Computers, Environment and Urban Systems 40 (2013) 46–55

Contents lists available at SciVerse ScienceDirect

Computers, Environment and Urban Systems

journal homepage: www.elsevier.com/locate/compenvurbsys

Developing a 3D cadastre for the administration of urban land use:

A case study of Shenzhen, China
Renzhong Guo a,b, Lin Li a,⇑, Shen Ying a, Ping Luo c, Biao He b, Renrong Jiang c
a
School of Resource and Environmental Science, Wuhan University, 129 Luoyu Road, Wuhan 430079, China
b
Urban Planning, Land and Resources Commission of Shenzhen Municipality, 518034 Shenzhen, China
c
Shenzhen Centre for Assessment and Development of Land and Estate, 518034 Shenzhen, China

a r t i c l e i n f o a b s t r a c t

Article history: Fast economic growth and rapid urbanization present a challenge to the limited urban land resources of
Available online 24 August 2012 China. Specifically, they encourage a shift in the concentration on the surface of land parcels to the space
above and below them. Recent development in land utilization in developed cities has broken the tradi-
Keywords: tional frame of the parcel-based or 2D cadastre. The three-dimensional (3D) cadastre is now accepted by
3D cadastre researchers as an effective technical means to support the administration of space. Taking Shenzhen, a
Cadastral survey quite developed city in China as a case study, this paper presents an ad hoc paradigm to develop a 3D
3D data model
cadastral system for the management of land space use in it. The demand for a 3D cadastre is shown
Land administration
Rights to land
through two actual instances of land space exploitation. Relying on current legislation, administration,
Property and 3D technology, we design a full 3D cadastral system and embed it in the current cadastral system
so that the administration of land space can be realized without an alteration in the frame of cadastral
data or in current administrative procedures. Presenting an application example, we demonstrate the
easy implementation and practicability of our 3D cadastral system. We find, however, that current insti-
tutions require some but not excessive modification when fully implementing it for land administration.
Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction dimensions (Agdas & Stubkjær, 2011; Benhamu & Doytsher,

The management of rapid urbanization, one of the signs and
motors of China’s economic development, has been and is a funda-
mental issue. As a developing country with a very large population,
China’s swift economic development has certainly reduced pov-
erty, even though about 90 million of its people still fall under
the poverty line, earning less than US$ 234 a year (Lu, 2011). How-
ever, a conflict between urban growth and limited land resources
has increasingly emerged; in effect, current urban development,
which takes the form of urban sprawls, is not sustainably afford-
able. Therefore, further urban expansion requires a change in land
use patterns, since the latter are considered critical for promoting
sustainable progress in developing countries (Foley et al., 2005;
Tilman et al., 2001; Turner, Lambin & Reenberg, 2007).
To meet this challenge, well-developed cities such as Shanghai,
Guangzhou, and Shenzhen turned several years ago to the vertical
use of land space, that is, the space above and below surface land
parcels. However, traditional two-dimensional cadastral technol-
ogy has proven to be deficient in the management of these newly
conceived urban spaces; thus, the 3D cadastre is emerging as an
effective means to support the administration of space in three
⇑ Corresponding author. Tel.: +86 13871504963.
E-mail address: lilin@whu.edu.cn (L. Li).
2003; Chong, 2006; Choon, Hussin, & Oon, 2010; Doner et al.,
2010; Hassan, Ahmad-Nasruddin, Yaakop & Abdul-Rahman,
2008; Stoter, Ploeger, Louwman, Oosterom, & Wünsch, 2004).
Shenzhen, our case-study city, is located in the southeastern
coastal region of China, adjacent to Hong Kong, and is one of the
most economically advanced cities of the country; it has experi-
enced rapid and steady economic growth and urbanization.
According to the Shenzhen Statistic Yearbook of 2010, its total pop-
ulation is about 8.9 million and its land area about 1,991 square
kilometers, covering 6 (restructured into 8) districts. The GDP of
Shenzhen has increased 3.75 times since 2000, reaching about
US$ 1,281 billion in 2009. The proportion of primary industry,
which was 37% of the GDP in 1979 and 0.7% in 2000, has decreased
to less than 0.1% today. The per capita GDP is about US$ 14,495, the
highest of all Chinese cities, and 4.3 times the national average.
With such accelerated growth, the urban area (construction acre-
age) has sprawled enormously. The construction acreage of the
core, four district (Futian, Luohu, Nanshan, and Yantian) urban area
of the Shenzhen Special Economic Zone increased by 64% from
1990 to 2000 but by only about 8% from 2000 to 2010. The city’s
outskirts, covered by two districts (Bao-an and Longgang [now
Bao-an, Guangming, Longgang and Pingshan]) became the major
construction area contributor in this last decade. Fig. 1 shows the
pattern in the change of construction areas from 1990 to 2010,

