Modelling the Road Network as an Expression of Historical Spatial

System Change in the Klaipeda Region (Lithuania)

Thomas Gloaguen

1,2 a

, Kęstutis Zaleckis

and Sébastien Gadal

Cultural and Spatial Environment Research Group, Faculty of Civil Engineering and Architecture,

Kaunas University of Technology, 51367 Kaunas, Lithuania

UMR 7300 ESPACE, CNRS, Aix-Marseille Université, Université Côte-d’Azur, Avignon Université,

84029 Avignon, France

Keywords: Spatial System, Network Structures, Generic City Concept, Space Syntax, Spatiotemporal Modelling,

Lithuania.

Abstract: Since the 20

century, the Klaipėda region (Lithuania) has undergone significant political, economic, social,

and cultural changes related to the disappearance of Prussia (1923), the Soviet occupation (1944) and the

restoration of independence (1991). This has led to radical transformations in the spatial system, of which the

road network is a key component. Networks could be analysed using space syntax methods by assimilating

them to a mathematical graph model, for which quantitative indexes based on topology can be applied to

assess their spatial configurations. The ‘Generic City’ approach, by simulating foreground and background

urban networks, can be modified innovatively on a territorial scale. Based on historical cartographies and

open geographic databases, indexes characterising both networks are derived to describe the length, angularity

or accessibility of the pre-Soviet, Soviet and post-Soviet networks. The analysis identified four main spatial

structures associated with the network by combining the indexes in a K-means machine learning algorithm.

They highlight the spatial impacts of collectivisation, industrialisation, and tertiarisation of the economy, post-

World Wars and post-Cold War geopolitical events, and their consequences as drivers of the territorial

organisation and dynamics such as the metropolisation or peri-urbanisation.

1 INTRODUCTION

Since the 20

century, the coastal region of Klaipėda

in Lithuania has undergone significant and rapid

transformations. These changes were initially

political, with the territory being divided between the

Kingdom of Prussia and the Russian Empire,

experiencing a brief period of independence during

the interwar years, being incorporated into the Union

of Soviet Socialist Republics (USSR) in 1944, and

finally seeing the restoration of independence in 1991

(Plakans, 2011). These geopolitical shifts have been

accompanied by economic, social, cultural, and

societal changes, including agrarian reforms such as

agrarian capitalism and collectivisation (Jespen et al.,

2015), urban expansion, particularly suburban sprawl

(Cirtautas, 2015), Soviet industrialisation, and the

https://orcid.org/0000-0003-1849-9615

https://orcid.org/0000-0001-9223-9956

https://orcid.org/0000-0002-6472-9955

subsequent deindustrialisation and tertiarisation after

the Soviet era (Idzelis, 1984; Köll, 2024), alongside

recent trends of population decline and migration

(Ubarevičienė and Van Ham, 2017), throughout the

and early 21

centuries.

The organisation of geographic space is

significantly influenced by societal changes, whether

directly or indirectly. The analysis of these territorial

transformations is rooted in the concept of spatial

systems, which refers to the distribution (structures)

of geographical entities — such as built environments,

road networks, land use, land cover, and land parcels

— in space according to either spontaneous or

regulated logics (Archaeomedes, 1998; Allain, 2004;

Ellisalde, 2004).

The objective of this research is to examine the

evolution of the spatial system in the Klaipėda region

Gloaguen, T., Zaleckis, K. and Gadal, S.

Modelling the Road Network as an Expression of Historical Spatial System Change in the Klaipeda Region (Lithuania).

DOI: 10.5220/0013249600003935

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 11th International Conference on Geographical Information Systems Theory, Applications and Management (GISTAM 2025), pages 61-72

ISBN: 978-989-758-741-2; ISSN: 2184-500X

(Lithuania) since the 20

century. To achieve this

objective, the road network has been modelled using

space syntax methods, with a particular focus on the

‘Generic City’ concept.

The analysis of spatial systems involves

examining their structural complexity, which

includes the participation of multiple agents; spatial

complexity characterised by various geographical

scales; and temporal complexity marked by non-

linear evolutions (Badariotti, 2005). Spatial analysis

and modelling offer methodologies for

comprehending the intricacies of spatial systems by

translating geographical reality into mathematical

forms (Georges and Verger, 2013). This

methodological approach has been employed to

describe “[…] the distribution, configuration and

covariation” of geographical objects in space”

(Voiron-Canicio, 1995).

