Identifying and characterising urban land use transformation processes using cadastral parcel geometries: A case study of Spanish cities

Abstract

Cadastral vector data has become particularly important in urban studies due to its ability to accurately represent different urban realities, such as vulnerability assessment, densification processes and the development of simulation models. This paper proposes a framework for analysing changes in cadastral parcels in Spanish cities, identifying transformation process involving geometric and land use changes. Land use maps were obtained for two dates, one in 2022 and another in 2024, covering 48 Spanish cities and consisting of six categories: single-family residential, multi-family residential, industrial, commercial and office, others and undeveloped. The framework combines thematic and spatial analyses, intersection operations, graph construction, and geometric characterisation of changes through different spatial metrics of parcels. The results revealed three main processes affecting cadastral parcels: Aggregating, Subdividing and Non-impactful. These processes mainly affected on undeveloped land with varying geometric characteristics. Large parcels with complex shapes were transformed into simple and elongated parcels through Subdividing processes, while average-sized parcels with simple shapes were transformed into compact, more complex parcels through Aggregating processes. These results provide insights into how urban spaces are evolving not only in terms of land use change, but also through transformations in parcel geometry. This opens several future research lines grounded on the monitoring of cadastral data, such as urban densification studies, policy impact assessment and urban growth simulation models coupled with parcel transformation algorithms.

Keywords

urban land use change cadastral data parcel geometry transformation process

Introduction

The sustainable development goals require a deep understanding of the complex dynamics that define cities (United Nations Human Settlements Programme, 2022). Land Use and Land Cover Change (LULCC) remains a primary focus of diverse research (Benhammou et al., 2022; Hussain et al., 2024; Sun et al., 2023), with literature ranging from the mapping of general long-term metropolitan patterns (Dadashpoor et al., 2019), to the utilisation of bottom-up morphological approaches to identify building-level growth patterns (Perez et al., 2021).

These studies are typically conducted using either (i) satellite imagery and remote sensing data (Wang et al., 2023) or (ii) polygon/line vector-based datasets (Yao et al., 2022). The first has a high capability to be conducted globally and in remote areas efficiently with no need for extensive ground surveys (Verburg et al., 2011). Due to the growing availability and update frequency of satellite imagery, they allow for a temporal evolution analysis (Herold et al., 2005). However, they also present some limitations, mainly related to the usage of pixel-based representation whose spatial resolution and precision must be carefully handled (Ez-zahouani et al., 2023).

On the other hand, studies that use highly detailed vector-based data benefit from a more accurate representation of urban realities. As urban areas have evolved, so have their dynamics, which are currently marked by several overlapping processes that do not fit well within a single, linear model of core-periphery expansion (Pradhan et al., 2025). Edge expansion often combines with redevelopment and gradual densification within the existing fabric, while leapfrogging incorporates discontinuities and fragmentation to growth patterns (Zhang et al., 2017). These changes can also produce spillover effects, where development pressures and their impacts extend beyond municipal limits and influence nearby areas (Kim, 2013). At the same time, growth can coexist with urban shrinkage, that reduces the sustainability and efficiency of the urban fabric (Haase et al., 2014). In these settings, urban function loss may appear through urban perforation, as gaps and vacant parcels accumulate within built-up areas (Usui, 2025), or through building abandonment and underuse, where structures remain but are no longer in active use (Caramaschi and Chiodelli, 2023). Thus, these complex dynamics are best addressed through bottom-up approaches, which are greatly strengthened by detailed vector-based data, enabling fine-grained changes in land use and form to be tracked over time and linked to broader urban trajectories (Fleischmann et al., 2022).

In this context, cadastral data offer rich and precise land parcel information that has recently increased in usage and reliance thanks to greater availability and to general enhanced hardware capabilities (Molinero-Parejo et al., 2024). Moreover, cadastral data are updated regularly to reflect changes in land ownership, including changes in urban uses and parcel geometries, ensuring that the information remains current and reliable. Hence, by comparing historical and current datasets, it is possible to identify trends in urban growth, land use changes and infrastructure development, shedding light on long-term patterns influenced by economic, policy and social shifts. Some applications of this data include urban vulnerability (Sánchez Ondoño and Prada Trigo, 2023), planning (Dalmau et al., 2014), sprawl (Mansour et al., 2024), densification processes (Domingo et al., 2023), land use change simulation (Barreira-González et al., 2019) and risk management (Grădinaru et al., 2023). However, few studies focus on the changes in urban land uses from a parcel geometry transformation perspective.

In this regard, identifying and understanding transformations processes affecting urban parcel geometry might be valuable for diverse urban studies, such as vertical growth pattern studies (Huang et al., 2023), accessibility studies (Kane et al., 2017) and Vector-based Cellular Automata (VCA) models. The latter specifically incorporate cadastral data as a spatial representation unit (Molinero-Parejo et al., 2023) and have successfully addressed land use change simulation (Stevens et al., 2007), but incorporating changes in parcel geometry is still at a very early stage (Yao et al., 2017). Algorithms aimed at assessing cadastral parcel geometry transformations that have been developed recently (Colomb et al., 2022; Dahal and Chow, 2014; Wickramasuriya et al., 2011) might benefit from the proposed analysis, incorporating knowledge derived from parcel geometry transformation analysis.

Within this framework, this study is focused on analysing geometry transformation in urban parcels, using Spanish urban areas as a case study. These urban areas represent an excellent case laboratory for two key reasons: first, they have a history of fast urban growth in recent decades, economic downturns and recovery, which was exacerbated before the 2008 housing crisis and is now resurging due to high demand and low supply in the real estate market (Cardesín Díaz and Araujo, 2017; Villar Navascués et al., 2023). Second, Spanish cadastral data is highly detailed and fully accessible to users. It also can handle large-scale downloads, which, when combined with the appropriate processing tools, enables the coverage of almost the entire country.