0198-9715/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.compenvurbsys.2012.07.006
 R. Guo et al. / Computers, Environment and Urban Systems 40 (2013) 46–55
Fig. 1. Construction acreage during 1990–2010.
and Fig. 2, which indicates land use in 2010, reveals that the core
urban area has very limited surface available for further
construction.
However, the core urban area is the economic and social nerve
center of the city, and land (space) is always in demand to support
its increasing activities. The use of space under or above its surface
is a rational alternative. Such space must be effectively and consis-
tently managed to prevent legal conflict with the existing cadastre.
To meet this objective, it is best to administer tri-dimensional ur-
ban space with a 3D cadastral system than with the current 2D
cadastral system.
Much research (Agdas & Stubkjær, 2011; Benhamu & Doytsher,
2003; Chong, 2006; Choon et al., 2010; Doner et al., 2010; Hassan
et al., 2008; Paulsson & Paasch, 2011; Sorensen, 2011; Stoter, 2004;
Stoter & Oosterom, 2005; Stoter, Ploeger, Louwman, Oosterom, &
Wünsch, 2011) has discussed the management of 3D property in
cadastral systems and acknowledged that a full 3D cadastre is
the best solution. However, most practical 3D cadastral systems
are configured either with multi-layer parcels, such as those in Is-
rael (Benhamu & Doytsher, 2003), or with 3D property objects reg-
istered on 2D cadastres, such as those in Australia, Norway (Stoter
& Oosterom, 2005), Malaysia (Hassan & Rahman, 2011), and the
Netherlands (Stoter et al., 2011). A full 3D cadastral system has
Fig. 2. The land use of Shenzhen city in 2010.
47
been rarely reported in the literature, and the lack of appropriate
3D toolkits or software systems for implementing the functionality
of 3D in a cadastre is one of the main reasons for its essential ab-
sence. Some 3D tools may be good at 3D modeling but not good
at visualization; some may provide precise visualization but fail
to manage effectively 3D geometrical objects; some packages
may provide effective geometrical computations but still lack mod-
ules to maintain 3D topologies.
This paper provides an ad hoc 3D cadastre solution for the man-
agement of urban land space in Shenzhen city. Its main contribu-
tion lies in the proposal of overall technical settings for a full 3D
cadastral system that is restricted by (1) existing data and available
techniques and (2) the current legal and administrative context.
The remainder of this paper is structured as follows. Its next sec-
tion presents two examples to illustrate that Shenzhen is now in
an administrative transition from surface to land space use; it is
followed by an outline of the context for the development of a
3D cadastre for this administration. Section 4 presents the main
technical issues for the achievement of a 3D cadastral system in
Shenzhen city. These include the primary geometrical hypotheses
for the use of space and the 3D data model and its implementation.
In Section 5, a simple application is presented. Finally, conclusions
are drawn, on the basis of the previous sections and some issues for
future study.
2. From the surface to the spatial use of land
The clarification of ownership and property rights are the keys
to land administration. When the ownership of a piece of land is
defined as all things within a spatial ‘‘cone’’ formed geometrically
from the earth’s center to the sky, ownership can be well specified
by 2-dimensional geometry. A parcel-based cadastral system can
unambiguously handle spatial relationships among various prop-
erty owners and the transitions among themselves and with other
parties. The rights to land within the ‘‘cone’’ are homogeneous and
can be projected onto the unique map layer of the 2D cadastre
without any confusion. With the evolution of society and economy,
land becomes precious, and rights to it must be refined to suit dif-
ferent purposes; thus, the nature of space below or above ground is
48 R. Guo et al. / Computers, Environment and Urban Systems 40 (2013) 46–55

no longer fully consistent with that on the ground. The emerging,

spatially heterogeneous rights within the ‘‘cone’’ break the spatial
homogeneity of land rights; consequently, they shatter the hypoth-
esis on which the parcel-based (2D) cadastre has rested.
In Shenzhen, a break with the framework of conventional land
administration based on the 2-dimensional cadastre occurred once
in 2005, when an underground space was independently leased for
commercial use, separate from the tract of land on its surface
(Shenzhen Special Zone Daily (2, February, 2005), 2005a). Since
then, quite a number of such cases have arisen, overrunning the
original domain of the 2D cadastre. The connotation of the term
land use has altered by the addition of vertical spatial elements,
requiring the upgrade of the cadastre of land administration. Two
examples are given here to show this urgent need.