In complex spatial systems, networks facilitate the

relationships between various points of attraction and

diffusion, such as urban centres (Georges and Verger,

2013). The mathematical graph model serves as the

theoretical basis for space syntax, wherein networks

are represented by nodes corresponding to the

network segments and links connecting these nodes.

This theoretical framework enables the simulation of

interrelationships between spatial configurations and

their associated urban functions (Laouar and

Manzouz, 2017; Abida, 2018). Syntactic indexes

based on the topology of network segments are

utilised to examine how spatial structure influences

human behaviour through simulated movement.

This research focuses on the Klaipėda region,

where spatial changes are particularly pronounced

compared to the rest of Lithuania. The unique

environment of this coastal area is attributable to its

position as a maritime interface. During the Soviet era,

the port of Klaipėda emerged as a pivotal facilitator

of connectivity between the Baltic and the Black Sea

regions, driving the development of strategic

economic activities alongside the growth of seaside

tourism. Despite the region experiencing partial

militarisation during this period, Lithuania’s

independence in the 1990s led to the enhancement of

these activities and an opening to the market economy.

A consequence of this transformation has been the

rapid suburbanisation of the metropolitan and coastal

areas that began in the 2000s (Vaitkus and

Vaikuvienė, 2005; Veteikis et al., 2011; Cirtautas,

2015). Furthermore, the region is historically divided

between two ethnographic areas: Samogitia

(Žemaitija) in the north and Lithuania Minor (Mažoji

Lietuva) in the south, each exhibiting distinct

functional and cultural characteristics (Ragauskaitė,

2019). The former corresponds to the territory that

was under Russian occupation, while the latter was

incorporated into Prussia prior to the Soviet era.

2 THE ‘GENERIC CITY’

CONCEPTUAL APPROACH

This analysis draws upon the concept of the ‘Generic

City’, developed within space syntax theories and

methodologies to identify the spatial and functional

similarities between cities despite their diverse

cultural contexts (Hillier, 2007; Hillier, 2010).

This concept proposes that each city exhibit a dual

network characterised by (1) “longer lines and nearly

straight connections”, indicating a foreground

network, and (2) “shorter lines with more near right

angle connections, and so more localised and with

less linear continuity”, referring to a background

network.

The foreground network tends to be similar across

various cities, while the background network is more

specific and sensitive to cultural context. These

networks, with their distinct physical and functional

attributes, are evaluated using specific indexes

(section 3.2).

For instance, Al Sayed (2009) compared the

foreground and background networks of Manhattan

and Barcelona to showcase the emergence of dynamic

and adaptive urban networks. In the present study,

however, the indexes traditionally used to

characterise these networks are applied not in an

urban context but in a territorial one. They prove to

be insufficiently suited to the research objectives due

to their low statistical reliability and limited

differentiation between foreground and background

networks. The diversity of urban and rural settlements

(including individual and familial farms, villages,

towns, and cities) in relation to road networks

presents varying degrees of appeal for movements

across the territory (Dringelis, 2015). Furthermore,

the region has been shaped by multiple cultural

influences — Prussian, Russian, Soviet, and post-

Soviet — which have successively affected territorial

organisation, complicating the modelling of these

networks. Consequently, it has become essential to

employ syntactic indexes that more accurately reflect

the conceptual approach of the Generic City in a

territorial context (Gloaguen et al., 2023).

GISTAM 2025 - 11th International Conference on Geographical Information Systems Theory, Applications and Management

3 SYNTACTIC TERRITORIAL

INDEXES FOR

SPATIOTEMPORAL

MODELLING OF NETWORK

STRUCTURES

3.1 Data Acquisition

The present analysis adopts a historical retrospective

approach, necessitating the identification of

representative analytical periods that reflect differing

political, economic, social, and cultural contexts.

According to the scientific literature and the

conceptualisations proposed by some Lithuanian

researchers regarding the urban structures of Baltic

cities (Cirtautas, 2013) and rural settlements and

landscapes (Bučas, 2001), three primary recurring

periods are typically used for analysing the 20

and

centuries: the pre-Soviet (before 1945), the

Soviet (1945-1990), and the post-Soviet (after 1990).