Consequently, the main objective of this study is to address the existing gap in the analysis of changes to parcel geometry by developing an analytical framework to detect, map, and characterise cadastral parcel transformations between two dates. This framework was tested for Spanish cities between 2022 and 2024 and allowed us to address the following research questions (RQs):

• RQ1: How geometry transformations undergone by cadastral parcel can be identified, classified and mapped?

• RQ2: How can these transformations be characterised through changes in cadastral parcel morphology (e.g. shape, size, and configuration) and associated land use?

• RQ3: Based on this characterisation, are there systemic differences between parcels targeted by and resulting from a transformation, and those that remain unchanged?

The paper is structured as follows: Section 2 presents the case study and cadastral data used. Section 3 describes the methodology employed. Section 4 presents the results and Section 5 discusses them, highlighting main limitations and future research lines. Finally, Section 6 provides the conclusions.

Datasets and study area

This paper studies 48 of the 52 provincial capital cities and autonomous cities of Spain (excluding Vitoria-Gasteiz, San Sebastián, Bilbao and Pamplona due to data availability limitations). These cities are home to a total of 15 million of Spanish citizens, representing more than 30% of Spain’s total population.

The Spanish urban landscape has witnessed dramatic changes since the early 2000s, initially driven by a real estate boom fuelled by low interest rates, easy credit and speculative investment (Burriel, 2011) and followed by rapid construction on urban outskirts (Andrés López, 2023). Later, the 2008 global financial crisis led to halted construction projects, urban sprawl and environmental degradation (Cardesín Díaz and Araujo, 2017). Meanwhile, recent years have seen a focus on policies aimed at redeveloping and regenerating existing urban areas (Rodríguez-Suárez et al., 2023).

This study employs Spanish cadastral data provided by the Directorate General for Cadastre (2024). This data is available on three levels: INSPIRE level, detailed level and historical level. INSPIRE level is easily accessible and usable, and it provides standardised and harmonised information about land parcels across Europe, enabling cross-border cooperation and consistency in spatial data usage.

Detailed level is a more complete version. It offers a much more precise and thorough description of parcel properties than the INSPIRE level, including specific construction types. It supports large-scale downloads since 2022 and requires advanced knowledge to properly be used since it is presented in a complex structure separated into two data types: geometries, composed of different types of spatial units such as blocks, parcels and buildings; and alphanumeric data, which includes information on urban land use (see Figure S1 in the supplementary material).

Lastly, the historical level gathers all the changes that the cadastral parcel has undergone over the past 20 years. This data is also presented in a complex format that requires advanced knowledge to use effectively. Its applications are more limited, particularly for large-area studies, since it provides no information beyond the geometrical properties of the parcels and does not support large-scale downloads.

With all this considered, detailed cadastral data was selected to conduct our study since it combines the richest amount of information with large-scale download. Thus, datasets for 2022 and 2024 (the latest available data) were obtained and processed with the Cadastral Classifier QGIS plugin (Shurupov et al., 2023). Custom classification option was selected to produce six urban land use classes from the original cadastral categories: 01 – Single-family Residential, 02 – Multi-family Residential, 03 – Industrial, 04 – Commerce and offices, 05 – Other uses and 06 – Undeveloped (a more detailed description of classes can be found in Table S1 of the supplementary material). This yielded two land use maps from 2022 to 2024 for each selected city. Figure 1 illustrates this data preprocessing with an example of the resulting classification for the city of Zaragoza in 2024.

Figure 1.

Cadastral data processing workflow with a classification map example.

Methodology

The research has been divided into three blocks, each designed to address one of the research questions. Firstly, a series of attribute and spatial analyses were conducted to identify, classify and map the Transformations Processes (TPs) that cadastral parcels undergo (RQ1). Secondly, a series of spatial metrics were calculated to measure the characteristics of parcel geometries, and a Factor Analysis (FA) was implemented to select those that best represented the different spatial dimensions of parcels. Subsequentially, each TP was then characterised by evaluating the magnitude of the alterations produced in the involved parcels (RQ2). Finally, to compare groups of parcels that participate in each TP and those that remained unchanged, an analysis of the metric distribution through boxplot combined with Kruskal–Wallis test, Dunn test and Cliff’s Delta, was used (RQ3). A comprehensive overview of these stages can be consulted in the supplementary material (Figure S2).

Identification of transformation processes

The aim of this stage is to display cadastral geometry changes and classify them in a proposed set of TPs. A three-step analysis was conducted. This consisted of (i) attribute analysis using Cadastral Reference (CR) and geometry; (ii) spatial analysis through the intersections of the parcel geometries from both years; and (iii) a spatial analysis employing a graph. To exclude detected changes resulting from minor boundary adjustments or routine cadastral data corrections, two thresholds were applied. First threshold consisted of a minimum of 1% area value change to consider that a parcel has undergone geometry modification, calculated as $0.99 < {A r e a}_{2023} {/ A r e a}_{2021} > 1.0 .$ Second threshold of 10% minimum intersecting area was used to avoid fictious spatial relations, calculated as ${A r e a}_{o v e r l a y} {/ A r e a}_{2021} > 0.1 & {A r e a}_{o v e r l a y} {/ A r e a}_{2023} > 0.1$ .

An illustrative diagram and detailed explanation of the steps can be found in the supplementary material (Figure S3).

