What do we visually focus on in a World Heritage Site? A case study in the Historic Centre of Prague

The study addresses how socio-economic development and modern construction threaten the visual integrity and Outstanding Universal Value (OUV) of World Heritage Sites (WHS). While heritage integrity and authenticity are central to UNESCO criteria, rapid urbanisation—especially tall, visually contrasting buildings—can adversely affect visual quality and potentially lead to delisting. Prior assessments often rely on expert judgment with limited time and few case studies focused on visitors’ visual perception. This paper investigates the Historic Centre of Prague, a WHS meeting criteria (i–vi), to explore visitors’ visual focus patterns across city, street, and building scales using eye-tracking to inform conservation and planning that balance protection and modernisation. The research questions are: (1) What characterises visual concerns for landmark buildings in heritage conservation planning (macro-perspectives)? (2) What are the patterns of visual concerns for street scenes in areas with high visual connectivity (meso-perspectives)? (3) What are the characteristics of visual concerns for typical architecture and sculptures (micro-perspectives)?

The paper situates visual integrity within the UNESCO OUV framework and the challenges posed by urban development, citing cases where visual impacts prompted delisting (e.g., Dresden Elbe Valley) or threats (e.g., Esfahan). Visual Impact Assessment (VIA) and Heritage Impact Assessment (HIA) are highlighted as critical tools. Eye-tracking (video-oculography) is established as a robust method to capture gaze behaviour, fixation locations, and durations for analysing visual attention to landscape elements. Prior works demonstrate eye-tracking utility in assessing visual attention to natural and built features, saliency, and human–environment interactions; and space syntax (axial maps, VGA) to quantify street integration and visual connectivity. Studies indicate that salient features (shape, colour, texture contrast), architectural style, and human activity (crowds) influence visual attention, while crowdsourced imagery and visibility modelling can map appealing views. The paper leverages these methods to examine WHS visual integrity from visitors’ perspectives.

The study integrates a multi-scale approach (macro/meso/micro) combining field survey, space syntax analysis, and eye-tracking. - Field survey: Over 200 photographs were captured on clear days using a Canon 500D from key observation points (KOP) and varied directions/perspectives reflecting visitors’ viewpoints. Images encompassed streets, architecture, sculptures, and river landscapes. Eight associate professors in landscape architecture helped select representative, high-quality images. - Case study site: Historic Centre of Prague (first Czech WHS; criteria i–vi). Planning context includes protected core and buffer zones; zoning for residential, public, and special uses; green spaces and roof landscapes; traffic networks; pollution and flood risk maps; and underground protected areas. Buffer zones preserve important view corridors and skylines. - Space syntax: Axial map and Visibility Graph Analysis (VGA) were produced in Depthmap (UCL) over an area ~1061 m × 802 m to analyse street integration and visual connectivity. Integration values were colour-coded (warmer = higher). Dense sight lines were identified at Parizska Street (Old Town Square), Krizovnicka Street (east of the Faculty of Arts), and the two bridges over the Vltava. - Eye-tracking experiment: 54 participants (24 male, 30 female) with normal or corrected vision viewed field-survey images on a screen; each image was shown for 10 s in random order. Eye movements were recorded with a Tobii TX300 (fixation locations, durations, scanpaths). Visual heatmaps were generated to identify hotspots. Data were analysed in SPSS. Tests were structured across three scales: (1) Macro: seven established panoramic viewpoints in the buffer/core areas; (2) Meso: selected streets from VGA hotspots (Old Town, Malá Strana, Charles Bridge approaches, Old Town Square, courtyards/greenery, pedestrian zones); (3) Micro: façades (e.g., Prague Castle), roofscapes (domes), sculptures (e.g., Cyril and Methodius), and river scenes (Vltava, bridges/boats).