The selection of these periods is also influenced

by the availability of cartographic data. Historical

cartographic resources are utilised for both pre-Soviet

and Soviet periods. Piškinaitė and Veteikis (2023)

note that cartographic materials are still underutilised

by researchers in Lithuania. For the initial period, the

series ‘Karte des Deutschen Reiches’ (Map of the

German Reich, 1870-1944) and ‘Karte des

Westlichen Russlands’ (Map of Western Russia,

1892-1921) are employed, both at a resolution of

1:100,000. For the subsequent period, Soviet military

topographic maps from 1985 at a resolution of

1:50,000 are used. The choice of these series is

predicated on several criteria, including their free

availability, the quality of their scanning, and the

inclusion of a legend for analysis and digitisation

purposes.

For the latest recent period (2023), free

geographic data from OpenStreetMap

(https://www.openstreetmap.org/) was extracted

using the ‘Over Pass Turbo’ tool (https://overpass-

turbo.eu/). Barrington-Leigh and Millard-Ball (2017)

estimate the completeness fraction of the

OpenStreetMap road network in Lithuania is

approximately 97% reflecting the accuracy of the

information compared to actual conditions.

The interactive, manually digitised data was then

processed to exclude unwanted networks, such as

footpaths, cycle paths, and occasionally utilised roads

like forest tracks. This process also involved the

removal of duplicate and isolated lines, as well as a

check for their connectivity.

3.2 ‘Generic City’ Indexes Adapted to

Territorial Analysis

The indexes are derived from the theoretical

framework that treats the network as a mathematical

graph, where nodes represent network segments and

links signify the connections between them.

Quantitative methodologies that prioritise the

topology of these segments are employed to assess

various spatial configurations, including centrality,

connectivity, accessibility, and movement.

Four indexes from the Spatial Network Analysis

Software (sDNA) library (https://sdna.cardiff.ac.uk/

sdna/) are utilised to characterise the foreground and

background networks (Cooper and Chiaradia, 2020;

Cooper, 2024). These indexes were initially chosen

based on experiments conducted in the same study

area, where their correlation with human activity data

— such as population density, amenities, and light

pollution — was assessed. The observed correlation

coefficients, which range from 0.61 to 0.74 on

average, reflect strong to extraordinarily strong

relationships, making them relevant for the analysis

(Gloaguen et al., 2022).

For the background network indexes, a radius of

1000 metres is applied to simulate local network

behaviour.

The Network Quantity Penalized by Distance

(NQPD) index is computed for both foreground and

background networks. This index quantifies the

number of accessible segments (network quantity)

within a specified radius of a given segment (equation

1). Each segment is regarded as equally attractive,

thereby allowing the focus to be solely on distance.

𝑁𝑄𝑃𝐷(𝑥) = 

(𝑊

(

𝑦

)

𝑃

(

𝑦

)



𝑑



(𝑥,𝑦)



∈



(1)

Where,

𝑅



is the set of polylines in the network radius

from the link 𝑥.

𝑊(𝑦) is the weight of a polyline 𝑦.

𝑃(𝑦) is the proportion of any polyline 𝑦

within the radius.

𝑑



(𝑥,𝑦) is the distance according to a metric

𝑀 along a geodesic defined by 𝑀, between an

origin polyline 𝑥 and a destination polyline 𝑦.

𝑛𝑞𝑝𝑑𝑛 and 𝑛𝑞𝑝𝑑𝑑 are the numerator and

denominator of NQPD, usually defined as 1.

The NQPD index emphasises the most accessible

areas of the network, which often align with the urban

central zone, characterised by high

multifunctionality.

Modelling the Road Network as an Expression of Historical Spatial System Change in the Klaipeda Region (Lithuania)

Three indexes are exclusively used to describe the

background network. First, the ‘Length’ index (LEN)

quantifies the cumulative length of segments within a

specified radius (equation 2).

𝐿𝐸𝑁 =  𝐿

(

𝑦

)

𝑃(𝑦)

∈



(2)

Where,

𝑅



is the set of polylines in the network radius

from the link 𝑥.

𝐿(𝑦) is the Euclidean length of polyline 𝑦.

𝑃(𝑦) is the proportion of any polyline 𝑦

within the radius.

This index can be likened to Peponis’s (2008)

‘Metric Reach’ index, as it underscores the physical

extent of connections and assesses the utilisation of

network areas.

Second, the ‘Angular Distance’ index (ANG)

captures the total angular curvature of segments

within the radius (equation 3).

𝐴

𝑁𝐺 =  𝑑



(𝑦



)

∈



(3)

Where,

𝑅



is the set of polylines in the network radius

from the link 𝑥.

𝑑



is the cumulative angular curvature along

the full length of the line in degrees.

