella, K.A., Patterson, C., Davis, P.T., Caffee, M., 1999. Mid-Pleistocene cosmogenic minimum-age limits for preWisconsinan glacial surfaces in southwestern Minnesota and southern Baffin Island: a multiple nuclide approach. Geomorphology 27, 25–39. Bishop, M., Shroder, J.F., 2004. Geographic Information Science and Mountain Geomorphology. Praxis Publishing, Chichester, UK. 486 pp. Bishop, M.P., Shroder, J.F., Hickman, B.L., Copland, L., 1998. Scaledependent analysis of satellite imagery for characterization of glacier surfaces in the Karakoram Himalaya. Geomorphology 21, 217–232. Bishop, M.P., Shroder, J.F., Hickman, B.L., 1999. High resolution satellite imagery and neural networks for information extraction in complex mountain environment. Geocarto International 14 (2), 17–26. Boulton, G.S., 1987. A theory of drumlin formation by subglacial sediment deformation. In: Menzies, J., Rose, J. (Eds.), Drumlin Symposium. Balkema, Rotterdam, pp. 25–80. Boulton, G.S., Clark, C.D., 1990. A highly mobile Laurentide ice sheet revealed by satellite images of glacial lineations. Nature 346, 813–817. Boulton, G.S., Smith, G.D., Jones, A.S., Newsome, J., 1985. Glacial geology and glaciology of the last mid-latitude ice-sheets. Journal of the Geological Society of London 142, 447–474. Boulton, G.S., Dongelmans, P., Punkari, M., Broadgate, M., 2001. Paleoglaciology of an ice sheet through a glacial cycle: the European ice sheet through the Weichselian. Quaternary Science Reviews 20, 591–625. Brook, E.J., Nesje, A., Lehman, S.J., Raisbeck, G.M., Yiou, F., 1996. Cosmogenic nuclide exposure ages along a vertical transect in western Norway: implications for the height of the Fennoscandian ice sheet. Geology 24, 207–210. Buckley, A., Hurni, L., Kriz, K., Patterson, T., Olsenholler, J., 2004. Cartography and visualization in mountain geomorphology. In: Bishop, M.P., Shroder, J.F. (Eds.), Geographic Information Science and Mountain Geomorphology. Praxis Publishing, Chichester, UK, pp. 253–285. Burrough, P., 1989. Fuzzy mathematical methods for soil survey and land evaluation. Journal of Soil Science 40, 477–492. Campolo, M., Andreussi, P., Soldati, A., 1999. River flood forecasting with a neural network model. Water Resources Research 35, 1191–1197. Chase, C.G., 1992. Fluvial landsculpting and the fractal dimension of topography. Geomorphology 5, 39–57. Clark, C.D., 1997. Reconstructing the evolutionary dynamics of former ice sheets using multi-temporal evidence, remote sensing and GIS. Quaternary Science Reviews 16, 1067–1092.

J. Napieralski et al. / Earth-Science Reviews 85 (2007) 1–22 Clark, C.D., 1999. Glaciodynamic context of subglacial bedform generation and preservation. Annals of Glaciology 28, 23–32. Clark, C.D., Meehan, R.T., 2001. Subglacial bedform geomorphology of the Irish Ice Sheet reveals major configuration changes during growth and decay. Journal of Quaternary Science 16, 483–496. Clark, C.D., Wilson, C., 1994. Spatial analysis of lineaments. Computers and Geosciences 20, 1237–1258. Clark, C.D., Knight, J.K., Gray, J.T., 2000. Geomorphologic reconstructions of the Labrador sector of the Laurentide Ice Sheet. Quaternary Science Reviews 19, 1343–1366. Clark, C.D., Evans, D.J.A., Khatwa, A., Bradwell, T., Jordan, C., Marsh, S.H., Mitchell, W.A., Bateman, M.D., 2004. Map and GIS database of glacial landforms and features related to the last British Ice Sheet. Boreas 33, 359–375. Clarke, K.C., Parks, B.O., Crane, M.P., 2002. Geographic Information Systems and Environmental Modeling. Prentice Hall, Inc., Upper Saddle River, NJ. 306 pp. Colgan, P.M., Principato, S., 1998. Distribution of Glacial Landforms and Sediments in Wisconsin and the Upper Peninsula of Michigan, an Application of GIS to Glacial Geology. Geological