Engineered from scratch: 40 years of geospatial intelligence in Flevoland

To mark the 40^th anniversary of the Netherlands’ youngest province, GIM International interviewed two geoinformation specialists who between them span the full arc of Flevoland’s existence: Alex Lucassen and Christiaan de Waal. Having started their careers at the national service that managed the polders before provincial status was formalized, they currently work on behalf of the Province of Flevoland, in today’s era of digital twins and AI-assisted analysis. Their reflections offer a rare inside view of how geodata collection, processing and application have evolved in parallel with the province itself.

There are very few places on Earth where you can say, with complete precision, exactly when the land came into existence. Flevoland, the 12^th and newest province of the Netherlands, is one of them. The ground beneath its cities and farms used to be at the bottom of the IJsselmeer lake (and, for centuries before that, of the shallow inland Zuiderzee sea in the north of the Netherlands, which was closed off and partially reclaimed by the Zuiderzee Works project) until engineers pumped it dry; Southern Flevoland emerged from the water in 1968, Eastern Flevoland in 1957, and Noordoostpolder in 1942. What followed was one of the most extraordinary experiments in deliberate nation-building ever undertaken: a province designed entirely from scratch, surveyed before a single field was ploughed, and precisely planned on the drawing board – right down to the placement of individual tree lines.

Established on 1 January 1986, Flevoland celebrates its 40^th anniversary as an official Dutch province this year. Over the past four decades, a polder born from hydrographic soundings and trigonometric survey stations has grown into a densely networked, digitally informed and still rapidly growing territory of 470,000 people, six municipalities and some of the Netherlands’ most ambitious spatial planning initiatives. Throughout that journey, geospatial data has not just supported decision-making – it has been the very language of the province’s development.

Designing land that did not exist

The story of Flevoland’s geospatial heritage begins well before the province itself. Cornelis Lely, the hydraulic engineer whose surveys and lobbying efforts made the Zuiderzee Works project possible in the 20^th century, understood from the outset that reliable geographic data was the key requirement for everything. His hydrographic surveys of the Zuiderzee, conducted from the 1890s onwards, captured water depths, bottom-sediment composition and tidal behaviour – precisely the information needed to determine whether and how a polder could be made habitable.

The Wieringermeer lake, which was drained in 1930 to create a polder, actually served as a working prototype. Every lesson learned there (about soil settlement, drainage spacing and parcel sizes suited to agricultural mechanization) was carried forward into the design of the Noordoostpolder and then, increasingly, into both flanks of Flevoland. “You can see how Walter Christaller’s Central Place Theory works in practice in the Noordoostpolder,” notes Alex Lucassen, referring to the planned settlement hierarchy of a central town surrounded by satellite villages within cycling distance. “By the time you get to the younger part of Flevoland, that model had already been adjusted – the car had arrived, and some of the planned villages were simply dropped from the design.”

The land-use framework for the Noordoostpolder – including the cable-drain network, the locations of farms, forests, roads and towns – was drawn up even before the polder was fully drained. Soil maps, derived from thousands of bore samples taken both before and during the drainage process, informed key decisions: where agriculture would be viable, where the subsiding peat-soils were better suited to forestry, and how the cable-drain system would manage water levels. This meticulous planning ensured that the polder’s infrastructure and land allocation were optimized from the start. The ancient buried channel of the Eemstroom – a river that once flowed northward and drained into what later became the Zuiderzee – influenced the alignment of primary drainage canals. The channel was invisible to the eye, but detectable in the sediment data. Geoinformation, in this sense, preceded Flevoland’s physical existence. This is a fitting illustration of how geological and hydrological data shaped the polder’s design before the land was even reclaimed.

Measurement-taking in this era was necessarily physical. Theodolites – precise optical instruments requiring stable ground stations – were among the first things placed on the newly exposed polder floors. A replica of one such ‘measuring chair’ still stands in the Kuinderbos forest. The triangulation network provided the geodetic backbone from which all subsequent survey work was referenced. “The philosophy was to start from a basic reference framework and then progressively densify,” says Lucassen. “That principle, which those early surveyors thought through so carefully, is essentially still what we do today. We just use different tools.”

From darkroom to digital plotter

The Rijksdienst voor de IJsselmeerpolders, the national agency responsible for polder management up until Flevoland’s provincial independence, operated what was effectively a vertically integrated geographic information machine. It collected and processed field data, produced maps, managed land cadastre and also planned spatial development, all under one institutional roof. The agricultural surveying department, known internally as the landbouwkundige afdeling, sat at the centre of this web, supplying spatial data to engineers, planners, ecologists and road designers alike.