𝑦



is the proportion of 𝑦 that falls within the

radius only.

Unlike the Length index, this index focuses solely

on closeness (geometrical proximity) and does not

account density.

Finally, the ‘Betweenness’ index (BT) (equation

4) assesses the frequency with which a segment

appears in the shortest paths connecting other

segments within a designated radius.

𝐵𝑇

(

𝑥

)

=𝑊

(

𝑦

)

𝑊

(

𝑧

)

𝑃

(

𝑧

)

𝑂𝐷

(

𝑦,𝑧,𝑥

)

∈



∈

(4)

Where,

𝑁 is the set of polylines in the global spatial

system.

𝑅



is the set of polylines in the network radius

from the link 𝑦.

𝑊(𝑦) and 𝑊(𝑧) are the weight of the

geodesic endpoints 𝑦 and 𝑧.

𝑃(𝑧) is the proportion of any polyline 𝑧 within

the radius.

𝑂𝐷(𝑦,𝑧,𝑥) are the geodesic paths between 𝑦

and 𝑧 that pass through a vertex 𝑥.

The Betweenness index reflects transit

movements, highlighting the key segments of the

network.

All indexes are standardised by dividing their

value by the standard deviation, which enables

comparisons across networks of varying sizes.

Standardisation also accommodates the spatial

resolutions of the geographic data utilised over

different periods.

3.3 Machine Learning Modelling of the

Spatial Structures of the Network

To identify the principal spatial structures associated

with the networks and their evolution over time, the

foreground and background indexes are integrated for

multivariate analysis.

The networks are overlaid on a 1km-by-1km grid,

and the average values of each index are calculated.

The grid size corresponds to the radius value used for

the background network indexes. However, it is

crucial to note that the selected grid size influences

the interpretation of results, as does the number of

clusters.

A multivariate analysis was conducted using a K-

means algorithm to minimise the variance of values

grouped into clusters. Effectively generalise and

visualise the results of the indexes, the number of

clusters was established at four. Each cluster

represents a spatial structure of the road network that

corresponds to the territorial organisation,

comparable both in time and space.

4 EVOLUTIONS OF

FOREGROUND AND

BACKGROUND NETWORKS

The foreground network has experienced

considerable evolution over time. As illustrated in

Figure 1, the centre of gravity of the network is

gradually shifting from the south during the pre-

Soviet period to the centre of the coastal region in the

post-Soviet period. The NQPD index indicates a

decline in regional accessibility of the network over

time, with average standardised values of 6.19 for the

pre-Soviet network, 5.24 for the Soviet network, and

4.22 for the post-Soviet network.

At the local level, two trends emerge: a decrease

in the NQPD index from the pre-Soviet (1.06) to

Soviet (0.94) periods, followed by an increase

between the Soviet and post-Soviet periods (1.20)

(Figure 2). These findings suggest a densification and

GISTAM 2025 - 11th International Conference on Geographical Information Systems Theory, Applications and Management

Figure 1: Spatial evolution of the foreground network between the (a) pre-Soviet, (b) Soviet, and (c) post-Soviet periods.

local expansion of the network, a trend also noted by

Strano et al. (2012) in their analysis of network

evolution in the suburbs of Milan. Notably, the

suburbs of Klaipėda witnessed significant

densification between the Soviet and post-Soviet

periods.

The Length and Angular Distance indexes, which

measure the potential for social interactions and the

organicity of the local networks, respectively, have

shown a decrease in their standardised values over

time. The average value for Length in the post-Soviet

period is 1.64, compared to 1.80 for the Soviet period

and 1.93 for the pre-Soviet period (Figure 2).

Regarding Angular Distance, the average value for

the post-Soviet period stands at 1.08, in contrast to

1.17 for the Soviet period and 1.21 for the pre-Soviet

period (Figure 2).

Figure 2: Quantitative evolution of background network

indexes.

The betweenness index, which highlights the

significance of a segment within the local network,

shows an increase in the average standardised values

from the pre-Soviet (0.33) to the Soviet period (0.49),

followed by a stabilisation between the Soviet and

post-Soviet periods (0.47) (Figure 2). This increase in

betweenness values suggests that the hierarchy

among routes is becoming more clearly defined over

time at the local level.