Identification of the changes made by TPs to parcels

To capture the changes, a set of spatial metrics that explain the different spatial characteristics of parcel geometries was initially selected. For this research, only metrics measuring spatial characteristics of parcels as independent objects were considered. These were grouped into three categories: size, complexity and shape. A detailed list can be consulted in the supplementary material (Table S2). A statistical analysis was conducted on all the metrics considered to select the most representative ones. This was achieved through correlation analysis to reveal potential redundancy and FA to determine which metrics had the greatest factor loadings, allowing to identify the following metrics as the most representative for the dataset:

- AR: directly related to the shape of the parcel geometry, specifically its elongation by comparing the length to the width of the geometry using the minimum rotated rectangle:

A R = \frac{{M R R}_{w i d t h}}{{M R R}_{h e i g h t}}

- cPAR: measures the intrinsic complexity of a parcel geometry by comparing its area and perimeter and adjusting it to square standard to solve size dependency:

c P A R = \frac{P}{\sqrt{A}}

- SCI: measures the extrinsic complexity, by comparing the area of the parcel geometry with the area of a geometry derived from the polygon of minimal geometry enveloping the parcel. However, it was shown to be related to the form metrics COMP and RECT, which also make a similar comparison, but with a circle and a rectangle, respectively:

S C I = 1 - \frac{A_{p a r c e l}}{A_{c o n v e x h u l l}}

- SK_NB: measures intrinsic complexity by analysing the internal complexity of parcel geometry’s skeleton:

S K_N B = \sum S k e l e t o n b r a n c h e s

- Additionally, AREA was also considered since it is directly related to the size of parcels.

More details of this selection can be consulted in the supplementary materials (Figures S4 and S5).

To complement these five spatial metrics, two more dimensions were considered: number of parcels and LU of parcels. The first one provides information on how a TP impacts the total number of parcels (i.e. how many new parcels a particular Subdivision produces), while the second provides information on whether a particular TP results in a change in the LU of the parcels (i.e. an ‘Undeveloped’ parcel that undergoes a Subdivision and results in a new set of ‘Single-family residential’ parcels). This yielded seven different types of changes, and the magnitude produced by each TP on these was analysed between the input parcels (targeted by TPs, from 2022) and the output parcels (resulting from TPs, from 2024), as summarised in Table 1.

Table 1.

Metrics defined to characterise transformation processes.

Metric change	Description	Formula	Labelling criteria
ΔAREA	Measure the changes in the spatial characteristics between input parcels and output parcels	$Δ M = \frac{{\bar{m}}_{p a r c e l s 2024}}{{\bar{m}}_{p a r c e l s 2022}}$ Where $Δ M$ is the change of a given metric M and $\bar{m}$ is the mean value of metric m for the parcels before (in, 2022) and after (out, 2024) the occurrence of the TP.	(↓↓): $Δ M$ < −50% (↓): $Δ M$ < −10% (=): $Δ M$ < 10% (↑): $Δ M$ > 10% (↑↑): $Δ M$ > 50%
ΔcPAR
ΔAR
ΔSCI
ΔSK_NB
ΔNP	Measures the change between the input number of parcels and output number of parcels	$PR = \frac{\sum {P a r c e l}_{2024}}{\sum {P a r c e l}_{2022}}$	(↓↓): PR < −0.5 (↓): PR < −0.1 (=): PR < 1.1 (↑): PR > 1.1 (↑↑): PR > 2.0
ΔLU	Measures the changes in land use (LU) between input parcels and output parcels	Seven different labels were implemented based on the following criteria: Diversification: Output parcels have a greater variety of LU. Extra undeveloped: Output parcels have new undeveloped parcels. Reclassification: Output parcels have a completely different LU to the input ones. Unchanged: The LU of the input and output parcels is the same. Undeveloped: The LU of input and output parcels is undeveloped. Consolidation: Undeveloped input parcels changed to some LU in the output. Homogenisation: Output parcels have a lower variety of LU.

Characteristics of parcels resulting from TPs

To analyse the characteristics of both parcels targeted by and resulting from the TPs, their metric distribution was obtained. To evaluate the differences in size, shape and complexity between them and unchanged parcels, a series of statistical tests were conducted: (i) a Kruskal–Wallis test to determine significant differences in parcel geometries across all TP; (ii) a Dunn’s post-hoc test to further analyse and show specific statistically significant differences in the selected metrics between TP pair-wise combination; and (iii) Cliff’s Delta to measure the magnitude of these differences. These tests provided precise results about the characteristics of parcels targeted by and resulting from each TP and how they compare to unchanged parcels. Additionally, a boxplot for all metrics was obtained, providing a graphic visualisation of the differences.

Results

Classification of the transformation processes: Mapping changes in urban parcels

The analytical framework developed allowed to identify six different spatial parcel transformation processes: (i) reshape, (ii) aggregation, (iii) subdivision, (iv) reconfiguration, (v) reconfiguration towards aggregation and (vi) reconfiguration towards subdivision. Additionally, a negligible/not intersected category grouped together parcels without any spatial relationships and those that did not meet the threshold criteria. Figure 2 illustrates an example of these results for the city of Zaragoza. Figure 2(a) shows the results of the first analysis conducted using the attributes of CR and Area attributes of the datasets. Figure 2(b)–(e) show the results of the second spatial analysis that uses intersection operation to detect single relationships of independent parcels. Finally, Figure 2(f) shows the classification with the third spatial analysis that incorporate a graph construction upon the previous single TP classification. In this case the relationships of single independent parcels are extended using connected components: leveraging graphs where each parcel is a node and connections each of the spatial relationships between them.