2.1. Tanglangshan

Tanglangshan is an example of multi-use land space above the
surface, where the underground and on ground construction is a
subway station and that above it a group of residential buildings
(condominiums) (Fig. 3). The subway company owns the station,
and the buildings consist of many privately owned apartments.
Here we have an example of the intensive use of land. The con-
struction, which is designed for both transportation and residence,
is marked by a clear vertical separation. The roof of the station is
about 16 m from the ground, below which space is utilized for sub-
way services; the residential buildings protrude from that of the
station (Fig. 4).
The parcel has two land development proposals (‘‘2011-203-
0002’’ and ‘‘2010-003-0072’’) (T506-0018), shown in Fig. 5. Pro-
posal ‘‘2011-203-0002’’ indicates that the residential buildings oc-
cupy this piece of land, which according to proposal ‘‘2010-003-
0072,’’ is also used for transportation. The competitive use of a par-
cel of land is not allowed in the conventional 2D cadastre, where
any parcel has the exclusive use on its defined planar space.
Recording either one of the land uses on a planar cadastral map
fails to tell the whole story of property rights. To deal with this case
within the frame of a 2D cadastral system, several solutions have
been suggested; in terms of land use, these form three groups: Pri-
mary/Secondary, Multi-Layer, and Combination.
2.1.1. Primary/secondary
One of a parcel’s uses is chosen as primary (for example, resi-
dential in this example) and recorded on the cadastral map; the
other use is noted as secondary (transportation). Proposals for land
development, including construction plans, are logged during the
building process.
Fig. 3. An overall image of construction of Tanglangshan.
Fig. 4. A profile image of the construction.
2.1.2. Multi-layer
Land space is vertically divided into a number of layers, each of
which records only one land use. For example, the cadastre in Israel
is divided into three layers: surface, below surface, and above surface
(Benhamu & Doytsher, 2003). In our Tanglangshan example, the
subway station can be recorded on the surface layer, the residential
buildings on the above surface layer, and the subway tunnel on the
below surface layer. An obvious merit of this solution is that no
change is needed in the current cadastral system beyond the addi-
tion of some equivalent layers. This solution is appropriate for the
existing data model in most GIS systems. However, the number of
layers may be flexible and the means of determining them remains
an open question, pointing to unresolved, complicated situations.
2.1.3. Combination
If the same parcel embraces two or more types of land use,
these may be combined into a new type of land use and all propos-
als for development linked with it. Architectural plans for the par-
cel may be attached so that real spatial configurations and
relations can be reconstructed from them. However, combined
land use (such as residential-transportation) is usually not ac-
cepted by the convention of classification in the current land
administration system. In addition, constructions for different pur-
poses may have different projected planar shapes or extents that
make the partition of parcels for the cadastral map very difficult
or even not applicable.
2.2. Wanxiangcheng
Wanxiangcheng is made up of several plaza buildings in which
international brands are sold. Among these, its two main buildings
are separated by a municipal road and connected by an arched
structure that crosses it (Fig. 6). The buildings are shown on a par-
cel-based cadastral map (Fig. 7). The land space used for the over-
ground arch is represented on this map by its projected-on-plane
parcel, H102-0037(B), which also indicates the spatial extent of
the underground parking lot and the commercial shops. Adjacent
to H102-0037(B), two parcels, H102-0037 and H102-0038, depict
the two main buildings. The buildings, the arch, and the under-
ground lot have the same owner, but the land space of the arch,
a public pedestrian corridor (a kind of easement), belongs to the
municipality. The underground shops above the parking lot are
owned by different individuals. In the current documentation,
plans for the construction are in the land development proposal,
and a text description of the construction is attached to it. Based
on the documentation, the vertical configuration is illustrated by
Fig. 8.
 R. Guo et al. / Computers, Environment and Urban Systems 40 (2013) 46–55 49

Fig. 5. A multi-use of a parcel (use of land space).

Fig. 8. A profile of the construction.

This example shows that it is hard to clarify the complex spatial

relations of property rights on the cadastral map or in the 2D
cadastre. None of the above-mentioned solutions can properly
handle this complicated situation. However, from the view of prop-
erty registration, the presentation of the complicated construction
Fig. 6. The arch over the municipal road.
as 3D property that is pasted onto the parcels is regarded as a good
alternative solution. Such a solution has been adopted in Australia,
Norway (Stoter & Oosterom, 2005), and Malaysia (Hassan & Rah-
man, 2011). A similar resolution was put forward to deal with a
complicated multi-level property in the Netherlands (Stoter et al.,
2011). Linking a planar parcel with the 3D representation of a com-
plex construction is one of the most effective current means to un-
cover the maze of complex constructions under the frame of a 2D
cadastre. However, much research on this issue indicates that any
amendment of the 2D cadastre is only a temporal solution. A 3D
cadastre with full 3D functions is the most effective response to
the technical problems of land administration, regardless of the
legislative challenges.