5 HIERARCHY AND DYNAMICS

OF NETWORK STRUCTURES

By integrating the foreground network and

background network indexes into a multivariate

analysis, the primary spatial structures of the

Klaipėda region’s spatial system are identified. These

structures are characterised by (1) the key local

centralities, (2) the territorial centre of gravity, (3)

secondary centralities in transition and (4) peripheral

areas. Their respective evolution is interpreted in

relation to existing scientific literature.

5.1 Main Local Centres

The first spatial structure is linked to grid tiles that

possess the highest average indexes, irrespective of

the type of network (Figure 3).

These grid tiles represent the main local

centralities within the network. Between the pre-

Soviet and post-Soviet periods, the networks

associated with the region’s primary centres

underwent a complete reconfiguration. In the pre-

Soviet era, the main local centres were situated in the

Modelling the Road Network as an Expression of Historical Spatial System Change in the Klaipeda Region (Lithuania)

Figure 3: Spatial evolution of networks relating to local centres between the (a) pre-Soviet, (b) Soviet, and (c) post-Soviet

periods.

south, with noteworthy development of individual

and family farms (Figure 3a and Figure 4).

In the post-Soviet era, the primary local centres

correspond to the main urban hubs of the coastal

region, specifically Klaipėda, Palanga, Kretinga,

Gargždai, and Šilutė (Figure 3c).

This transformation can be attributed to the

agricultural and land planning policies implemented

by the Soviet Union, which resulted in the dissolution

of individual and family farms and their replacement

with collective farms, such as ‘sovkhoz’ and

‘kolkhozes’ in the early 1950s (Gadal, 2011; Köll,

2024). The structure of the pre-Soviet individual and

family farms is comparable to these main urban

centres in terms of accessibility, underscoring their

historical significance in the economic and territorial

development of the Klaipėda region. Accessibility

analysis supports this conclusion: the top 25% of

values for the pre-Soviet network exceed 1.30, while

the post-Soviet network shows values of 1.70,

indicating a more organic and interconnected network

prior to Soviet policies of land collectivisation.

The reconfiguration of the network illustrates the

process of metropolisation taking place in the Baltic

countries (Ubarevičienė, 2018). This process

commenced during the Soviet period when territorial

planning adhered to a ‘hierarchical centre-periphery’

model (Vanagas et al., 2002). Klaipėda was

designated as a regional centre, intended to be linked

Figure 4: Network associated with local centres: the case of

individual/family farms during the pre-Soviet period.

GISTAM 2025 - 11th International Conference on Geographical Information Systems Theory, Applications and Management

Figure 5: Spatial evolution of networks relating to territorial centre of gravity between the (a) pre-Soviet, (b) Soviet, and (c)

post-Soviet periods.

to secondary regional centres through a network of

smaller cities like Palanga or Kretinga. In the post-

Soviet period, this process accelerated as the Soviet

territorial framework persisted, leading to heightened

urbanisation of the main urban centres. The findings

from the Betweenness index at the local level (section

4) are particularly illuminating, revealing a more

pronounced hierarchy of networks emerging from the

Soviet era onwards.

5.2 Territorial Centre of Gravity

The second spatial structure is associated with grid

tiles that exhibit high averages for the foreground

network index (Figure 5).

These identified grid tiles correspond to the

territorial centre of gravity, highlighting the most

efficient networks at the regional level in terms of

accessibility — networks that bolster regional

economic development. The shift of the territorial

centre of gravity from the south to the centre of the

region is the main change observed between the pre-

Soviet and post-Soviet periods (Figure 5).

Two complementary interpretations can be made

regarding this spatial evolution. The first pertains to

geopolitical changes within the region: during the

pre-Soviet period, the southern part of the coastal

region was under Prussian control and was an integral

part of East Prussia, with Königsberg (now

Kaliningrad) as its capital (Figure 9a). The

establishment of the Lithuanian state after World War

I led to a severance of the historical connections with

present-day Kaliningrad oblast. The designation of

Klaipėda as the administrative centre of the coastal

region (currently known as ‘Klaipėda apskritis’)

marks a significant shift in the network’s centre of

gravity toward the heart of this coastal region.

A second interpretation pertains to the economic

transformations that have occurred since the 20

century. During the pre-Soviet period, the region’s

economy relied on agriculture, with the most

developed networks corresponding to areas of intense

agricultural activity. The presence of individual and

family farms as local centres (section 5.1) supports

this notion. However, in the post-Soviet era, the

economy underwent a substantial transformation,

realigning itself around industrial and service sectors.