Figure 2.

TPs identified. Example for Zaragoza.

To compile the results, an interactive web viewer was developed. The outputs for all Spanish capital cities were mapped and can be accessed via the following link https://nikoshur.github.io/Spanish_parcel_transformations/, where the 2022 and 2024 maps can be compared. The viewer includes multiple visualisation options for the results of each analytical step, from the initial land use classification to the final classification of TPs.

As regards TP prevalence, Figure 3 shows an overview of the TPs that occurred in all Spanish capital cities between 2022 and 2024. The left graph shows the raw amount TPs, the central graph shows the number of input (targeted by a TP) and output (resulting from a TP) parcels impacted by those TPs and the right graph accounts for the total area of these participating parcels.

Figure 3.

TP frequencies.

Of the six TPs identified, Reshape has been shown to be the most frequent, accounting for ∼68% of total individual TPs and ∼46% of total parcels (both targeted and resulting). However, it tended to focus on small, localised parcels that often require routine cadastral adjustments, which have no impact on large portions of urban land (∼19%). Conversely, Subdivision and Reconfiguration towards subdivision had the greatest impact in terms of area, covering a large amount of urban land (∼48% of total area) and involving a significant number of parcels (∼28% of total resulting parcels). Aggregation, Reconfiguration and Reconfiguration towards aggregation, meanwhile, are less frequent (∼25% of total parcels created) and cover a smaller area of urban land (∼33% of total area).

Changes produced on parcel geometries by each TP

Changes produced by each TP are summarised in Figure 4 (labels defined Table 1). Percentages represent the number of parcels involved in each process undergoing high increase (↑↑), low increase (↑), stability (=), low decrease (↓) and high decrease (↓↓) in the corresponding metric.

Figure 4.

Changes produced by each TP on the parcel’s geometric characteristics and LU.

The Reshape and Reconfiguration TPs show a high level of stability in most metrics. Although, several changes can be observed in terms of the complexity of parcels. In particular, these TPs tend to increase SK_NB without clear SCI increase, which indicates that these TPs involve an increase of the number of vertices and thus the internal complexity, without making the geometries more complex in terms of shape.

Aggregation and Reconfiguration towards aggregation present very noticeable similarities. Both processes involve a merging of geometries that result in larger parcels that also entails an increase in the complexity from an internal (increase of SK_NB by more than 50% in most cases, which implies more vertices) and shape perspective (increase of SCI by more than 50% in most cases). This is accompanied by a relative tendency to increase AR (with cases showing an increase between 10 and 50%).

Subdivision and Reconfiguration towards subdivision also showed a level of similarity. Although the latter had a lower impact than the former (i.e. Subdivision reduced parcel size by more than 50% in most cases, whereas the reconfiguration counterpart reduced size by between 10 and 50%). Unlike Aggregation, these TPs result in a reduction in the complexity of parcel geometries, both in internal structure of their skeletons (decreases of more than 50% SK_NB in most cases) and the shape (which showed greater variability, with reductions of between 10 and 50% being the most frequent), also tending to lower the AR (in most cases, a decrease of between 10 and 50%) thereby increasing the elongation of new parcels.

Finally, regarding LU changes produced by these TPs, several findings were obtained. Reshape and Reconfiguration very rarely resulted in changes to LU (less than 10% involved any change to the original LU). On the other hand, Subdivision and Reconfiguration towards subdivision primarily targeted ‘Undeveloped’ parcels (∼40% and ∼36%, respectively), while also bringing with it a minor overall urban diversification, with the appearance of some new undeveloped (∼22% and 10%, respectively) or new use (∼13% and ∼10%, respectively) parcels. Lastly, Aggregation and Reconfiguration towards aggregation had varied effects on LU, leaning towards a consolidation process.

Characteristics of the parcels targeted by and resulting from the transformation processes

This section completes the results with the characteristics of the parcels targeted by (input) and resulting from (output) each of the TPs, comparing them with the unchanged parcels. Rather than evaluating the changes produced by each TP on the parcels involved, independent groups of parcels (i.e. Aggregation input parcels, Subdivision output parcels, etc.) are analysed and compared between them and with unchanged parcels.

These results are illustrated in the Figure 5, where the distributions of the metrics for each group of parcels are shown through a set of boxplots. The statistical comparison results are summarised using labels that combine colour, to indicate the existence of a significant statistical differences (Kruskal–Wallis and Dunn post-hoc tests) and arrows to indicate the magnitude and direction of such differences (Cliff’s Delta). Proportion of LU of each group and for unchanged parcels is also provided for further analysis of impact of the TPs on LU change. More detailed information of these statistical comparisons can be found in the supplementary material (Figures S6 to S10).

Figure 5.

Summary of TPs metric distribution and statistically significant difference, with median values shown above each corresponding boxplot.

Looking at these findings, from the TP perspective, Reshape and Reconfiguration didn’t show remarkable differences between the spatial characteristics of either targeted or resulting parcels with unchanged parcels. Moreover, LU distribution was also very similar to Unchanged parcels, so both processes didn’t seem to either target or produce parcels with a specific pattern.