3. Context for 3D cadastre

3.1. Legal context

3.1.1. Legal base

When land use extends spatially to a vertical from a horizontal
plane, land administration confronts the challenge of determining
if property laws and regulations must be amended. The extent of
Fig. 7. The cadastral map. such amendment is proportional to the 3D cadastre available for
50 R. Guo et al. / Computers, Environment and Urban Systems 40 (2013) 46–55
land space use. China is currently perfecting her legislation and
administration, which offers her room to adapt to new situations.
The current property laws have not be defined in detail so that the
application of a 3D cadastre will not produce large legal problems
or turmoil in land rights registered in land administration systems.
Before 2007, when the Real Right Law of People’s Republic of Chi-
na was issued and took effect (October 1, 2007), no law stated
whether the right to land for construction could be separated from
the surface parcel. The rights to land are founded on the apprehen-
sion of the principles of the parcel-based cadastre, which link the
use of land space under or above ground to land at the surface.
In most cases, the right to use underground space is, in the light
of the Land Administration Law and Construction Law, assigned as
another property right. The Real Right Law is the first to state
clearly and explicitly that rights to use space under or above the
surface parcel can be independent from surface construction. Arti-
cle 136 states that ‘‘the right to use construction land may be cre-
ated separately on the surface of or above or under the land. The
newly-established one may not injure the usufructuary right that
has already been established.’’ Article 138 further states that land
space occupied by buildings, fixtures, and affiliated facilities shall
be contained in a contract with the transfer of rights.
The separation of property rights for construction above and
underground from that on the surface implies that uses of above
and underground spaces may be different from those of the surface
and that the parcel space may be multi-level, across boundary, or
without geometrical limitation. Under this law, ownership of
underground construction may differ from that of surface build-
ings. It provides a good legal basis for local governments to create
their own rules and regulations for land use, ones adapted to par-
ticular situations in land administration systems. Therefore, there
is no legal obstacle to develop a full 3D cadastral system to support
the management of land space use.
3.1.2. Property
The key issue in land administration is the management of var-
ious property or spatial rights. Property rights are often associated
with construction or natural resources. These rights are embodied
in the concept of immovable property, which may have different
national meanings (Kalantari, Rajabifard, Wallace & Williamson,
2008; Stubkjær, 2004) that are largely dependent on legal systems
(Paulsson & Paasch, 2011). Ownership is the crux of property
administration. Some countries—such as the Netherlands, Ger-
many, the United Kingdom, France, and Belgium—define owner-
ship as the right to the ground and of all space above and below
it, including groundwater and fixtures (van der Molen, 2003).
Other countries understand ownership in a way that does not in-
clude mines and groundwater. Some jurisdictions may not allow
separate rights to a parcel from construction on it; for example,
in the Netherlands, ownership of a parcel includes the buildings
and constructions on and in it (Stoter et al., 2011). Others nations,
such as in Demark, accept, through leasing, different ownerships
for land and for buildings; in fact, the formation of a property
‘‘on top of another property’’ can be implemented under a special
procedure (Sorensen, 2011).
China has a quite different juridical system with regard to prop-
erty rights from that of the western countries. According to the
Land Administration Law, any urban land is the property of a unique
owner, the State, and this ownership cannot be altered; others en-
joy ownership of the buildings or other constructions on the land
parcel and legal rights over construction land space. A parcel of
land cannot be registered as property independently from the
buildings on it. Thus, property ownership is focused on the edifices
created in development. A property embodies the ownership of a
building or buildings and the usufruct of land. According to the
current laws in China, buildings may be owned forever, but the
usufruct of land is tenured for some years (40–70 years). This jurid-
ical norm means that any property is composed of two legal items.
Property is assigned attributes that are represented by spatial
features. These features are adopted on the basis of existing tech-
nology, which changes over time. A spatial attribute is character-
ized by a spatial dimension. In the 2D cadastre, the construction
area of a building, the building’s plan, and the location of the parcel
on which it sits are adopted to describe the spatial attribute. In the
3D cadastre, the volume of a property object and its 3D model may
be chosen as an embodiment of the spatial attribute.
3.2. Land administration
3.2.1. Land development
Land development, which in this paper refers to the physical
creation of any new property (real estate) object, is very important
in Chinese land administration. As just indicated, the Land Admin-