Consequently, the network’s centre of gravity is now

found in the primary industrial and tourist hotspots,

Modelling the Road Network as an Expression of Historical Spatial System Change in the Klaipeda Region (Lithuania)

Figure 6: Spatial evolution of networks relating to secondary centres in transition between the (a) pre-Soviet, (b) Soviet, and

specifically the port of Klaipėda and the seaside

resorts of Palanga and Šventoji.

It is also noteworthy that these economic shifts are

intricately linked to the development of specific

transport infrastructures. In this respect, motorways

have played a crucial role in defining the network’s

centre of gravity in the coastal region (Figure 5c). The

first key route (east-west) is the ‘A1’ motorway,

which connects Klaipėda (on the Baltic Sea) with

Odesa (on the Black Sea) and Moscow. According to

the Soviet military maps from 1985, the motorway’s

construction was incomplete, aside from the sections

near Klaipėda. The second major route (south-north)

is the ‘A13’ motorway, which links Klaipėda and the

northern part of the coastal region with Latvia. The

third route (east-west) establishes a connection

between Palanga and the regional centre of Šiauliai

(outside the study area) via the ‘A11’ motorway,

passing through Kretinga. The development of these

routes has further influenced the local hierarchy of the

network, as evidenced by the Betweenness index

during both the Soviet and post-Soviet periods

(section 4).

The trends for these first two spatial structures

mirror those identified by Strano et al. (2012) in the

suburbs of Milan, showcasing a transformation of the

territory from rural areas to urban-industrial zones,

and eventually to metropolitan regions.

5.3 Secondary Centres in Transition

The third spatial structure is associated with grid tiles

that exhibit, on average, low to medium background

network indexes when compared to other spatial

structures (Figure 6). Across all periods analysed, the

average standardised indexes for the spatial structure

of the main local centres consistently exceed those of

this structure: 4.15 compared to 0.66 for NQPD, 2.72

compared to 1.02 for the Length, and 4.57 compared

to 0.32 for Betweenness.

This spatial structure denotes networks that are in

a ‘transition phase’ towards main local centres

(section 5.1) or peripheral areas (section 5.4). For

instance, during the post-Soviet period, this cluster

was situated in peri-urban zones surrounding key

urban centres like Klaipėda or Kretinga (Figure 6c),

which can be classified as secondary centres.

GISTAM 2025 - 11th International Conference on Geographical Information Systems Theory, Applications and Management

Figure 7: Spatial evolution of networks related to peripheral territories between the (a) pre-Soviet, (b) Soviet, and (c) post-

Soviet periods.

5.4 Peripheral Territories

The fourth spatial structure is characterised by grid

tiles exhibiting low average indexes, irrespective of

the type of network (Figure 7).

The grid tiles identified within this spatial

structure are linked to the peripheral networks at the

regional scale. For instance, during the post-Soviet

period (Figure 7c), these peripheral networks were

found in the agricultural and semi-natural areas of the

Niemen delta, as well as the riparian zones along the

Niemen River, delineating the southern border with

Russia. In comparison to the pre-Soviet era, contem-

porary agricultural practices are more modern but less

intensive, reflecting a trend of agricultural decline

(Lekavičiūtė and Gadal, 2009; Spiriajevas, 2014).

Moreover, the motorway infrastructure, which

facilitates more efficient access to the network, is

confined to the northern part of the coastal region.

This overall marginalisation of the southern region

may partially account for the observed decline in the

foreground network index.

During the pre-Soviet period (Figures 7a and 9a),

it is noteworthy that the peripheral networks

corresponded to territories under Russian control.

Figure 8: Patterns of spatial organisation between Prussia

and Russia in pre-Soviet Lithuania.

Two distinct patterns of spatial organisation

appear to emerge between the two countries (Figure

8): one explanation for this could relate to the

Modelling the Road Network as an Expression of Historical Spatial System Change in the Klaipeda Region (Lithuania)

Figure 9: Spatiotemporal modelling of spatial structures relating to road network in the Klaipeda region between (a) the pre-

Soviet period, (b) the Soviet period, and (c) the post-Soviet period.

differing timelines of agrarian reform. In Prussia, the

reform aimed at developing the familial farming

model was implemented earlier (in the early 19

century) compared to Russia, where the reform was

introduced later (in the late 19

century), and the

traditional street village model was retained (Jespen

et al., 2015).

6 CONCLUSIONS AND

PERSPECTIVES

This research analyses the evolution of spatial

structures and territorial organisation in the Klaipėda

region since the 20

century, utilising the Generic

City concept to model road networks.