On the other hand, the remaining TPs show significant differences in either spatial metrics or land use distribution, some of them reinforcing the results from previous sections. First, important similarities are observed between Subdivision and Reconfiguration towards subdivision, which target parcels that are much larger (1853 m² and 1376 m² vs 263 m²) and more complex (comparing medians of SK_NB, 37 and 33 vs 25) than the rest and produce parcels that are still larger (365 m² and 603 m² vs 263 m²) but with a complexity that is reduced to the point of being simpler than any other (comparing medians of SK_NB, 15 and 19 vs 25) and shape is more elongated (0.48 AR for parcels resulting from Subdivision vs 0.54 for unchanged). The reverse pattern is observed for Aggregation and Aggregation towards aggregation TPs, which target parcels with low complexity (19 and 16.5 vs 25 median SK_NB) and average size (218 m² and 189 m² vs 237 m²) and produce parcels that are larger (601 m² and 483 m² vs 237 m²), more complex (comparing medians of SK_NB, 35 and 32 vs 25) and less elongated (0.61 median AR for parcels resulting from Aggregation vs 0.54 for unchanged).

Regarding the LU distribution of the input and output parcels, ‘Undeveloped’ are the most frequently targeted parcels by most of the TP, with some having a higher incidence (i.e. Subdivision and Reconfiguration towards subdivision). Moreover, Subdivision and Reconfiguration towards subdivision appear to reduce the total developed land, since ‘Undeveloped’ proportion experiment a noticeable increase from input to output parcels for these processes. Additionally, ‘Single-family residential’, which is the most common developed LU, is strongly represented in targeted parcels of Reconfiguration and Reshape processes.

Discussion

Transformation processes patterns

Drawing on the results, the total number of parcels increased between 2022 and 2024 (higher number of resulting parcels than targeted) due to large areas being redistributed through the dominance of Subdivision and Reconfiguration towards subdivision. However, this increase does not unequivocally indicate the development of new urban functions, since it mainly affects undeveloped land in terms of both target and resulting parcels. In terms of geometric alterations, the effect of these two TPs is clear: large parcels are split to greatly reduce their size and produce parcels with simple, elongated geometries. Together with the observed LU changes, these TPs often facilitate new developments, which may be more diverse within already developed zones. These patterns highlight the structural role of the aforementioned TPs in urban dynamics, as they are directly associated with the intensification of urban land, and, consequently, the adaptation of the urban fabric to finer-grained urban configurations that are often accompanied by infill growth and the re-parcelling of previously underutilised land that needs to be prepared for future development.

By contrast, Aggregation and Reconfiguration towards aggregation produced opposite but more varied effects: integration of small simple-shaped geometries into large and more complex geometries. In some cases, these changes were accompanied by the consolidation of urban spaces through the agglomeration of LU and the reduction of heterogeneity. In other cases, they appeared to prepare large areas for future large-scale projects. This could suggest that these processes are part of a more complex land redistribution process involving different sequential TPs stages, with aggregation TP preceding subsequent subdivision TP.

A third group of Reshape and Reconfiguration TPs produced different results and are best interpreted as non-impactful in terms of urban changes. They minimally alter LU and do not substantially alter the parcel geometry and thus the urban system. Although different thresholds have been implemented, significant proportion of these processes seem to be linked to cadastral adjustments and geometric corrections (boundary refinements, alignment fixes, and precision updates). As these are the most frequent group of identified processes, it suggests that cadastral updates mainly focus on correcting data with no meaningful land development changes.

Taken together, the results support a straightforward synthesis of three TP families, each with a distinct urban meaning: Non-impactful Processes (Reshape and Reconfiguration) which are primarily associated with cadastral corrections; Subdividing Processes (SP, Subdivision and Reconfiguration towards subdivision), which are linked to fragmentation and the production of finer-grained parcels, and are often associated with intensification/infill growths; and Aggregating Processes (AP, Aggregation and Reconfiguration towards aggregation), which are linked to consolidation or the preparation of urban land for future large-scale coordinated projects.

We believe that the proposed conceptual framework enables the following: (i) a rigorous definition of TPs based on spatial analysis tools and their mapping across a large proportion of Spanish cities; (ii) the spatial characterisation of changes in parcel geometries; and (iii) the identification of target parcels undergoing transformation processes.

Moreover, the identification, mapping, and description of these TPs could contribute to urban studies in several ways. Firstly, examining the location of these transformation processes within urban areas (e.g. centre vs periphery) may contribute to understanding their relationship with spatial growth patterns such as infilling, edge expansion, and outlying development (Liu et al., 2010), both internationally and within the Spanish context. In this regard, analysing TPs in connection with specific urban expansion and planning policies may be particularly relevant, including the Urban Action Plans (PAUs) in Madrid (Martinez Perez and Walker, 2018), large residential development projects such as the Madrid Nuevo Norte project (Metaxas et al., 2021), current rehabilitation, regeneration, and renewal agendas in Spain (Rodríguez-Suárez et al., 2023), and the long-term European objective of achieving ‘no net land take’ by 2050 (Evers, 2026).

Secondly, the definition of TPs may also be useful in the field of urban simulation using vector-based cellular automata (VCA) models (Molinero-Parejo et al., 2023). These models, which can incorporate parcel transformation algorithms, could simulate both urban growth/change and parcel transformations (Colomb et al., 2022; Dahal and Chow, 2014; Wickramasuriya et al., 2011) more accurately by identifying parcels subject to transformation processes and applying appropriate subdivision or aggregation algorithms to reproduce the spatial characteristics of such TPs.

Finally, the proposed framework could also support policy impact assessment and studies of urban densification or urban sprawl at a fine spatial scale when monitoring future urban processes. As observed, ‘undeveloped’ parcels undergo most TPs as a preliminary step prior to urban development. This may be highly valuable for detecting potential urban expansion before it occurs and before it can be captured by satellite imagery. Such capability would facilitate the control of outlying growth and the early identification of potential infill development.