Map production in the 1970s and the early 1980s was literally a craft. Survey data, computed on punched cards using time on a shared mainframe called SARA in Amsterdam, was output of an automated plotting table which engraved linework onto large red plastic sheets. These sheets were then taken to a darkroom (a restricted space reached through a revolving door) where photographic negatives were being made and paper prints produced. Thematic content was added by hand by cartographers using ecoline ink to shade polygons, each mapping a meticulous watercolour of spatial policy.

The transition to digital was gradual and, by modern standards, quite painstaking. In the early 1980s, the Rijksdienst was among the first Dutch government organizations to bring in computer-aided drafting. A visionary head of section called H.W. van de Zijpp – described by Christiaan Lucassen and Christiaan de Waal as a formative mentor – insisted that digitization was not merely about replicating analogue lines on screen. “He had this idea that when you digitize a line, you should attach intelligence to it,” Lucassen recalls. “You make polygons, you put attribute data inside them. That was already happening in the 1980s.” The system used was an early GRADIS 2000 workstation, and the data it produced was stored on large magnetic disks retrieved by a specialist operator in the computing centre.

Alongside survey works, the Rijksdienst produced the large-scale base mapping – the Grootschalige Basiskaart Nederland (GBKN) – for Eastern and Southern Flevoland. Coordinates were digitized by using a planimeter table, points clicked and recorded, then connected by hand with rotring pens of specified thicknesses before the automated plotter took over for bulk production. The same office ran the quasi-cadastral registration of land parcels, producing official transfer documents – typed in full by hand, accompanied by drawn plans – that were dispatched by a special courier to the land-registry keeper in the city of Zwolle.

New province, new geoinformation challenge

When Flevoland became province on 1 January 1986, the competences were not transferred from the Rijksdienst instantaneously. The old agency’s surveying and mapping functions continued in parallel with the new provincial administration for two to three transitional years. But the fundamental logic of who owned and managed geographic data did change rapidly in that time.

The new provincial government organization initially maintained two surveyors in-house within its Roads & Traffic department. It quickly became clear, however, that the small organization – of barely 300 staff at that time – could not financially sustain a full surveying operation. Instead, the province repositioned itself as a client and coordinator, commissioning field surveys from external specialists while building internal capacity for data processing and spatial analysis. “We don’t have field surveyors anymore,” Lucassen explains. “Now, we are more the director. The municipalities have their own surveying services, and we are the ones who commission, receive and process the data.”

A few years later, the geoinformation function was consolidated into a dedicated bureau: Bureau Kartografie en Grafisch Ontwerp (BKGO), staffed by four specialists covering land surveying, cartography and graphic design. This small team served the entire organization. It was here that some of Flevoland’s most distinctive cartographic work was produced – provincial maps that, according to Lucassen and De Waal, became reference points in not only Dutch cartographic circles but also international ones.

The cartographic philosophy developed at BKGO appeared deceptively simple on the surface, but it was actually astonishingly rigorous beneath, where artistry met the relentless precision of science. Large place-name labels were aligned to invisible horizontal and vertical grids; symbols were designed to sit in visual balance; colour palettes were calibrated, so that background information receded while thematic content remained legible. One set of provincial maps won awards from the Nederlandse Vereniging voor Kartografie. The province’s visual system was later cited as a model by international colleagues. “The complexity of the landscape was expressed in a simple way. That was the strength of it,” reflects De Waal. The logo of the province itself – the kiekendief (marsh harrier) – was designed by the same cartographic hand.

The GIS revolution: from layers to policy integration

Geographic information systems (GIS) arrived in the province of Flevoland’s workflow in the early 1990s, initially running alongside the existing CAD-based cartographic software in a dual-track arrangement. Data managed in the GIS environment was exported into Adobe Illustrator and similar packages for final cartographic refinement. The two streams served different masters: one optimized for analysis, the other for visual communication.

The analytical power of GIS was demonstrated early, and dramatically. An internal presentation showing spatial conflict analysis, where proposed industrial development impinged on nature areas, produced an unexpectedly strong reaction from senior management. “They were a bit alarmed,” Lucassen recalls with a smile. “With one press of a button, you could immediately show where conflicts arose. That was moving too fast for them.” The visual bluntness of binary GIS outputs – a pixel is either in conflict or it is not – required diplomatic handling. Later planning documents deliberately employed gradient shading and soft transitions to communicate the ambiguity that policy genuinely involves.