By distinguishing between foreground and

background networks with indexes tailored for

regional analysis, this study measures the relationship

between the spatial configuration of space and its

associated functions, thereby illustrating various

processes at play within the Klaipėda territory.

The significant territorial restructuring observed

during the pre-Soviet and post-Soviet periods is

particularly evident in the integration of local urban

centres along the coastal region, driven by the

dynamics of metropolisation. This metropolisation

was accompanied by a shift in the region’s economic

function from agriculture to industry and services.

The development of motorway infrastructure further

bolstered this territorial growth, enhancing regional

accessibility while marginalising agricultural areas

that once formed the core of the territory.

Additionally, the collectivisation reforms and

geopolitical shifts following the disintegration of the

former province of East Prussia contributed to this

marginalisation.

In this study, the Generic City indexes —

modified and adapted for territorial analysis —

proved effective in highlighting the interconnected

political, economic, social, and cultural factors that

influence the spatial and territorial organisation of the

Klaipėda region.

The reliability of the modelling was partly

validated by the alignment of results with existing

scientific literature on this region. Incorporating

supplementary data, such as built-up areas, land use,

and land cover, could further strengthen the model if

similar trends are identified. However, the indexes

associated with the Generic City concept, modified

GISTAM 2025 - 11th International Conference on Geographical Information Systems Theory, Applications and Management

for regional analysis, can only be generalised through

examination in other territorial contexts exhibiting

comparable dynamics. In this respect, the Baltic Sea

region presents intriguing opportunities, particularly

considering the rapid, intense and complex impacts of

political, economic, social, and cultural changes on

spatial and territorial organisation.

In this regard, the historical application of

generative models remains constrained when they are

trained on overly general databases. The approach

developed in this research, which utilises

mathematical graph models, has been tailored from

generic indexes to conduct a thorough analysis of

existing structures at a territorial scale. This method

considers the complexities of the Baltic region,

characterised by the interplay of various cultural

influences, including Prussian, Russian, Soviet, and

post-Soviet elements.

From a methodological standpoint, this research

enables a connection between historical spatial

structures and specific territorial functions, such as

agricultural activities. Furthermore, spatiotemporal

modelling of road network structures could serve to

enhance missing or incomplete historical data in

retrospective modelling studies.

ACKNOWLEDGEMENTS

This Research is supported by the Science Council of

Lithuania's “Next Generation of Multi-Objective

Satellite Image Recognition Algorithms for Climate

Monitoring” program (LMT NEUTRINO) and the

CNES TOSCA THRISNA ISEULT.

REFERENCES

Al Sayed, K., Turner, A., Hanna, S. (2009). Cities as

emergent models: the morphological logic of

Manhattan and Barcelona. In Proceedings of the 7

International Space Syntax Symposium. Royal Institute

of Technology (KTH), Stockholm.

Allain, R. (2004). Morphologie urbaine. Géographie,

aménagement et architecture de la ville. Armand Colin.

Abida, H. (2018). Syntaxe Spatiale comme un outil

d’analyse de l’espace architectural. International

Journal of Innovative Technical and Applied Sciences,

2(1), 22-30.

Archaeomedes. (1998). Des oppida aux métropoles.

Economica Anthropos.

Badariotti, D. (2005). Des fractales pour l’urbanisme ?

Quelques pistes de réflecion à partir de l’exemple de

Strasbourg-Kehl. Cahiers de géographie du Québec,

49(137), 133-156.

Barrington-Leigh, C., Millard-Ball, A. (2017). The world’s

user-generated road map is more than 80% complete.

PloS one, 12(8).

Bučas, J. (2001). Kraštotvarkos pagrindai. Technologija.

Cirtautas, M. (2015). Changing periphery of the Baltic

cities: Lithuanian case. Journal of Architecture and

Urbanism, 39(1), 56-69.

Cooper ; C., Chiaradia, A. (2020). sDNA: 3-D spatial

network analysis for GIS, CAS, Command Line &

Python. Software X 12, 100525.

Cooper, C. (2024). Spatial Design Network Analysis

(sDNA) Version 4.1 Manual. Cardiff University.

Retrieved from http://sdna.cardiff.ac.uk/sdna/software/

documentation

Dringelis, L., Ramanauskas, E., Povilaitienė, I.,

Mačiukėnaitė, J. (2015). Exploration and respectation

of the spatial structure of cities, towns, townships and

villages as a significant formant of their identity.