Limitations and future research

Two limitations of this study should be highlighted: errors in cadastral data and time period considered in the analysis. Regarding the first one, cadastral data is not free of errors. Parcels with duplicate or erroneously designated CRs were found. This was solved with simple assignation of new unique CRs, which might result in overestimating new and extinguished parcels. Other errors were related to the geometry of parcels that lead to ‘translation phenomena’. This refers to a fictitious change in geometry caused by errors in the geometric cadastral definition of parcels. To avoid misclassification of these adjustments, different thresholds were tested, and the ones that yielded the best results were selected. Nevertheless, some fictitious processes are still present, overestimating the total amount of Non-Impactful processes, as previously discussed.

Regarding the second limitation, although the proposed methodology enables an exhaustive comparison between two cadastral parcel maps, the period considered is relatively short. Alternatively, historical cadastral data could have been used to analyse a longer temporal span; however, this would have limited the working scale and would not have allowed the inclusion of information on parcels’ land use. Nevertheless, since the main objective of this study was to establish a methodological framework for analysing TPs and their implications for cadastral parcels, the time period examined has been sufficient to define the main characteristics of the transformation processes.

Future studies may therefore explore longer temporal horizons and their relationship with urban expansion processes, urban planning policies, and the spatial location of urban processes (e.g. centre-periphery dynamics). We consider that making the results of this study publicly available through the interactive viewer will facilitate the development of these and other lines of research in this field.

Likewise, the richness of cadastral data enables new approaches to be explored when defining cadastral parcel TPs. Information on buildings within parcels could be used to derive detailed variables, such as building footprints or volumes. This could support fine-grained analyses of urban densification processes and their potential relationship with the proposed TPs.

Conclusions

In this study, we introduced an analytical framework to compare two highly detailed cadastral datasets for 48 Spanish cities. Its application highlighted a series of transformation processes in urban parcels from a geometric perspective, revealing three main types of processes: subdividing processes, aggregating processes, and non-impactful processes. While non-impactful processes were the most frequent, they accounted for a small proportion of total urban land and had very limited effects on parcel geometries. Subdividing processes, however, were the most impactful, both in terms of the total area of urban land affected and the number of parcels generated. In these cases, very large parcels with complex shapes were targeted, resulting in larger-than-average parcels with simpler geometries and a tendency toward high elongation, a common feature of residential parcels. Aggregating processes, by contrast, affected small and simple-shaped parcels, producing highly compact and geometrically complex parcels. Taken together, these processes predominantly concentrated on undeveloped land and in most cases the resulting parcels were not assigned a new LU. This finding indicates that geometric changes and LU changes do not necessarily occur together and may respond to partly different urban mechanisms. Recognising this decoupling is analytically important, because it positions cadastral geometric change as an autonomous component of urban transformation and not simply as an expression of LU change.

Supplemental material

Supplemental Material - Identifying and characterising urban land use transformation processes using cadastral parcel geometries: A case study of Spanish cities

Supplemental Material for Identifying and characterising urban land use transformation processes using cadastral parcel geometries: A case study of Spanish cities by Nikolai Shurupov and Francisco Aguilera-Benavente in Environment and Planning B: Urban Analytics and City Science

Footnotes

Acknowledgements

This paper has been developed in the context of ‘SIM4PLAN: Desarrollo de un prototipo multiusuario de ayuda a la planificación’ project, funded by the Spanish Government under National R + D Plan (2021-2023). Grant Agreement no. PDC2021-121568-C22 and the research program ‘Estímulo a la Excelencia para Profesores Universitarios Permanentes’ funded by University of Alcalá and the Community of Madrid (EPU-INV/2020/009). First author has been supported by the Spanish Ministry of Science, Innovation and Universities (MICIU), through a FPU doctoral fellowship (grant no. FPU21/06590). The authors would like to thank Linguaservice for English language revision. The authors would also like to thank the reviewers for their careful reading of the manuscript and their constructive comments and suggestions, which helped improve the paper’s quality and clarity.

ORCID iDs

Nikolai Shurupov

Francisco Aguilera-Benavente

Author contributions

Shurupov Nikolai: Writing – original draft, Writing – review & editing, Methodology, Investigation, Formal analysis, Data curation, Conceptualization, Software, Visualization, Validation. Aguilera-Benavente Francisco: Supervision, Writing – original draft, Writing – review and editing, Project administration, Funding acquisition, Conceptualization, Resources.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: FPU doctoral fellowship (grant no. FPU21/06590), ‘SIM4PLAN: Desarrollo de un prototipo multiusuario de ayuda a la planificación’ Project (grant no. PDC2021-121568-C22) and ‘Estímulo a la Excelencia para Profesores Universitarios Permanentes’ research program (grant no. EPU-INV/2020/009).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The data used in this study are publicly available from the by the Directorate General for Cadastre through its official data dissemination portal: https://www.sedecatastro.gob.es/. The entire methodology was implemented using the Python programming language, with the main support of Geopandas, scikit-learn, NetworkX and Matplotlib libraries. The source code, as well as the visualisation of the generated datasets, can be accessed via the following link to the paper repository: .

Supplemental material

Supplemental material for this article is available online.

Author biographies

Nikolai Shurupov holds a BSc in Environmental Sciences and is a PhD candidate in the Doctoral Programme in Geographical Information Technologies at the University of Alcalá. His research focuses on urban analytics and spatial modelling, with a particular interest in simulating land-use changes, analysing urban growth patterns, and developing geospatial applications for urban planning.

Francisco Aguilera-Benavente holds a BSc in Environmental Sciences and a PhD in Geographical Information Technologies. He is an Associate Professor of Human Geography at the University of Alcalá (Spain), where he directs the Master's Programme in Geographical Information Technologies and teaches courses on spatial databases, Geographic Information Systems (GIS), and spatial analysis. His main research interests include urban growth simulation models, land-use change modelling, and the analysis of urban growth patterns.