istration Law stipulates that ‘‘Land in the urban areas of cities is
owned by the State’’ and cannot be transferred to another party.
The only property right that individuals may exercise over such
land is an entitlement of usufruct for construction, which can be
transferred in accordance with the laws governing purchases, sales,
and transfers. The right to construct is issued by a municipality in
conformity to certain general requirements (termed a ‘‘planning
outline’’) on which the developer bases an architectural design.
The planning outline usually prescribes the use, tenure, geographic
location, and total construction area. An approval routine deter-
mines if the architectural design has been complied according to
the planning outline and can be approved. If approved, the design
is recorded and employed as the reference for a final check and
acceptance of the completed construction. Amendments to the de-
sign are allowed, but these must go through the approval routine.
In this process, the spatial attribute of property is formed
according to the planning outline, and its form is shaped by the
architectural design. Therefore, the depiction of the spatial attri-
bute based on the planning outline and the approval checks are
the two places where the spatial features of property are adopted
and depicted. When upgrading from a 2D cadastre to a 3D cadastre,
the process governing the spatial features may not require change
but the features themselves and their treatment in the process
have to be somewhat altered. A change in these aspects may pro-
duce a crack in current smooth routines. Therefore, filling this
‘‘crack’’ must be considered in developing a 3D cadastre.
3.2.2. Registration
When a construction is completed and accepted in the process of
land development, a new property object is created. The property
will be authorized through a process of registration in which its
ownership, land usufruct, location, and area are recorded and autho-
rized. Although the property can be transferred and some of its attri-
butes may change, its spatial features will not alter during its
lifecycle. Therefore, the presentation of spatial features is an impor-
tant element in its registration. Both the registration and the devel-
opment procedures require a spatial/geometric representation of
property and may use the same representation. The current 2D rep-
resentation, such as plans and drawings, is not adequate to 3D real-
ity (buildings), since it demands an additional mental process to
integrate individual drawings so as envision the physical construc-
tion in grasping full property rights (Stoter et al., 2011). In a 3D
cadastral system, 3D representation of reality is an inevitable choice.
(Fig. 9 depicts an example of a 3D representation of a property.)
3.3. Three dimensional GIS technology
3D GIS techniques, including 3D data capturing, 3D data
management, and 3D data models and visualization have been
 R. Guo et al. / Computers, Environment and Urban Systems 40 (2013) 46–55
Fig. 9. The 3D representation of property: the transparent faces represent the spatial extent of land space use and the building model is drawn within the extent.
substantially studied. Hundreds of references can be found when a
few key words, such as ‘‘3D model’’ and ‘‘3D GIS,’’ are used to
search the Internet. 3D techniques have been extensively applied,
creating many impressive products in the fields of GIS and CAD,
which support various applications. In fields such as urban plan-
ning, architectural design, and engineering, most of required tech-
niques of 3D cadastre systems can be found in the various
applications of 3D GIS. However, only a very few cases in the field
of land administration have so far adopted 3D techniques. Current
3D GISs may present good modules for storage and visualization of
3D models, but they lack effective 3D geometric operators, such as
3D buffering and intersection, which is shown by the very limited
3D geometric computation functionality of commonly used GIS
software, such as ArcGIS. Thus, there is a large potential to utilize
3D techniques to improve management in land administration,
especially in dealing with spatial issues.
Many existing 3D GIS theories, methods, and tools can be uti-
lized to develop a 3D cadastre; however, a deep study of the 3D
model, visualization, and functionality is still required, for these
elements must be tailored and adopted to the administration of
land rights. They offer a foundation for developing a full 3D cadas-
tral system.
4. Developing a 3D cadastral system
4.1. Geometry of 3D property objects
As mentioned previously, the registered property includes
information on both the construction and the parcel. In the case
of the 2D cadastre, the inclusion of an orthographic notion of the
construction in addition to plans in the planar boundary of the par-
cel is an important recorded spatial relation. This spatial relation is
used to check the validity of the construction in the approval rou-
tine. It must be introduced into the 3D cadastre for the same pur-
pose. Thus, geometric computation shows if the construction is
contained within the spatial extent.
In this sense, the geometry of a 3D property consists of two 3D
geometries: a 3D model of the construction and a 3D container (a
term used in the remainder of this paper) that represents the spa-
tial extent of land space used by the construction. Since a 3D model