Society of America Abstract with Programs, vol. 30, no. 2, p. 11. Colgan, P.M., Mickelson, D.M., Cutler, P.M., 2003. Ice-marginal terrestrial landsystems: Southern Laurentide Ice Sheet. In: Evans, D.A., Rea, B.R. (Eds.), Glacial Landsystems, Edwin Arnold, London, pp. 111–142. Cook, D.B., Regis, R.S., 2005. Morphological analysis of the Menominee drumlin field using GIS. Abstracts with Programs — Geological Society of America 37 (5), 15. Copland, L., Sharp, M., 2001. Mapping thermal and hydrological conditions beneath a polythermal glacier with radioecho sounding. Journal of Glaciology 47, 232–242. Croot, D.G., Michalak, W., 1993. The European Glaciotectonic Data Base Project. In: Aber, J.S. (Ed.), Glaciotectonics and Mapping Glacial Deposits, Proceedings of the INQUA Commission on Formation and Properties of Glacial Deposits, pp. 251–258. Cuffey, K.M., Conway, H., Gades, A.M., Hallet, B., Lorrain, R., Severinghaus, J.P., Steig, E.J., Vaughn, B., White, J.W.C., 2000. Entrainment at cold glacier beds. Geology 28, 351–354. Desmet, P., 1997. Effects of interpolation errors on the analysis of DEMs. Earth Surface Processes and Landforms 22, 563–580. Diakau, R., 1989. The application of a digital relief model to landform analysis in geomorphology. In: Raper, J. (Ed.), Three Dimensional Applications in Geographic Information Systems. Taylor & Francis, pp. 51–77. Djuokic, D., Maidment, D.R., 2000. Hydrologic and Hydraulic Modeling Support with Geographic Information Systems. ESRI Press, Redlands, California. Duncan, C.C., Klein, A.J., Masek, J.G., Isacks, B.L., 1998. Comparison of Late Pleistocene and modern glacier extents in Central Nepal based on digital elevation data and satellite imagery. Quaternary Research 49, 241–245. Dunlop, P., Clark, C.D., 2006. The morphological characteristics of ribbed moraines. Quaternary Science Reviews 25, 1668–1691. Dyke, A.S., Prest, V.K., 1987. Late Wisconsian and Holocene history of the Laurentide ice sheet. Géographie physique et Quaternaire 41, 237–263. Ehlers, J., Gibbard, P., 2003. Extent and chronology of glaciations. Quaternary Science Reviews 22, 1561–1568. Ehlers, J., Gibbard, P.L. (Eds.), 2004a. Quaternary Glaciations — Extent and Chronology, Part I: Europe. Developments in Quaternary Science, vol. 2a. Elsevier, Amsterdam. 448 pp.


Ehlers, J., Gibbard, P.L. (Eds.), 2004b. Quaternary Glaciations — Extent and Chronology, Part II: North America. Developments in Quaternary Science, vol. 2b. Elsevier, Amsterdam. 440 pp. Ehlers, J., Gibbard, P.L. (Eds.), 2004c. Quaternary Glaciations — Extent and Chronology, Part III: South America, Asia, Africa, Australasia, Antarctica. Developments in Quaternary Science, vol. 2c. Elsevier, Amsterdam. 388 pp. Embleton, C., Hamann, C., 1988. A comparison of cirque forms between the Austrian Alps and the Highlands of Britain. Zeitschrift für Geomorphologie. Supplementband 70, 75–93. Engel, J.M., Pair, D.L., 2001. Combining surficial mapping with GIS to investigate the glacial history of the Skaneateles, New York region. Abstracts with Programs — Geological Society of America 33 (1), 25. Ensminger, S., Evenson, E., Lawson, D., Alley, R., Larson, G., 2002. Using GIS to test for the occurrence of glaciohydraulic supercooling and basal freeze-on processes on the Laurentide ice sheet. Abstracts with Programs — Geological Society of America 34 (2), 43. Evans, I.S., 1972. General geomorphometry, derivatives of altitude, and descriptive statistics. In: Chorley, R. (Ed.), Spatial Analysis in Geomorphology. Methuen, London, pp. 17–90. Evans, I.S., 1977. Worldwide variations in the direction and concentration of cirque and glacier aspects. Geografiska Annaler. Series A. Physical Geography 59, 151–175. Evans, I.S., Cox, N.J., 1974. Geomorphometry