Model documentation

Download the Word document

PORTFOLIO / TECHNICAL RECORD · 2026

A paired classroom resource combining a 10 m DEM, custom cartography, urban land-cover data, clickable settlements and a simplified Three.js pupil viewer.

Interactive model and cartography: Mr Lane · Year 5 geography · Mount Vesuvius, Campania

1. Project brief

The Vesuvius model was developed as the human-geography counterpart to the finished Mount Etna resource. The two models use the same broad visual language and geographic scale, but answer different parts of the enquiry: Etna foregrounds farming; Vesuvius foregrounds settlement and urbanisation around an active volcano.

The main teaching context is a Year 5 lesson on why people live near volcanoes. Vesuvius is especially useful because dense settlement reaches close to the volcano on several sides. Rather than presenting this only as a statement, the model allows pupils to rotate the terrain, switch built-up land on and off, and click individual settlements to connect mapped patterns with real places.

Educational aims

Show the relationship between relief, the Bay of Naples and dense settlement around Vesuvius.

Keep the base terrain readable first, then add human geography deliberately through an urban overlay.

Use clickable settlements as simple, pupil-friendly gateways to local geographical information.

Pair directly with the Etna model so pupils can compare two active volcanoes with different surrounding land uses.

Work reliably on classroom computers, iPads and phones without exposing pupils to the stock technical Qgis2threejs interface.

Final pupil experience

The viewer opens with the terrain and the Vesuvius summit marker visible. The built-up-area layer and town spheres are hidden by default. When pupils switch on Built-up areas, both the CORINE urban polygons and the clickable settlement spheres appear together. Reset view restores the original camera, switches the urban layer and town spheres off, keeps the Vesuvius summit marker visible, closes open information cards and hides the legend.

2. Core spatial specification

The model was deliberately kept in EPSG:32632 to preserve a clean TINITALY workflow and consistency with the Etna model. Reprojecting the terrain to UTM zone 33 introduced rotated/no-data corners; at this model scale, remaining distortion in zone 32 was accepted as the better practical compromise.

Extent choice

The extent was moved deliberately south and west so that the Vesuvius summit would not lie on the seam of the planned 2 × 2 physical print. The final footprint also keeps Naples, Pompeii, the Bay of Naples, the Sorrento Peninsula and Capri in a useful comparative frame.

3. Terrain, sea and cartography

Terrain construction

As with Etna, a continuous terrain surface was needed rather than a land-only DEM. The working Vesuvius raster therefore retained real land elevations while assigning offshore cells a constant elevation of 0 m. This produces a flat sea surface without false vertical cliffs at the coastline.

Clip the full terrain to the exact final model extent.

Create or retain a 0 m sea surface using the same extent, CRS and raster resolution.

Ensure valid land elevations overwrite the constant-zero raster.

Verify that offshore cells read 0 m and land cells retain real elevation values.

Use the merged/clipped raster as the terrain source in Qgis2threejs.

Hypsometric colour ramp

The colour ramp was tuned to stay within the same visual family as Etna while fitting Vesuvius’s much lower summit elevation. A grey summit was tested and rejected because it made the mountain appear too brown-grey and visually disconnected from the paired Etna model.

Hillshade and sea

A duplicate DEM layer was rendered as hillshade above the coloured terrain and blended using Multiply. An additional NoData value of 0 was set on the hillshade copy so that the sea remained the exact Etna blue rather than being greyed by relief shading.

Base-map content

Principal settlements shown with dark-blue circular symbols; Naples given greater prominence.

Hand-traced principal roads styled with a thin pink-red line and pale casing to match Etna.

The summit access road was retained but drawn more lightly/thinly than the main roads.

Mount Vesuvius labelled in bold italic; Ancient Pompeii marked separately as a heritage site.

Gulf of Naples labelled over the sea; a second Tyrrhenian/Mediterranean label was deliberately omitted to avoid clutter.

North arrow, scale bar and a vertical “Height above sea level (m)” key retained for consistency with Etna.

Texture and print-layout proportions

A white strip initially appeared when the exported texture was draped over the terrain. The PNG itself was complete; the problem was an aspect-ratio mismatch. The DEM extent ratio is 49,940 ÷ 52,300 = 0.95488, while the first PNG was proportionally wider. The layout was corrected to the same physical ratio as the intended 3D print: 380.00 mm × 397.96 mm, with the map item filling the page exactly.

4. Urban overlay and settlement interaction

CORINE urban land

CORINE Land Cover 2018 supplied the thematic urban layer. The source data were filtered in EPSG:3035, reprojected to EPSG:32632 and geometrically clipped to the exact Vesuvius model extent. The layer contains the two pupil-relevant urban-fabric classes below.

The urban polygons are exported as Qgis2threejs Overlay geometry, relative to the actual Vesuvius DEM, with an altitude offset of about 20 m. This avoids z-fighting with the terrain surface. The final styling uses a muted urban tone with no border and is hidden on first load.

Settlement click targets

Locality polygons were tested but abandoned because complex polygon triangulation was unreliable in Qgis2threejs. The robust solution was to keep the urban polygons as the visible thematic layer and use settlement point features as separate click targets. These are exported as semi-transparent spheres above the terrain.

