From viewsheds to viewscapes: Trends in landscape visibility and visual quality research

Highlights

• First multi-disciplinary systematic review of visibility & visual quality research.

• VVQ research increased 21-fold over past two decades.

• Advances in GIS allow more efficient, accurate visibility models.

• Bare-earth elevation models still dominate VVQ analysis despite limitations.

• Present 4-step framework for reporting + conceptual guidelines for future research.

Abstract

The study of visibility and visual quality (VVQ) spans scientific disciplines, methods, frameworks and eras. Recent advances in line-of-sight computation and geographic information systems (GIS) have propelled VVQ research into the realm of high performance computing via a cache of geospatial tools accessible to a broad range of research disciplines. However, in the disciplines that use VVQ analysis most (archaeology, architecture, geosciences and planning), methods and terminology can vary markedly, which may encumber interdisciplinary progress. A multidisciplinary systematic review of past VVQ research is timely to assess past efforts and effectively advance the field. In this study, we summarize the state of VVQ research in a systematic review of peer-reviewed publications spanning the past two decades. Our search yielded 528 total studies, 176 of which we reviewed in depth. VVQ analysis in peer-reviewed research increased 21-fold in the last 20 years, applied primarily in archaeology and natural resources research. We found that methods, tools and study designs varied across disciplines and scales. Research disproportionately represented the Global North and primarily employed medium resolution bare-earth elevation models, despite their known limitations. We propose a framework for standardized reporting of methods that emphasizes cross-disciplinary collaboration to propel visibility research into the future.

Introduction

Visibility and visual quality analysis is a critical aspect of human-environment interaction research. An observer’s visual field is fundamental to the formation of spatial preferences (Nijhuis et al. 2011) and affects almost every aspect of human-environment experience on a vast range of scales, from internal emotions (Millar et al., 2021) to tourism economics (Schirpke, Timmermann, Tappeiner, & Tasser, 2016). Visibility and visual quality—which we will refer to as VVQ—analysis can be traced back more than half a century in multiple scientific domains, where it has become common practice in assessing and understanding visual experiences (Nijhuis et al. 2011). Given the growing accessibility and increasingly widespread use of VVQ analysis across a range of disciplines, there is an emerging need for cross-disciplinary understanding of current progress and future directions for the field.

VVQ analysis consists of calculating spatial models of what geographical areas can be seen from a set point, known as a viewshed (Fig. 1). The term “viewshed” was first published in 1967 by surveyor and landscape architect Clifford Tandy (Tandy, 1967). In 1968, computer scientists at the U.S. Forest Service developed a program called VIEWIT (Amidon & Elsner, 1968) that computed “seen area” based on gridded elevation cells. While VIEWIT’s documentation doesn’t use the terms “viewshed,” “raster,” or “GIS,” it was a harbinger of VVQ research to come. The use of geospatial computation in modeling visibility has proliferated in recent decades, with advances in Geographic Information Systems (GIS) making visibility modeling accessible by operationalizing viewshed algorithms with easy-to-use and widely distributed software (ESRI., 2021, GRASSGIS, 2021). Historically, calculating line-of-sight for VVQ analysis has been a computationally intensive process (Zhao, Padmanabhan, & Wang, 2013). Researchers continue to develop new algorithms, leveraging high performance computing and GPU processing to reach new heights in computational efficiency and accuracy (Zhao et al., 2013, Zhu et al., 2019). Thus, GIS-based visibility models have become a mainstay of human-environment interaction research such as landscape planning (Anderson and Rex, 2019, Inglis and Vukomanovic, 2020), architecture (Rød and van der Meer, 2009, Weitkamp, 2011), archaeology (Garcia-Moreno, 2013, Van Dyke et al., 2016) and natural resources (Chamberlain and Meitner, 2013, Depellegrin, 2016, Aben et al., 2018).