and the operational definition of ciques. Area 6 (2), 150–153. Evans, D.J.A., Lemmen, D.S., Rea, B.R., 1999. Glacial land systems of the Southwest Laurentide Ice Sheet: modern Icelandic analogues. Journal of Quaternary Science 14, 673–691. Evans, D.J.A., Clark, C.D., Mitchell, W.A., 2005. The last British Ice Sheet: a review of the evidence utilized in the compilation of the Glacial Map of Britain. Earth Science Reviews 70, 253–312. Fabel, D., Harbor, J., 1999. Application of in situ produced cosmogenic radionuclides techniques in glaciology and glacial geomorphology. Annals of Glaciology 28, 103–110. Fabel, D., Stroeven, A.P., Harbor, J., Kleman, J., Elmore, D., Fink, D., 2002. Landscape preservation under Fennoscandian ice sheets determined from in situ produced 10Be and 26Al. Earth and Planetary Science Letters 201, 397–406. Federici, P.R., Spagnolo, M., 2004. Morphometric analysis on the size, shape, and areal distribution of glacial cirques in the Maritime Alps (Western French–Italian Alps). Geografiska Annaler 86A (3), 235–248. Fenton, M.M., Andriashek, L.D., Pawlowicz, J.G., Paulen, R.C., 2003. Application of three dimensional mapping of drift in Alberta for resource development. Abstract with Programs — Joint Annual Meeting of the Canadian Quaternary Association and the Canadian Geomorphology Research Group, Halifax, Nova Scotia. Fisher, T.G., Jol, H.M., Boudreau, A.M., 2005. Saginaw Lobe tunnel channels (Laurentide Ice Sheet) and their significance in south-central Michigan, USA. Quaternary Science Reviews 24, 2375–2391. Gao, J., Liu, Y., 2001. Application of remote sensing, GIS and GPS in glaciology: a review. Progress in Physical Geography 25, 520–540. Gautam, M., Watanabe, K., Saegusa, H., 2000. Runoff analysis in humid forest catchment with an artifical neural network. Journal of Hydrology 235 (1–2), 117–136. Goodchild, M.F., 1992. Geographical information science. International Journal of Geographical Systems 6, 31–45. Goodchild, M.F., Parks, B.O., Steyaert, L.T. (Eds.), 1993. Environmental Modeling with GIS. Oxford University Press, New York.


J. Napieralski et al. / Earth-Science Reviews 85 (2007) 1–22

Gordon, J.E., 1977. Morphometry of cirques in the Kintail–Affric– Cannich. Geografiska Annaler. Series A. Physical Geography 59, 177–194. Gosse, J.C., Phillips, F.M., 2001. Terrestrial cosmogenic nuclides: theory and applications. Quaternary Sciences Review 20, 1475–1560. Gulamali, M.Y., Lenton, T.M., Yool, A., Price, A.R., Marsh, R.J., Edwards, N.R., Valdes, P.J., Wason, J.L., Cox, S.J., Krznaric, M., Newhouse, S., Darlington, J., 2003. GENIE: Delivering e-Science to the environmental scientist. UK e-Science All Hands Meeting, pp. 145–152. Gustavsson, M., 2006. Development of a detailed geomorphological mapping system and GIS geodatabase in Sweden. In: Digital Comprehensive Summaries of Uppsala Dissertation from the Faculty of Science and Technology, 236. 36 pp. Uppsala. ISBN 91-554-6705-9. Hallet, B., 1990. Spatial self-organization in geomorphology: from periodic bedforms and patterned ground to scale-invariant topography. Earth Science Reviews 29, 57–75. Harbor, J., 1992. Numerical modeling of the development of U-shaped valleys by glacial erosion. Geological Society of America Bulletin 104, 1364–1375. Harbor, J., 1993. Glacial geomorphology: modeling processes and landforms. Geomorphology 7, 129–140. Harbor, J., Stroeven, A., Fabel, D., Clarhäll, A., Kleman, J., Li, Y.K., Elmore, D., Fink, D., 2006. Cosmogenic nuclide evidence for minimal erosion across two subglacial sliding boundaries of the late glacial Fennoscandian ice sheet. Geomorphology 75, 90–99. Hart, J.K., 1995. Recent drumlins, flutes and lineations at VestariHagafellsjökull, Iceland. Journal of Glaciology 41, 