- Macro (city scale): Heatmaps from seven key viewpoints showed that views to principal historic attractions are generally unblocked and preserved by the buffer zone, indicating high authenticity and visual integrity. However, visually prominent modern tall structures (e.g., Žižkov TV Tower) attract strong attention and pose visual concerns. Gaze concentrates on the skyline’s highest points, building tops, and prominent nodes with contrasting colours/styles; greenery and large uniform roof areas attract little attention. - Meso (street scale): Space syntax indicated low integration for most streets, with concentrated visual connectivity at Parizska Street, Krizovnicka Street, and bridge approaches. Eye-tracking on representative scenes revealed: (a) Residential/commercial streets: attention to façade frescoes, shop decorations, and street ends; minimal attention to sky, ground, and stationary vehicles. (b) Old Town Square: the old town hall and surrounding classical buildings dominate attention; faces of nearby people and distant street ends are salient; patterned flooring attracts little attention. (c) Courtyard greenery: gaze prioritises commanding heights (domes, prominent forms) and nearby visitors over the greenery; sky, ground, and same-colour roofs receive little attention; greenery functions effectively as background. (d) Pedestrian zones to Charles Bridge: attention centres on the street end, pedestrians’ faces, and nearby bicycles; ground is neglected; bicycles can disrupt visual quality depending on placement. - Micro (building scale): Façade attention clusters along the façade midline and entrances; mural decorations and garlands are hotspots. Sculpture viewing concentrates on faces. Roofscapes attract attention to complex domes over the sky. Along the Vltava, gaze targets far bridge spans and the highest points on nearby boats; river surface, sky, and greenery receive minimal attention. - Quantitative details: 54 participants; seven macro viewpoints tested; space syntax integration included one red axis (10–11), six yellow (8–9), six green (6–7); remaining axes <6. Findings consistently show complexity and prominence (height/contrast) drive attention, while homogeneous elements (ground, uniform roofs, sky, much greenery) attract little.

Findings across scales directly address the research questions. Macro-level results confirm that buffer zones help maintain visual integrity of key heritage skylines, yet highly salient modern tall structures command attention and can threaten perceived integrity. Meso-level analyses show that street terminations, façade decorations, human faces, and even bicycles become visual focal points, informing crowd and mobility management as part of visual conservation. Micro-level results emphasise the importance of complex, symbolically rich elements (façade midlines, murals, domes, sculpture faces) as key carriers of visual attention and, by extension, perceived heritage value. Implications for planning and VIA/HIA include: restricting tall/high-contrast new buildings within or near WHS; preserving outer contours and skyline integrity; prioritising conservation/maintenance of visually complex heritage features; managing crowd density and bicycle parking to reduce visual clutter; and treating greenery as a flexible, supportive background element while maintaining existing green spaces. The integrated framework demonstrates how eye-tracking combined with space syntax can operationalise visitor-centred visual integrity assessments, guiding design, conservation, and management to protect OUV.

The study contributes a visitor-centred, multi-scale visual assessment of a WHS using eye-tracking and space syntax. Key contributions: (1) Macro: Buffer zones generally maintain unblocked views of major attractions; attention clusters at skyline high points and prominent nodes; tall/high-contrast modern buildings (e.g., Žižkov TV Tower) draw disproportionate attention—new buildings should remain below average height with muted colours. (2) Meso: Visual focus at street ends, façade decorations, human faces, and bicycles suggests managing tourist density and bicycle placement; ground, roofs, and much greenery draw limited attention, supporting greenery’s role as visual background. (3) Micro: Complex artistic features (façade midlines, domes, mural centres, sculpture faces) are consistent hotspots warranting enhanced protection and maintenance. The approach supports VIA/HIA and heritage management plans, with broader applicability to other WHS. Future research should extend beyond vision to multisensory assessment (audio-visual interactions, olfaction), include physiological measures, and leverage wearable/VR-based methods for comprehensive environmental appraisal.

The study focuses solely on visual perception using static images and laboratory eye-tracking, without directly measuring multisensory influences (e.g., soundscapes, smells) or in-situ dynamic experiences. Ground, sky, and greenery assessments rely on still-image stimuli and may differ under varied environmental or seasonal conditions. Future work will incorporate audio-visual interactions, physiological measurements, and wearable/VR techniques for more comprehensive assessments.