VVQ analysis is applied with a wide variety of standards and practices, owing to its multidisciplinary applications. Various vocabulary terms represent overlapping concepts: for example, “isovist,” “visualscape,” “viewshed” and “line-of-sight”, are scattered throughout the literature, which could lead to disparate interpretations (Table 1). For example, the term “viewshed” often refers to the area that can be seen from an observer point, but some studies interpret viewshed area as the quality of “openness” (Wilson et al., 2008, Weitkamp et al., 2011). More recently, the term “viewscape” aims to characterize not just what can be seen, but how humans visually connect to their surrounding 3-dimensional terrain and built environment (Vukomanovic, Singh, Petrasova, & Vogler, 2018). While perhaps consistent in their respective fields, these discrepancies in terminology may encumber cross-disciplinary understanding and promote advancements in VVQ research that remain in disciplinary silos.

Interdisciplinary inquiry is integral to the geospatial sciences (Gilbert, 1909, Baerwald, 2010) and VVQ research exemplifies this synergy between disciplines. Questions about visibility and visual quality often involve phenomena that cross disciplines, and that can benefit from cross-fertilization of ideas, technology, frameworks and insight (Baerwald, 2010). Given discrepancies in terminology (Table 1), and the rapid advancement of geospatial tools and technology—such as GPU-based parallel processing (Osterman, Benedičič, & Ritoša, 2014), lidar-derived surface models (Vukomanovic et al., 2018) and a glut of spatial big data (Lee & Kang, 2015) —the lack of recent and multidisciplinary reviews represents a mounting barrier in VVQ research. Now is the opportune moment for a literature review that takes stock of the state of the field, highlights research opportunities that leverage emerging geospatial technology, and serves as a launchpad for novel and interdisciplinary applications.

Past reviews of VVQ literature are either decades old or focused on a specific niche of VVQ analysis. A qualitative literature review in 2003 explored the use of GIS-based viewshed analysis in archaeology, where researchers use visibility models to locate historic sites and confirm theoretical suppositions (Lake, Woodman, & Mithen, 1998). Another qualitative review evaluated the influence of GIS-based visibility analysis across disciplines (Bishop, 2003). However, the technological advances in GIS and computing since this review suggest a timely opportunity to examine the growing field of VVQ research through a multidisciplinary lens. In 2018, researchers reviewed the usage of VVQ analysis in wildlife ecology research, encouraging future ecology researchers to consider viewsheds as a rich source of ecological information about the spatial patterns and behaviors of animals (Aben et al., 2018). A more recent literature review examined studies related to the visual assessment of landscapes, however it was limited to the journals Landscape and Urban Planning, Urban Ecology and Landscape Planning (Gobster, Ribe, & Palmer, 2019).The authors called for more systematic reviews with expanded definitions of what the visual quality research field entails beyond visual assessment. Our study directly builds on these past reviews and recommendations by conducting the first known multidisciplinary, systematic review of visibility and visual quality research.

In this study, we systematically reviewed and analyzed a subset of peer-reviewed research papers published from 2000 to 2019 to ask the following questions: 1) What are the geographic and temporal trends in VVQ research, and how do they vary across scientific discipline?, 2) What types of computational resources and algorithms are visibility researchers using?, and 3) How do different disciplines approach characterizing the visual quality and human dimensions of viewsheds? We describe the systematic review methods in section 2, and share descriptive statistics and visualizations of quantitative analysis results in section 3. In section 4, we elaborate on these results with a discussion of the papers reviewed and the implications of the trends and applications we observed. Finally, in section 5, we synthesize these concepts into a framework of suggested steps for VVQ researchers to consider in future studies.

The visual field is the primary way humans connect to our surroundings (Kandel et al., 2000), so visibility analysis could be a common lens through which to study human-landscape connection. Given advancements in geospatial technology and growing data availability, VVQ research has the potential to have far-reaching impacts in scientific research beyond architecture, archaeology and natural resources, and shape growing interdisciplinary research across human-environment systems. This study aims to promote insight into past VVQ research in order to empower researchers to bridge disciplinary divides and propel visibility analysis into the future.