596–606. Hättestrand, C., 1997. Ribbed moraines in Sweden — distribution pattern and Palaeoglaciological implications. Sedimentary Geology 111, 41–56. Hubbard, A., 1999. High-resolution modeling of the advance of the Younger Dryas ice sheet and its climate in Scotland. Quaternary Research 52, 27–43. Hubbard, A., Blatter, B., Nienow, P., Mair, D., Hubbard, B., 1998. Comparison of a three-dimensional model for glacier flow with field data from Haut Glacier d'Arolla, Switzerland. Journal of Glaciology 44, 368–378. Humlum, O., 1997. Younger Dryas glaciation in Söderåsen, South Sweden: an analysis of meterologic and topographic controls. Geografiska Annaler. Series A. Physical Geography 79, 1–15. Hutchinson, M.F., Gallant, J.C., 1999. Representation of terrain. In: Longley, P.A., Goodchild, M.F., Maguire, D.J., Rhind, D.W. (Eds.), Geographical Information Systems. John Wiley & Sons, New York, pp. 105–124. Kaser, G., 2001. Glacier–climate interaction at low latitudes. Journal of Glaciology 47, 195–204. Kelly, M., Buoncristiani, J.-F., Schlüchter, C., 2004. A reconstruction of the last glacial maximum (LGM) ice-surface geometry in the western Swiss Alps and contiguous Alpine regions in Italy and France. Ecologae Geologicae Helvetiae 97, 57–75. King, C.A.M., 1974. Morphometry in glacial geomorphology. In: Coates, D.R. (Ed.), glacial Geomorphology: Proceedings of the 5th Annual Geomorphology Symposium Series, 26–28 September 1974, Binghampton, New York, pp. 147–162. Kleman, J., 1992. The palimpsest glacial landscape in northwestern Sweden — Late Weichselian deglaciation forms and traces of older west-centred ice sheets. Geografiska Annaler. Series A. Physical Geography 74, 305–325. Kleman, J., 1994. Preservation of landforms under ice sheets and ice caps. Geomorphology 9, 19–32. Klempe, H., 2004. Identification of Quaternary subsurface glacial deposits using 3D databases and GIS. Norwegian Journal of Geography 58 (2), 90–95.

Kleman, J., Borgström, I., 1994. Glacial landforms indicative of a partly frozen bed. Journal of Glaciology 135, 255–264. Kleman, J., Borgström, I., 1996. Reconstruction of paleo-ice sheets: the use of geomorphological data. Earth Surface Processes and Landforms 21, 893–909. Kleman, J., Hättestrand, C., 1999. Frozen-bed Fennoscandian and Laurentide ice sheets during the Last Glacial Maximum. Nature 402, 63–66. Kleman, J., Hättestrand, C., Borgström, I., Stroeven, A., 1997. Fennoscandian paleoglaciology reconstructed using a glacial geological inversion model. Journal of Glaciology 43, 283–299. Kleman, J., Hättestrand, C., Clarhäll, A., 1999. Zooming in on frozenbed patches: scale dependent controls on Fennoscandian ice sheet basal thermal zonation. Annals of Glaciology 28, 189–194. Kleman, J., Hättestrand, C., Stroeven, A., Jansson, K.N., De Angelis, H., Borgström, I., 2004. Reconstruction of paleo-ice sheets: inversion of their glacial geomorphological record. In: Knight, P.G. (Ed.), Glacier Science and Environmental Change. Blackwell Publishing Ltd. Klimaszewski, M., 1990. Thirty years of geomorphological mapping. Geographia Polonica 58, 11–18. Knight, J.K., 1996. A Geographical Information System-based Synthesis of the Labrodor Sector of the Laurentide Ice Sheet. Unpublished Ph.D. thesis, University of Sheffield, UK, 206 pp. Lal, D., 1991. Cosmic ray labeling of erosion surfaces: in situ nuclide production rates and erosion models. Earth and Planetary Science Letters 104, 424–439. Lanier, A.J., Norton, K.P., 2003. GIS based analysis of the Chautauqua Drumlin Field, Northwest Pennsylvania and Western New York. Abstracts with Programs — Geological Society of America 35 (6), 425. Levin, S.A., 1992. The problem of pattern and scale in ecology. Ecology 73, 