of construction is included in a container formed by the spatial
51
extent, the spatial relation of a property with other constructions
(of different properties) is replaced by the spatial relation among
the containers of different properties. The geometry of a 3D model
is determined by the architectural configuration of the construc-
tion that may have some influence on land space rights. However,
this influence cannot be explicitly depicted in geometry. For exam-
ple, the gates of the construction will shape the geometry of ease-
ment on neighboring spaces. Regarding applications, the 3D model,
which may take any geometrical shape, is regarded as an indepen-
dent geometry. Building 3D models from plans or drawings may
use complex geometric computations, but many tools (such as
3D Max or SketchUp) are now available and effectively perform
these.
As in the partitioning of land parcels in the 2D cadastre, the par-
titioning of land space in three-dimensional space determines the
geometry of the 3D container and defines the complexity of a full
3D cadastral system. It may be impossible to treat the entire range
of possible 3D geometries, since 3D geometric computations re-
main some disputed issues in the field of computational geometry.
Like many other practical applications, it is rational and reasonable
to place some restrictions on the 3D geometry of containers. The
primary hypotheses of this paper are listed below:
(1) A container is a polyhedron
As with the parcel-based cadastre, no pure curve is included on
its boundary line, which consists of a set of polylines; no purely
curved surface is a part of the surface of the container that is
represented by a set of faces enclosing a limited space. Although
land space use may sometimes have no clear roof boundary, we
may designate the spatial extent of its physical construction
that has a limited space. More importantly, the faces of a poly-
hedron and the edges of the faces must satisfy the generalized
Jordan Curve Theorem that refers to the orientability of these
geometric elements.
(2) The interior of the container is connected, which means that
any container is ‘‘simple’’ and no compound or multiple con-
tainers is allowed. If a container can be divided into two or
more independent containers, each of the latter is treated
as a simple one.
These two hypotheses will rid us of dealing with some very
complex situations. However, we still cannot guarantee the
52 R. Guo et al. / Computers, Environment and Urban Systems 40 (2013) 46–55
avoidance of facing abnormalities when we make geometric com-
putations with polyhedra of this kind.
4.2. Compatible 3D data models
There are many data models for 3D objects in GIS, such those
represented by simplexes (point, line, triangle, and tetrahedron)
(Carlson, 1987), by a 3D Formal Data Structure (FDS) (Molenaar,
1990), by tetrahedronized irregular networks (Penninga, Oosterom,
& Kazar, 2006), by polyhedrons (Arens, Stoter, & van Oosterom,
2005; Stoter, 2004; Wenninger 1974; Zlatanova, 2000), by polyhe-
dral regular polytopes (Thompson, 2007), and by a Constructive
Solid Geometry (CSG) and B-rep approach in computer graphics.
All of those models are used to capture the spatial features of 3D
objects and not their semantic attributes. The semantic data in
the field of land administration are used to regulate and coordinate
relationships among people and property under a given society,
economy, and legal system. A high priority should be placed on
keeping these data consistent when developing a full 3D cadastre
and extending its spatial dimension from two-dimensions. There-
fore, the data model in a 3D cadastral system should be compatible
with the current data model in 2D parcel-based cadastral systems
so that the meanings of the semantic data recorded in the latter
will not change. In addition, the 3D data model with a complete
3D topological structure is regarded as the most advanced imple-
mentation for 3D cadastres (Doner et al., 2010).
Topologically speaking, the data model of the 2D cadastre in-
cludes topological features—faces, edges, and nodes—that are com-
monly used in most of GIS applications. Consequently, a 3D data
model (Guo & Ying, 2010) is designed for the 3D cadastral system
by adding a 3D topological feature-volume to the 2D data model. A
volume (that is, a container or polyhedron) consists of a set of faces
that enclose a 3D space. The data model is operationally structured
with a 3D piecewise linear complex (PLC), a commonly used geo-
metric data structure in computer graphics (Cohen-Steiner, Colin
De Verdiere, & Yvinec, 2004; Miller, Talmor, Teng, Walkington &
Wang, 1996; Si & Gartner, 2005). It matches well with the current
2D data model (Fig. 10).
4.3. Generating 3D model data
Except in their land related spatial aspect, cadastral data in the
3D cadastre are the same as those of current land parcels; they in-
clude, for example, information on ownership, tenure, location, and
so on. There are normal work routines to generate and maintain
those data. Only 3D models for the container and the construction
need to be created for the 3D cadastre.
4.3.1. 3D model of container
So far, the examples reveal that the geometry of land space is
simple: it comprises a prism or a combination of prisms that have
vertical faces and flat up or bottom faces. This simplicity results
from the fact that land space for above or underground construction
Fig. 10. 3D data model with topology (From Guo & Ying, 2010).