References

Andrés López

(2023) Recent transformations in the morphology of Spanish medium-sized cities: from the compact city to the urban area. Land 12(7): 1276. https://doi.org/10.3390/land12071276

Barreira-González

Aguilera-Benavente

Gómez-Delgado

(2019) Implementation and calibration of a new irregular cellular automata-based model for local urban growth simulation: the MUGICA model. Environment and Planning B Urban Analytics and City Science 46: 243–263. https://doi.org/10.1177/2399808317709280

Benhammou

Alcaraz-Segura

Guirado

, et al. (2022) Sentinel2GlobalLULC: a Sentinel-2 RGB image tile dataset for global land use/cover mapping with deep learning. Scientific Data 9(1): 681. https://doi.org/10.1038/s41597-022-01775-8

Burriel

(2011) Subversion of land-use plans and the housing bubble in Spain. Urban Research and Practice 4(3): 232–249. https://doi.org/10.1080/17535069.2011.616743

Caramaschi

Chiodelli

(2023) Reconceptualising housing emptiness beyond vacancy and abandonment. International Journal of Housing Policy 23(3): 588–611. https://doi.org/10.1080/19491247.2022.2074268

Cardesín Díaz

Araujo

(2017) Historic Urbanization Process in Spain (1746–2013). Journal of Urban History 43(1): 33–52. https://doi.org/10.1177/0096144215583481

Colomb

Tannier

Perret

, et al. (2022) Parcel Manager: a parcel reshaping model incorporating design rules of residential development. Transactions in GIS 26(6): 2558–2597. https://doi.org/10.1111/tgis.12970

Dadashpoor

Azizi

Moghadasi

(2019) Analyzing spatial patterns, driving forces and predicting future growth scenarios for supporting sustainable urban growth: evidence from Tabriz metropolitan area, Iran. Sustainable Cities and Society 47: 1–15. https://doi.org/10.1016/j.scs.2019.101502

Dahal

Chow

(2014) A GIS toolset for automated partitioning of urban lands. Environmental Modelling & Software 55: 222–234. https://doi.org/10.1016/j.envsoft.2014.01.024

10.

Dalmau

Garcia-Almirall

Domínguez

, et al. (2014) From raw data to meaningful information: a representational approach to cadastral databases in relation to urban planning. Future Internet 6(4): 612–639. https://doi.org/10.3390/fi6040612

11.

Directorate General for Cadastre (2024) Dissemination of cadastral data. In: Cadastral Electronic Site. https://www.sedecatastro.gob.es/

12.

Domingo

van Vliet

Hersperger

(2023) Long-term changes in 3D urban form in four Spanish cities. Landscape and Urban Planning 230: 104624. https://doi.org/10.1016/j.landurbplan.2022.104624

13.

Evers

(2026) Taking land seriously: tracing the politicization of the European ‘no net land take in 2050’ target. Journal of European Integration: 1–24. https://doi.org/10.1080/07036337.2026.2632709

14.

Ez-zahouani

Teodoro

El Kharki

, et al. (2023) Remote sensing imagery segmentation in object-based analysis: a review of methods, optimization, and quality evaluation over the past 20 years. Remote Sensing Applications: Society and Environment 32: 101031. https://doi.org/10.1016/j.rsase.2023.101031

15.

Fleischmann

Feliciotti

Kerr

(2022) Evolution of urban patterns: urban morphology as an open reproducible data science. Geographical Analysis 54(3): 536–558. https://doi.org/10.1111/gean.12302

16.

Grădinaru

Badea

A-C

Dragomir

, et al. (2023) Integrating cadastral data with seismic risk data in an online building database for the historical Centre of Bucharest City. Land 12(8): 1594. https://doi.org/10.3390/land12081594

17.

Haase

Rink

Grossmann

, et al. (2014) Conceptualizing urban shrinkage. Environment and Planning A 46(7): 1519–1534. https://doi.org/10.1068/a46269

18.

Herold

Couclelis

Clarke

(2005) The role of spatial metrics in the analysis and modeling of urban land use change. Computers, Environment and Urban Systems 29(4): 369–399. https://doi.org/10.1016/j.compenvurbsys.2003.12.001

19.

Huang

Lieske

Liu

(2023) Factors influencing vertical urban development at the parcel scale: the case in Brisbane, Australia. Environment and Planning B: Urban Analytics and City Science 50(3): 694–708. https://doi.org/10.1177/23998083221129283

20.

Hussain

Mehmood

Anees

, et al. (2024) Assessing forest fragmentation due to land use changes from 1992 to 2023: a spatio-temporal analysis using remote sensing data. Heliyon 10(14): e34710. https://doi.org/10.1016/j.heliyon.2024.e34710

21.

Kane

Hipp

Kim

(2017) Analyzing accessibility using parcel data: is there still an access–space trade-off in Long Beach, California? The Professional Geographer 69(3): 486–503. https://doi.org/10.1080/00330124.2016.1266951

22.

Kim

(2013) Measuring the containment and spillover effects of urban growth boundaries: the case of the Portland metropolitan area. Growth and Change 44(4): 650–675. https://doi.org/10.1111/grow.12028

23.

Liu

Chen

, et al. (2010) A new landscape index for quantifying urban expansion using multi-temporal remotely sensed data. Landscape Ecology 25(5): 671–682. https://doi.org/10.1007/s10980-010-9454-5

24.