The province migrated through a succession of GIS platforms over the following decade. The first was Intergraph MGE, a pioneering GIS platform from the 1980s/1990s, designed for large-scale professional use in mapping, utilities and infrastructure. It merged vector and raster analysis, CAD precision and enterprise database integration, becoming a go-to system for engineering and utility sectors. Then came Bentley Microstation, and then an Esri ArcView environment backed by an Oracle database – a configuration that still remains in use today. Unusually, Flevoland and Limburg were the last Dutch provinces to align with the Esri standard that others had adopted. “We visited Limburg, borrowed their database structure one to one, and built from there,” Lucassen notes approvingly. “Somebody had designed something clever – why invent it again from scratch?”

Turning point

A turning point in the organizational positioning of geoinformation came when the team was moved from the IT department into the central policy directorate. This was, Lucassen suggests, a structurally significant decision that actually distinguished Flevoland from many peer organizations. “In other organizations, GIS was seen as a technical tool – a hobby of the IT department. We were positioned inside the policy process. We knew what was going on. We could act as the connector between all the thematic domains.” Geoinformation became, in the phrase Lucassen uses repeatedly, the ‘cement’ bonding together policy fields that would otherwise operate as silos.

The Geopolis project, an early European collaborative effort, produced one of the Netherlands’ first digital map viewers. This tool enabled a user to ‘probe’ the landscape by drawing a polygon and immediately see which thematic layers it intersected. As the polygon moved toward the city of Lelystad, more conflicts lit up; as it shifted to open agricultural land, the screen turned green. This primitive spatial decision-support tool prefigured much of what digital twins now promise. It was presented at European conferences as an example of how cartographic integration could serve integrated spatial policy.

Elevation, archaeology and the third dimension

Without any doubt, one of the datasets that has had the most transformative impact on provincial spatial management is the Actueel Hoogtebestand Nederland (AHN). This national airborne laser scanning archive provides centimetre-resolution elevation models of the entire country. For Flevoland, a province whose very existence depends on controlled water levels and whose soft peat soils are measurably subsiding, elevation data is not just an analytical luxury but an operational necessity.

The first time the AHN data was visualized with hill-shading, it revealed the hidden topographic memory of the pre-polder landscape: the sinuous trace of the ancient Eemstroom riverbed, the micro-relief of former tidal flats, the subtle outlines of mediaeval settlements drowned by the Zuiderzee centuries before the polders were created. “The archaeologist was completely enthusiastic,” De Waal remembers. “He said, this is of inestimable value. You could see exactly where the Eemstroom runs.” Researcher Yftinus van Popta subsequently used Lidar-derived AHN data to locate submerged mediaeval villages in Flevoland, refining hypotheses about settlement locations which had been developed from historical documents.

Modelling soil subsidence

The same elevation data now feeds into long-range scenario modelling for soil subsidence – a pressing challenge, as Flevoland’s peat-based soils continue to compress and oxidize following drainage. Comparing historic survey benchmarks with current AHN measurements is allowing the province to quantify subsidence rates by area and to model future implications for infrastructure, water management and building foundations. The original survey data from the polder-construction era, painstakingly collected on punched cards and magnetic tapes, has regained contemporary relevance as a historical baseline.

Three-dimensional visualization has also reshaped the way in which spatial planning is communicated to decision-makers. Whereas a two-dimensional conflict map once had to be explained with lengthy annotations, a 3D model allows policymakers to ‘walk through’ a proposed development, experiencing its scale and implications in a way that a flat diagram cannot convey. “You take them along through the landscape,” De Waal comments. “And then all of a sudden they say, well, we’ve decided something – we need to look at this again.” The province currently uses Tygron, Esri Web Scenes and Erdas Imagine (a raster-based platform by Hexagon) for different 3D applications, deliberately maintaining platform plurality to avoid vendor lock-in.

The province as data broker

Flevoland’s approach to geodata publication reflects both its scale advantages and its philosophical commitments. With six municipalities, one water authority, one Rijkswaterstaat district and ProRail as primary partners, the provincial network is compact enough to enable genuine coordination rather than mere consultation. Land use data, which includes confidential municipal housing development plans, is aggregated in a shared platform that allows the province to track regional development trajectories and align provincial policy accordingly.

Public data dissemination operates through a geo-server environment running open-source software, exposing datasets via Open Geospatial Consortium (OGC) web services as well as newer API formats that allow developers and programmers to query and consume provincial data directly. A public map portal provides thematic viewers accessible to all. The province participates in the national INSPIRE reporting framework through the inter-provincial coordination body IPO, and is actively working towards compliance with INSPIRE Plus, the EU’s high-value dataset requirements.