Journal of Architecture and Urbanism, 39(1), 79-100.

Elissalde, B. (2004). Spatial structure. Hypergeo. Retrieved

from https://hypergeo.eu/spatial-structure/?lang=en

Gadal, S. (2011). Métropolisations, territoires émergents et

systèmes géographiques d’information et de

modélisation territoriale [Habilitation à Diriger des

Recherches]. Université Paris Diderot Paris 7.

Georges, P., Verger, F. (2013). Dictionnaire de la

géographie, 4

édition. Presses universitaires de France.

Gloaguen, T., Zaleckis, K., Gadal, S. (2023). Development

of new indexes of the ‘Generic City’ in the Baltic

coastal city network. In XXIX International Seminar on

Urban Form ISUF 2022: Urban Redevelopment and

Revitalisation. A Multidisciplinary Perspective. Łódź

University of Technology Press, 365-378.

Hillier, B. (2007). Space is the machine: a configurational

theory of architecture. Space Syntax.

Hillier, B. (2010). The Need for A Spatial Ontology for

Societies (The Space Syntax Workshop). The European

Archaeological Association Conference, The Hague,

Amsterdam.

Idzelis, A. (1984). Industrialisation and population change

in the Baltic republics. Lituanus, 30(2).

Jespen, M.R., Kuemmerke, T., Müller, D., Erb, K.,

Verburg, P.H., Haberl, H., …, Reenberg, A. (2015).

Transitions in European land-management regimes

between 1800 and 2010. Land use policy, 49, 53-64.

Köll, A. M. (2024). Reinterpretation of economic history in

the Baltic countries, 30 years after independence.

Scandinavian Economic History Review, 1-10.

Laouar, D., Mazouz, S. (2017). La carte axiale, un outil

d’analyse de l’accessibilité spatiale : cas de la ville

d’Annaba. Synthese : Revue des Sciences et de la

Technologie, 35, 111-123.

Lekavičiūte, J., Gadal, S. (2009). Forest Cover Changes

during 1975-2005 in the Lithuanian Coastal Region.

Rural Development 2008, 241.

Peponis, J., Bafna, S., Zhang, Z. (2008). The connectivity

of streets: reach and directional distance. Environment

and Planning B: Planning and Design, 35(5), 881-901.

Piškinaitė, E., Veteikis, D. (2023). The Results of

Digitizing Historical Maps: Comparison of Lithuanian

Modelling the Road Network as an Expression of Historical Spatial System Change in the Klaipeda Region (Lithuania)

Land-Use Structure in the 19

and 21

Centuries. Land,

12(5), 946.

Plakans, A. (2011). A concise history of the Baltic states.

Cambridge University Press.

Ragauskaitė, A. (2019). Ethnocultural regionalization of

Lithuania or all the truth about our regions. Acta

Museologica Lithuanica, 4, 162-168.

Spiriajevas, E. (2014). Hindrances and suggestions for

sustainable development of Lithuanian Coastal Strip

(Zone). Regional Formation and Development Studies.

6(1), 125-136.

Strano, E., Nicosia, V., Latora, V., Porta, S., Barthélémy,

M. (2012). Elementary processes governing the

evolution of road networks. Scientific reports, 2(1),

296.

Ubarevičienė, R. (2018). City systems in the Baltic states:

the Soviet legacy and current paths of change. Europa

Regional, 25(2), 15-29.

Ubarevičienė, R., Van Ham, M. (2017). Population decline

in Lithuania: who lives in declining regions, and who

leaves? Regional Studies, Regional Science, 4(1), 57-

79.

Vanagas, J., Krišjane, Z., Noorkoiv, R., Staniunas, E.

(2002). Planning urban systems in Soviet times and the

era of transition: the case of Estonia, Latvia and

Lituania. Geographia Polonica 75(2), 75-100.

Vaitkus, G., Vaitkuvienė, D. (2005). Land cover changes in

the Lithuanian coastal zone during 1975-2000. Acta

Zoologica Lituanica 15(2), 183-187.

Veteikis, D., Šabanovas, S. Jankauskaitė, M. (2011).

Landscape structure changes on the coastal plain of

Lithuania during 1998-2009. Baltica, 24(2), 107-116.

Voiron-Canicio, C. (1995). Analyse spatiale et analyse

d’images par la morphologie mathématique. GIP

Reclus.

GISTAM 2025 - 11th International Conference on Geographical Information Systems Theory, Applications and Management