Mansour

Souiah

Gacemi

MEA

, et al. (2024) Evolutions and threats to land management in the peri-urban area of the city of Sidi Bel Abbes (Algeria) - an analysis based on spatial cadastral data. Geoadria 28(2): 173–190. https://doi.org/10.15291/geoadria.4145

25.

Martinez Perez

Walker

(2018) Around Madrid: The Continuing Influence of Historical Urban Development Plans on Today’s Periphery. Journal of Urban Design, 1–20. https://doi.org/10.1080/13574809.2017.1413934

26.

Metaxas

Juarez

Gavriilidis

(2021) Planning and marketing the City for sustainability: the Madrid nuevo norte Project. Sustainability 13(4): 2094. https://doi.org/10.3390/su13042094

27.

Molinero-Parejo

Aguilera-Benavente

Gómez-Delgado

, et al. (2023) Combining a land parcel cellular automata (LP-CA) model with participatory approaches in the simulation of disruptive future scenarios of urban land use change. Computers, Environment and Urban Systems 99: 101895. https://doi.org/10.1016/j.compenvurbsys.2022.101895

28.

Molinero‐Parejo

Aguilera‐Benavente

Gómez‐Delgado

(2024) Adapting moving‐window metrics to vector datasets for the characterization and comparison of simulated urban scenarios. Transactions in GIS 28(2): 326–348. https://doi.org/10.1111/tgis.13139

29.

Perez

Ornon

Usui

(2021) Classification of residential buildings into spatial patterns of urban growth: a morpho-structural approach. Environment and Planning B: Urban Analytics and City Science 48(8): 2402–2417. https://doi.org/10.1177/2399808320974377

30.

Pradhan

Rajabifard

Aryal

, et al. (2025) Dissecting multidimensional morphology of urban sprawl: framework for policy intervention. Cities 157: 105627. https://doi.org/10.1016/j.cities.2024.105627

31.

Rodríguez-Suárez

Álvarez-Del-Valle

Díez-Bermejo

(2023) Planeamiento y regeneración urbana: la articulación entre políticas urbanísticas y sectoriales como respuesta a los retos urbanos. Ciudad y Territorio - Estudios Territoriales 55(217): 859–872. https://doi.org/10.37230/CyTET.2023.217.15

32.

Sánchez Ondoño

Prada Trigo

(2023) Medición de la vulnerabilidad urbana a partir del catastro: Aplicación a la ciudad de Valladolid. Scripta Nova: Revista Electrónica de Geografía y Ciencias Sociales 27(3): 143–181. Available at: https://doi.org/10.1344/SN2023.27.42092

33.

Shurupov

Molinero-Parejo

Rodríguez-Espinosa

, et al. (2023) Cadastral Classifier: QGIS plugin for the classification of urban land use at parcel level. Boletin de La Asociacion de Geografos Espanoles (97): 1–37. Available at: https://doi.org/10.21138/bage.3280

34.

Stevens

Dragicevic

Rothley

(2007) iCity: a GIS–CA modelling tool for urban planning and decision making. Environmental Modelling & Software 22(6): 761–773. https://doi.org/10.1016/j.envsoft.2006.02.004

35.

Sun

, et al. (2023) Coupled impacts of climate and land use changes on regional ecosystem services. Journal of Environmental Management 326: 116753. https://doi.org/10.1016/j.jenvman.2022.116753

36.

United Nations Human Settlements Programme (2022) World Cities Report 2022. United Nations. https://doi.org/10.18356/9789210028592

37.

Usui

(2025) Morphological dynamics of cost-efficient urban areas: expansion and shrinkage in greater Tokyo from 2003 to 2020. Cities 157: 105468. https://doi.org/10.1016/j.cities.2024.105468

38.

Verburg

Neumann

Nol

(2011) Challenges in using land use and land cover data for global change studies. Global Change Biology 17(2): 974–989. https://doi.org/10.1111/j.1365-2486.2010.02307.x

39.

Villar Navascués

Baños Castiñeira

Hernández Hernández

, et al. (2023) Territorial planning in times of crisis (2008-2022): paradigm shift and re-growth on the Coast of Alicante (Spain). Boletín de La Asociación de Geógrafos Españoles (99): 1–33. Available at: https://doi.org/10.21138/bage.3450

40.

Wang

Sun

Cao

, et al. (2023) A review of regional and global scale Land Use/Land Cover (LULC) mapping products generated from satellite remote sensing. In: ISPRS Journal of Photogrammetry and Remote Sensing. Elsevier B.V, Vol. 206, 311–334. https://doi.org/10.1016/j.isprsjprs.2023.11.014

41.

Wickramasuriya

Chisholm

Puotinen

, et al. (2011) An automated land subdivision tool for urban and regional planning: concepts, implementation and testing. Environmental Modelling and Software 26(12): 1675–1684. https://doi.org/10.1016/j.envsoft.2011.06.003

42.

Yao

Liu

, et al. (2017) Simulating urban land-use changes at a large scale by integrating dynamic land parcel subdivision and vector-based cellular automata. International Journal of Geographical Information Science 31(12): 2452–2479. https://doi.org/10.1080/13658816.2017.1360494

43.

Yao

Cheng

Sun

, et al. (2022) VecLI: a framework for calculating vector landscape indices considering landscape fragmentation. Environmental Modelling and Software 149: 105325. https://doi.org/10.1016/j.envsoft.2022.105325

44.

Zhang

Wrenn

Irwin

(2017) Spatial heterogeneity, accessibility, and zoning: an empirical investigation of leapfrog development. Journal of Economic Geography 17(3): 547–570. https://doi.org/10.1093/jeg/lbw007

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

1.58 MB