Main clickable settlements include Naples, Portici, Ercolano, Torre del Greco, Torre Annunziata, Pompei, Castellammare di Stabia, Somma Vesuviana, Ottaviano, San Giuseppe Vesuviano, Nola, Sarno and Sorrento. The web viewer stores the explanatory text in JavaScript rather than bloating the QGIS attributes.

Mount Vesuvius summit marker

A separate Vesuvius sphere remains visible when the model first loads. It acts as a persistent volcano click target and is not tied to the urban toggle. Its information card explains that Vesuvius is active but currently quiescent and last erupted in 1944.

5. Pupil-facing web viewer

The stock Qgis2threejs controls were replaced with a simplified interface matching the Etna resource. The technical layer tree and raw query output are hidden from pupils; interaction is reduced to a small set of meaningful classroom controls.

Show built-up areas toggle controls both the CORINE urban overlay and town spheres.

Urban layer and town spheres are off on first load; the Vesuvius sphere remains visible.

Clickable town spheres open locality-specific cards with short geographical explanations and a single “Look closely” question.

Clicking the urban overlay opens a generic built-up-area card rather than technical feature attributes.

Clicking the Vesuvius sphere opens a volcano-specific card.

Reset view restores camera, urban toggle, legend and open-card state while retaining the volcano marker.

The interface is responsive for desktop, iPad and phone screens.

A compact Credits panel records data, image and software attribution.

6. Physical model specification

The physical Vesuvius model is designed to match Etna’s broad horizontal scale rather than simply filling the printer bed independently. The values below are the target geometry derived from the final extent. They should be treated as the intended production specification until the final Vesuvius STL/print is measured and verified.

The extent was chosen specifically so the summit remains comfortably within one tile rather than being split by the central seam. Final STL dimensions, terrain height, slicer settings, filament and print time should be added to this record after the physical Vesuvius print has been produced and measured.

7. Problems solved during development

Wrong UTM zone looked cleaner in theory but worse in practice — Reprojecting to EPSG:32633 produced rotated/no-data corners. The model remained in EPSG:32632 for a cleaner raster workflow and direct consistency with Etna.

Texture left a white band on the terrain — The map export had the wrong aspect ratio. Matching the print layout exactly to the DEM extent ratio removed the uncovered strip.

Hillshade greyed the sea — Setting 0 as an additional NoData value on the hillshade layer prevented shading from affecting the flat 0 m sea surface.

Generated road data did not look right — The automatically produced road layer was rejected and principal roads were hand-traced for cartographic control and consistency with Etna.

Nearest-place joins produced misleading “Naples” areas — CORINE polygons can be large and connected, so attaching a single nearest settlement to each polygon was geographically unhelpful.

Voronoi and locality polygons added complexity without improving the lesson — Voronoi territories were arbitrary and ISTAT locality polygons proved unreliable in Qgis2threejs. The final design separates visible urban cover from simple point-based settlement interaction.

Stock Qgis2threejs interaction was too technical — The exported HTML/CSS/JavaScript was simplified into a pupil-facing interface with linked thematic controls and custom information cards.

8. Outcome, reproducibility and portfolio value

The Vesuvius resource complements Etna rather than duplicating it. Etna asks pupils to notice how people use fertile volcanic slopes; Vesuvius makes the density of settlement around an active volcano immediately visible. Together they provide a stronger comparative enquiry than either model alone.

What the project demonstrates

Raster processing and practical CRS judgement using a high-resolution national DEM.

Accurate control of scene extent, aspect ratio and texture draping.

Cartographic simplification for a primary-school audience rather than maximal GIS detail.

Vector filtering, reprojection and geometric clipping of CORINE land-cover data.

3D overlay design that avoids terrain z-fighting.

A robust alternative to complex locality polygons through lightweight point-based interaction.

Front-end modification of an exported Qgis2threejs scene for a specific classroom workflow.

A deliberately paired design language across two volcano case studies.

Reproducibility checklist

Use the exact extent: 924380, 4500680 : 974320, 4552980 in EPSG:32632.

Keep Fix aspect ratio to 1:1 switched off in Qgis2threejs.

Use the continuous DEM with offshore elevation at 0 m.

Keep the map/texture at the same 49,940 : 52,300 aspect ratio as the terrain.

Use CORINE 111/112 for the urban overlay and raise it approximately 20 m above terrain.

Keep town spheres hidden until Built-up areas is turned on; keep the Vesuvius sphere visible from load.

Do not reintroduce locality polygons unless there is a specific new teaching reason to do so.

Credits and data sources

Interactive model and cartography: Mr Lane, 2026.

Terrain: TINITALY 1.1, © INGV, CC BY 4.0. Tarquini, S., Isola, I., Favalli, M., Battistini, A. & Dotta, G. (2023), TINITALY, a digital elevation model of Italy with a 10 metres cell size (Version 1.1). DOI: 10.13127/tinitaly/1.1.

Land cover: CORINE Land Cover 2018, Copernicus Land Monitoring Service / ISPRA.

Settlement locations: point locations prepared for the project from GeoNames-based data; educational text added in the web viewer.

Photography: Wikimedia Commons / Wikipedia lead images where used; individual source and licence information should remain visible in the final viewer credits.

Software: QGIS, Qgis2threejs, Three.js, DEMto3D and Bambu Studio.