A known tension in the current setup is the fragmentation between the public map portal and the geo-register, which both operate as separate environments. The province is therefore working towards integration: a unified platform creating a single interface where users – whether specialists or citizens – could search across all provincial spatial data, potentially assisted by AI-driven search. “We see ourselves as a kind of data broker,”  explains De Waal. “We know where everything is, literally. We bring it together. And increasingly, we want to make that access seamless.”

The Basisregistratie Grootschalige Topografie (BGT), the Netherlands’ national large-scale topographic base register, is maintained in Flevoland through a formal contractual arrangement in which the six municipalities act as surveyors and the province as commissioning client. Periodic coordination meetings, held three to four times a year, align quality standards and resolve any discrepancies. The arrangement illustrates the collaborative model that Flevoland developed: the province sets standards and holds data, the municipalities execute fieldwork, and the resulting dataset serves all parties.

Towards a digital twin

The concept of the digital twin, as a dynamic computational model of a physical territory that can be used to simulate and evaluate policy scenarios, correlates with Flevoland’s origins with unusual precision. A province that was designed from scratch on a drawing board, whose spatial layout was the product of deliberate computation, is in a sense already an analogue twin of itself. Flevoland’s aspiration with a digital twin is to close the loop: to maintain a continuously updated computational representation of the province that can be probed about the consequences of future decisions.

Current work is proceeding through a series of use cases under the Digital Model Infrastructure (DMI) programme. An already operational time-lapse application visualizes land-use change across the province over recent decades. Pilot projects in the town of Dronten have tested the Tygron environment for neighbourhood-level scenario planning, with the model subsequently shared with other organizations as an open framework. The province’s largest city by far, Almere – which faces severe urban heat island and flooding challenges – is engaged in a separate digital twin project to model precipitation response at city scale.

A Digital Experience Center, which is planned for completion in 2027 in the new provincial offices, will provide a purpose-built physical space where large-format digital displays present digital twin scenarios to politicians, citizens and partner organizations. The intent is explicitly collaborative: it should serve as a place where municipalities, the water authority and civil society organizations can gather around shared scenario-based evidence to discuss complex spatial challenges. “Seeing is believing, often,” Lucassen observes. “But you also have to make sure people understand that these are models. They come with uncertainty. At the provincial scale, we deliberately stay one abstraction level higher.”

Computational capacity remains a practical constraint. Running 3D scenario calculations for an entire province requires processing resources that exceed what the organization holds in-house. The organization has used Tygron’s cloud computing infrastructure for province-wide exercises, working within a European data sovereignty framework by specifying that computed resources remain within EU boundaries. The team is also experimenting with point cloud data – currently applied to bridge and viaduct surveys at small scales – anticipating that decreasing costs and increasing processing capacity will make province-wide Lidar analysis routine within the planning horizon.

AI and imagery analysis

Artificial intelligence tools are being piloted across several application areas. The province holds an annual ‘development week’ in which the geospatial team suspends normal operations to experiment with emerging technologies – a deliberate innovation budget which has management support. Recent outputs include: AI-assisted classification of solar panels and solar farms from aerial imagery, generating provincial-scale estimates of realized and potential renewable energy capacity; detection of asbestos-containing roofing from spectral imagery data, supporting a provincial programme to accelerate asbestos removal; and network analysis using both open-source QGIS tools and Esri’s ArcGIS suite in order to compare performance characteristics.

Spatial analysis of the energy transition is one example of how these tools are applied to complex policy problems. Gas pipeline networks serving farming operations were overlaid with agricultural enterprise distributions to identify potential hub locations for the activity of ‘green’ gas production from manure. The locations must have sufficient concentrations of livestock operations to justify pipeline upgrades and biodigester investment. “Through AI techniques, we can generate that kind of assessment much faster than before,” De Waal notes. “It’s still an experiment, limited in scope, but it’s something we’re exploring.”

The team maintains deliberate openness about platform selection, running parallel comparisons of commercial and open-source tools before committing to any particular solution for operational use. This philosophy of “we don’t lock ourselves in” extends to data architecture; where possible, services are exposed through open standards that allow downstream systems to be replaced without data loss. It is a policy shaped partly by experience of past migrations, and partly by a commitment to data sovereignty and interoperability with the growing European spatial data infrastructure.