is plotted depending on a planar parcel. Tools are available for cre-
ating such prisms by extruding planar parcels or drawings. When
creating a 3D model of a container, its 3D topology will be also cre-
ated and checked so that spatial operations based on the topology
are implemented.
4.3.2. 3D model of construction
3D architectural models are much more complex than those of
containers. The former may have various roofs and irregular sides.
Through more complicated to create, such models are based on
their architectural plans or drawings, and methods for doing so
are found in existing research (Lewis & Sequin, 1998; Lu, Tai,
Bao, Su & Cai, 2005; Mendez et al., 2008; Tanaka, Anthony, Kaneeda
& Hirooka, 2004; Yin, Wonka, & Razdan, 2009). In general, a build-
ing can be automatically modeled by extruding its footprint plan, if
it has a flat or simply shaped roof. For a building with a compli-
cated roof, interactive editing must be inserted in the modeling
process by extruding each edge, according to the profile of the
architecture (Kelly & Wonka, 2011).
Currently, 2D plans or drawings for construction are formatted
in AutoCAD and recorded in the process of land development. The
3D cadastre needs only the 3D shell-model and not the exact 3D
architectural model. Therefore, interactive processing with auto-
matic extruding is designed to generate 3D construction models
from extracted 2D plans by geo-coding the plans for extrusion
(He, Li, Guo & Shi, 2011; Ying, Li & Guo, 2011). A completed 3D
model is individually stored in a spatial database that is embedded
in the whole cadastral system.
4.4. Three dimensional topologies
3D topological model is not a new topic in GIS; substantial dis-
cussion has occurred on different aspects of topological theory,
including their implementation and application. From the perspec-
tive of database management systems, Zlatanova, Rahman and Shi
(2004) provide a brief but informative review of these models and
conclude that application characteristics are essential in choosing a
topological model. Penninga and Oosterom (2008) put forward a
computational model for realizing 3D topology by tetrahedroniza-
tion of 3D objects, based on Poincare’s theory of simplicial homol-
ogy. A conspicuous advantage of this approach is its solid
mathematical foundation. However, its application involves some
risk, since some 3D shapes may not be tetrahedronized. In addi-
tion, the topological relations between 3D objects must be derived
from the topological relations of their tetrahedronization, a step
that involves extra cost to maintain the data of some applications.
Our 3D topological data model, which is drawn from that of the
2D data model, is suited to applications of a full 3D cadastre and is
effective for topological query and analyses, such as the search for
neighboring objects and the incident faces or edges of the object. In
the current implementing phrase of the 3D cadastral system, two
functions are designed for building and maintaining 3D topology:
face-based and volume-based, respectively (Li, Zhao, Guo & He,
2012).
(1) Face-based: 3D source data may come from available 3D
tools (3D MAX or SketchUp), in which all the faces for 3D
objects are prepared but they are discrete and offer no infor-
mation on which compose a volume (container). This func-
tion may suit the case where data can be distributedly
prepared and integrated into a complete topological data
set. (Fig. 11 depicts the organization of discrete faces into
volumes.)
(2) Volume-based: A 3D volume or a 3D object is built and topo-
logical relations among nodes, edges, and faces with the
volume (or object) are recorded. However, the topological
 R. Guo et al. / Computers, Environment and Urban Systems 40 (2013) 46–55
Fig. 11. Organizing faces into volumes by the face-based method.
relations among individual volumes are unknown and need
to be created. This function is suitable for maintaining topo-
logical consistency when some 3D data are updated or a
newly created 3D object is merged into whole 3D data set,
while keeping 3D topology. (Fig. 12 depicts the merging of
an individual volume into the whole 3D configuration of
3D objects.)
4.5. Implementation of 3D cadastral system
4.5.1. System configuration
In order to deal with both the 2D parcel-based cadastre and 3D
spatial use of land, the 3D cadastral system is embedded in the cur-
rent-operational system as a sub-system, which it parallels. The
system configuration is depicted in Fig. 13. 3D data are stored in
the newly created 3D spatial database (3D data), paralleling the
2D planar-parcel database (2D data). Attribute data for the 3D
and 2D cadastres are stored in the same attribute database and
are linked by the OID of 3D objects. The 2D data (parcel data) are
used for the display of the location of 3D properties relevant to
2D parcels in 3D space. In this frame, a uniform interface for data
access is designed for data exchange and communication when
processing these data. Skyline TerraGate (a 3D tool) is used for
3D data viewing, and SketchUp is used for locally editing 3D geom-
etry. The implementation of 3D topology generation and updating
quite evidently depends on robust and effective 3D geometric
computation.
4.5.2. Geometric computation
3D geometric computation, unlike it 2D counterpart, increases
greatly the complexity of computation. First, geometric calcula-
tions have an inherited flaw in that important operators, such as