1943–1967. Li, Y.K., Harbor, J., Stroeven, A.P., Fabel, D., Kleman, J., Fink, D., Caffee, M., Elmore, D., 2005. Ice sheet erosion patterns in valley systems in northern Sweden investigated using cosmogenic nuclides. Earth Surface Processes and Landforms 30, 1039–1049. Li, Y.K., Napieralski, J.A., Harbor, J., Hubbard, A., 2007. Identifying patterns of correspondence between modeled flow directions and field evidence: an automated flow direction analysis. Computer & Geosciences 33, 141–150. Lifton, N.A., Chase, C.G., 1992. Tectonic, climatic and lithologic influences on landscape fractal dimension and hypsometry: implications for landscape evolution in the San Gabriel Mountains, California. Geomorphology 5, 77–114. Lundqvist, J., 1969. Problems of the so-called Rogen moraine. Sveriges Geologiska Undersökning, Series C, p. 648. Luo, W., 2001. LANDSAP: a coupled surface and subsurface cellular automata model for landform simulation. Computers and Geosciences 27, 363–367. Maclachlan, J.C., Eyles, C.H., 2005. Quantitative morphological analysis of the Peterborough drumlin field using GIS. Abstracts with Programs — Geological Society of America 37, 7. Maclachlan, J.C., Eyles, C.H., Terlaky, V., 2003. Understanding the origin of glacial landscapes using GIS imaging of deposits formed along the southern margin of the Laurentide ice sheet. Abstracts with Programs — Geological Society of America 35 (6), 298. Maidment, D.R., 2002. ArcHydro: GIS for Water Resources. Environmental Systems Research Institute Press, Redlands, CA. 203 pp. Maidment, D., Djokic, D., 2000. Hydrologic and Hydraulic Modeling Support with Geographic Information Systems. Environmental Systems Research Institute Press, Redlands, CA. 216 pp.

J. Napieralski et al. / Earth-Science Reviews 85 (2007) 1–22 Mark, D.M., Smith, B., 2004. A science of topography: from qualitative ontology to digital representation. In: Bishop, M.P., Shroder, J.F. (Eds.), Geographic Information Science and Mountain Geomorphology. Praxis Publishing, Chichester, UK, pp. 75–97. Mark, D.M., Smith, B., Egenhofer, M., Hirtle, S., 2004. Ontological foundations for geographic information science. In: McMaster, R.B., Usery, L. (Eds.), Research Challenges in Geographic Information Science. John Wiley & Sons, New York. Markgren, M., Lassila, M., 1980. Problems of moraine morphology, Rogen moraine and Blattnick moraine. Boreas 9, 217–274. Martini, I.P., Brookfield, M.E., Sadura, S., 2001. Principles of Glacial Geomorphology and Geology. Prentice-Hall, Inc., New Jersey. 381 pp. McBratney, A., Odeh, I., 1997. Application of fuzzy sets in soil science: fuzzy logic, fuzzy measurements and fuzzy decision. Geoderma 77 (2–4), 85–113. Meehan, R.T., Warren, W.P., Gallagher, J.D., 1997. The sedimentology of a Late Pleistocene drumlin near Kingscourt, Ireland. Sedimentary Geology 111, 91–105. Montgomery, D.R., 2004. GIS in tectonic geomorphology and landscape evolution. In: Bishop, M.P., Shroder, J.F. (Eds.), Geographic Information Science and Mountain Geomorphology. Praxis Publishing, Chichester, UK, pp. 425–453. Moody, A., Katz, D.B., 2004. Artificial intelligence in the study of mountain landscapes. In: Bishop, M.P., Shroder, J.F. (Eds.), Geographic Information Science and Mountain Geomorphology. Praxis Publishing, Chichester, UK, pp. 219–247. Nalepa, N., Napieralski, J., 2007. Using GIS to ascertain efficient spatial and vertical resolution for determining drumlin morphometry: a case study of the Palmyra Drumlin Field. Association of American Geographers Annual Meeting, San Francisco, CA, April 17–22. Napieralski, J.A., 2005. Integrating a Numerical Model of the Scandinavian Ice Sheet with Field Data using GIS. Unpublished PhD thesis, Purdue University, 1–133 pp. Napieralski, J., 2007. GIS and field-based spatio-temporal analysis for