What Flevoland teaches the geospatial world

So what can the story of Flevoland teach geospatial professionals working in more conventionally evolved territories? Several themes go beyond the specific circumstances of the polders. The first is the enduring value of geodetic foundations. The triangulation network laid out on bare polder mud in the 1950s and 1960s established a reference framework whose logic – start from reliable base points, densify progressively – remains unchanged in the GPS era. Data collected then, with instruments that seem primitive today, has retained value as a historical baseline against which contemporary measurements reveal half a century of landscape change. The investment in rigorous ground control is paying back with compound interest across generations.

The second theme is organizational positioning. The decision to locate the geoinformation function within the policy directorate rather than the IT department had long-lasting consequences. It meant spatial analysts were present at the earliest stages of policy development, able to contribute analytical input before decisions crystallized rather than only illustrating conclusions already reached. It also meant that the team developed fluency in policy language, learning to present probabilistic and gradient results in ways which served deliberation rather than excluding it.

The third is the power of collaborative scale. Flevoland’s compactness – six municipalities, one water authority, short lines of communication, relationships built across careers – has enabled a degree of data integration and joint working that larger, more fragmented regions struggle to achieve. The BGT maintenance arrangement, the shared housing development monitoring platform, the joint digital twin pilots… they all depend on the same small network of professionals that have worked together for years. “We can adapt quickly,” Lucassen says. “Short lines. We know each other.” This is a structural advantage that cannot be entirely replicated by technology alone.

The fourth is the importance of a pioneering culture. Flevoland was built by people who were, by definition, doing something that had never been done before. That spirit of experimentation, the annual development week, the willingness to test multiple platforms in parallel, the comfort with ‘intelligent failure’, it all appears to be a genuine institutional inheritance from the Rijksdienst era. “People with a geo background always stay open to new developments,” Lucassen reflects. “They just go along with it. They know that if they want to stay relevant, they have to keep up with the pace of change.”

Conclusion

Four decades ago, in 1986, the Dutch state formally recognized that what had been built on the former Zuiderzee sea floor was not merely reclaimed agricultural land, but a genuine province with its own identity, institutions and future. The geoinformation professionals who have been serving that province throughout those 40 years have been more than technicians providing maps on demand. They have been, in the deepest sense, the memory-keepers and intelligence-providers of a territory that has no ancient cadastral record, no centuries-old survey tradition and no geological landscape predating human intention.

From hydrographic soundings in a shallow sea to AI-assisted solar panel classification from aircraft, and from punched-card polygons on shared mainframes to cloud-computed digital twin scenarios accessible on a wall of screens, the arc of Flevoland’s geospatial journey mirrors – and in some ways anticipates – the broader trajectory of the discipline. A province drawn before it even existed is now modelling its future in three dimensions.

The drawing board has become a digital twin. But the underlying imperative – to understand the land before you act upon it, and to share that understanding as openly as possible – has not changed at all.

Acknowledgements

This article is based on an interview conducted with two senior geoinformation specialists (Alex Lucassen and Christiaan de Waal) working on behalf of the Province of Flevoland. Additional context has been drawn from published documentation by Batavia Land (formerly the archive of the Rijksdienst voor de IJsselmeerpolders) and the Waterschap Zuiderzeeland water authority, amongst other sources.

Alex Lucassen is a geoinformation advisor at the Province of Flevoland in the Netherlands. He has built an extensive career in land and spatial development, beginning at the Rijksdienst voor de IJsselmeerpolders (RIJP) – the Dutch government agency responsible for the development of newly reclaimed land in the drained IJsselmeer polders – where he worked from November 1981 to December 1987. Since January 1988, he has been with the Province of Flevoland, bringing over 38 years of dedicated expertise in geoinformation to the organization.

Christiaan de Waal is a geospecialist and graphic designer with over 20 years of experience at the Province of Flevoland in the Netherlands. He creates clear and functional visualizations. He is the 'guardian' of the Flevoland cartography, including the provincial map. His principle is simple: "less is more". This way, complex information becomes easier to understand.

Pioneering cartography

Jan Schilder (1950, Kampen, the Netherlands) is a pioneering cartographer and graphic designer, celebrated as the fundeer of Flevocartography. After starting his career in 1967 at the Rijksdienst voor de IJsselmeerpolders (RIJP), he played a key role in visualizing and shaping the Netherlands’ newly reclaimed polders. As head of the Province of Flevoland’s Bureau of Cartography and Graphic Design (from 1987 onwards), Schilder redefined mapmaking by merging topographic precision with bold artistic expression, drawing inspiration from Flevoland’s landscapes. His distinctive style gained global acclaim, and his maps adorn public spaces worldwide. In 2015, Schilder was knighted in the Order of Orange-Nassau for his transformative impact on cartography.