evaluation of paleo-ice sheet simulations. The Professional Geographer 59, 173–183. Napieralski, J.A., Li, Y.K., Harbor, J., 2006. Comparing predicted and observed spatial boundaries of geologic phenomena: Automated Proximity and Conformity Analysis (APCA) applied to ice sheet reconstructions. Computers and Geosciences 32, 124–134. Napieralski, J., Hubbard, A., Li, Y.K., Harbor, J., Stroeven, A.P., Kleman, J., Alm, G., Jansson, K.N., 2007. Towards a GIS assessment of numerical ice sheet model performance using geomorphological data. Journal of Glaciology 53, 71–83. Näslund, J.O., Rodhe, L., Fastook, J.L., Holmlund, P., 2003. New ways of studying ice sheet flow directions and glacial erosion by computer modeling — examples from Fennoscandia. Quaternary Science Reviews 22, 245–258. Oberthur, T., Doberman, A., Aylward, M., 2000. Using auxiliary information to adjust fuzzy membership functions for improved mapping of soil qualities. International Journal of Geographic Information Science 14, 431–454. O'Sullivan, D., Unwin, D.J., 2003. Geographic Information Analysis. John Wiley & Sons, Inc., New Jersey. 436 pp. Pandey, S., Harbor, J., Engel, B., 2000. Internet Based Geographic Information Systems and Decision Support Tools. Urban and Regional Information Systems, Park Ridge, Illinois. 36 pp. Paulen, R.C., McClenaghan, M.B., 2000. Integration of surficial mapping, stratigraphy and GIS to facilitate drift prospecting in glaciated terrain. Abstr. Programs — Geo Canada 2000, Calgary, Alberta. Peuquet, D.J., 2002. Representations of Space and Time. Guilford Press, New York. 380 pp.


Plummer, M.A., Phillips, F.M., 2003. A 2-D numerical model of snow/ice energy balance and ice flow for paleoclimatic interpretation of glacial geomorphic features. Quaternary Science Reviews 22, 1389–1406. Punkari, M., 1993. Modelling of the dynamics of the Scandinavian ice sheet using remote sensing and GIS methods. In: Aber, J.S. (Ed.), Glaciotectonics and Mapping Glacial Deposits, Proceedings of the INQUA Commission on Formation and Properties of Glacial Deposits, pp. 232–250. Punkari, M., 1995a. Function of ice streams in the Scandinavian ice sheet: analyses of glacial geological data from southwestern Finland. Transactions of the Royal Society of Edinburgh. Earth Sciences 85, 283–302. Punkari, M., 1995b. Glacial flow systems in the zone of confluence between the Scandinavian and Novaya Zemlya ice sheets. Quaternary Science Reviews 14, 589–603. Raper, J., Livingstone, D., 1995. Development of a geomorphological spatial model using object-oriented design. International Journal of Geographical Information Systems 9, 359–383. Raseman, S., Schimdt, J., Schrott, L., Dikau, R., 2004. Geomorphometry in mountain terrain. In: Bishop, M., Shroder, J. (Eds.), Geographic Information Science and Mountain Geomorphology. Praxis Publishing, Chichester, UK, pp. 101–145. Rhoads, B.L., 1999. Beyond pragmatism: the value of philosophical discourse for physical geography. Annals of the Association of American Geographers 89, 760–771. Rhoads, B.L., Thorn, C.E., 1993. Geomorphology as science: the role of theory. Geomorphology 6, 287–307. Rizzo, D., Dougherty, D., 1994. Characterization of aquifer properties using artificial neural networks — neural kriging. Water Resources Research 30 (2), 483–497. Seibert, J., 2000. Multi-criteria calibration of a conceptual runoff model using a genetic algorithm. Hydrology and Earth Systems Sciences 4, 215–224. Seibert, J., Uhlenbrook, S., Leiundgut, C., Halldin, S., 2000. Multiscale calibration and validation of a conceptual rainfall-runoff model. Physics and Chemistry of the Earth. Part B: Hydrology, Oceans and Atmosphere 25 (1), 59–64. Sharp, M.J., 1988. Surging glaciers: geomorphic effects. Progress in Physical Geography 12, 533–559. Sharpe, D.R., Russell, H.A.J., Pullan, S.E., Pugin, A., 1999. An approach to 3-D environmental geology mapping: a case from the Oak Ridges Moraine, southern Ontario. Abstracts with Programs — Geological Society of America 31 (7), 81. Sharpe, D.R., Russell, H.A.J., Logan, C.E., 2003. The role of expert knowledge in the construction of a 3-D geological model of the Oak Ridges Moraine Region, Canada. Abstracts with Programs — Geological Society of America 35 (6), 76. Shaw, J., Sharpe, D.R., 1987. Drumlin formation by subglacial erosion. Canadian Journal of Earth Sciences 24, 2316–2322. Smalley, I.J., Unwin, D.J., 1968. The formation and shape of drumlins and their distribution and orientation in drumlin fields. Journal of Glaciology 7, 377–390. Smith, M.J., Clark, C.D., 2005. Methods for the visualisation of digital elevation models for landform mapping. Earth Surface Processes and Landforms 30, 885–900. Smith, B., Mark, D.M., 2001. Geographic categories: an ontological investigation. International Journal of Geographical Information Science 15, 591–612. Smith, M.J., Clark, C.D., Wise, S.M., 2001. Mapping glacial lineaments from satellite imagery: an assessment of the problems and development of best procedure. Slovak Geological Magazine 7, 263–274.


J. Napieralski et al. / Earth-Science Reviews 85 (2007) 1–22

Stokes, C.R., Clark, C.D., 2001. Identifying paleo-ice stream tracks using remote sensing. Slovak Geological Magazine 7, 247–262. Stokes, C.R., Clark, C.D., 2003. The Dubawnt lake paleo-ice stream: evidence for dynamic ice sheet behaviour on the Canadian Shield and insights regarding the controls on ice-stream location and vigour. Boreas 32, 263–279. Stroeven, A.P., Fabel, D., Hättestrand, C., Harbor, J., 2002. A relict landscape in the centre of Fennoscandian glaciations: cosmogenic radionuclide evidence of tors preserved through multiple glacial cycles. Geomorphology 44, 145–154. Sung, Q.C., Chen, Y.C., 2004. Self-affinity dimensions of topography and its implications in morphotectonics: an example from Taiwan. Geomorphology 62 (3–4), 181–198. Tarasov, L., Peltier, W.R., 2004. A geophysically constrained large ensemble analysis of the deglacial history of the North American ice-sheet complex. Quaternary Science Reviews 23, 359–388. Treague, J.J., Syverson, K.M., 2002. GIS analysis of northern Wisconsin glacial till units. Abstracts with Programs — Geological Society of America 34 (6), 505. Trenhaile, A.S., 1976. Cirque morphometry in the Canadian Cordillera. Annals of the Association of American Geographers 66, 451–462.

Truffer, M., Echelmeyer, K.A., Harrison, W.D., 2001. Implications of till deformation on glacier dynamics. Journal of Glaciology 47, 123–134. Unwin, D.J., 1973. The distribution and orientation of corries in northern Snowdonia, Wales. Transactions of the Institute of British Geographers 58, 85–97. Walsh, S.J., Butler, D.R., Malanson, G.P., 1998. An overview of scale, pattern, process relationships in geomorphology: a remote sensing and GIS perspective. Geomorphology 21, 183–205. Wise, S.M., 1998. The effect of GIS interpolation errors on the use of digital elevation models in geomorphology. In: Lane, S., Richards, K., Chandler, J. (Eds.), Landform Monitoring. Modelling and Analysis, vol. 99. John Wiley & Sons, Chichester, pp. 139–164. Wright, D.J., Goodchild, M.F., Proctor, J.D., 1997. Demystifying the persistent ambiguity of GIS as “tool” versus “science”. Annals of the Association of American Geographers 87, 346–362. Zelčs, V., 2003. Glaciotectonic map of Latvia. Abstract with Programs — XVI INQUA Congress, Paper No. 39-4. Zevenbergen, L.W., Thorne, C.R., 1987. Quantitative analysis of land surface topography. Earth Surface Processes and Landforms 12, 47–56.