set intersection and union among 3D objects, are not closed for a
set. Thus, 3D computation must deal with complex situations to
Fig. 12. Merging an individual volume into the whole topological configuration.
53
Fig. 13. System configuration.
produce results in a required form. Second, some geometric pro-
cessing is required during or after calculations. For example, in or-
der to guarantee correct topological relations between edges and
faces, the Jordan Curve Theorem must be employed to check any
newly generated edge with all other edges on the same plane. Con-
sequently, one edge of a face cannot become an edge (boundary) of
another face, even if the edge is on that face. Third, some digital
calculations may also raise problems for sequential geometric
computation. For example, in building topology, node-matching
causes little problem in sequentially forming a polygon in 2D sys-
tems, but it may introduce a serious problem in building 3D topol-
ogies. Fig. 14 depicts such an example, where in merging the nodes
of two faces into one, the nodes are no longer be on a plane and the
normal direction of the face may change.
Therefore, the robustness of geometric computation is an
imposing issue and remains an open question for all fields of 3D
modeling and geometric computations. Any available solution to
it must be application-oriented. Finding a more effective and ro-
bust methodology is always encouraged and inspired by a scien-
tific curiosity.
5. An application example
In the practice of land administration, every routine is designed
according to explicit working specifications in which the items to
be recorded and the order of working steps are predefined and
determined; these cannot be arbitrarily altered. The 3D cadastre
cannot be applied to the whole procedure of land administration
without organizational amendments. With the development of
the city, some local government regulations, such as the ‘‘Provi-
sional Measures on Shenzhen Underground Space Development
and Utilization’’ (Shenzhen Special Zone Daily (11, August, 2005),
2005b) were issued to meet the new above and underground spa-
tial uses of land. They are a complement to the existing 2D cadastre
framework. However, the emerging new technology may be a fresh
impetus to improve institutions so that these will be more suited
to manage urban space. A simple example is given here to show
how the 3D cadastral system supports the management of land
space use in the process of land development.
According to the procedure of the land administration, the auc-
tion listing of land is the initial phase for land development. In the
2D cadastre, the spatial attributes for an auctioned parcel are is
boundary, the construction area, and the planning outline for the
construction. No spatial (3D) elements are considered, even though
these may cause problems. For example, the absence of predefined
roof heights and underground depths may permit a developer to
build very high rooms that may eventually raise taxation issues
or to dig in a manner that impinges on underground infrastruc-
tures, such as pipelines.
In 2011, Shenzhen auctioned a piece of underground land for
parking cars (in central Houhai in the Nanshan district). The
54 R. Guo et al. / Computers, Environment and Urban Systems 40 (2013) 46–55
Fig. 14. Node-matching may alter the normal direction of a face.
Fig. 15. Description of spatial attributes in the published document in auction.
surface land is already designated for public use as ‘‘green land.’’
According to the planning outline, two stories (layers) of about
16,000 m2 will be built underground. With the 3D cadastre system,
its spatial extent can be easily described in the published docu-
ments, together with other geographic information. Fig. 15 shows
the documents on the official bidding website. The 3D representa-
tion provides a clearer and less ambiguous guideline for a devel-
oper to make good use of the land space. These documents will
be recorded and will be followed in the process of land develop-
ment. By applying the 3D cadastral system when the construction
is completed, this 3D data for the underground space can be also
used to check the conformity of the underground construction to
the planning outline.
6. Conclusion
A shift of focus from parcel-based to space-based land reflects
the refinement of property rights management in rapidly develop-
ing urban areas. Although some rights to land may be completely
or partially unclear with respect to space (Bennett, Wallace & Wil-
liamson, 2006), the nature of the rights characterized by spatial
features is crucial in managing and clarifying them. The use of
the vertical space above and below ground rather than horizontally
ruled, surface parcels is the key concept pushing property rights
from a two to a three-dimensional framework. A full 3D cadastre
emerges as an effective and adequate technological means to pro-
mote the management of such land space.
Our case study shows that upgrading the current 2D cadastral
system into a 3D cadastral system is a feasible and effective
management solution to the problem of space in urban contexts.
The case study provides a paradigm for integrating current 3D
techniques into a full 3D cadastral system within a certain context.
Embedded in the existing cadastral system, the data of the 3D
cadastral system is compatible with those of its predecessor and
focuses on 3D modeling and computation. This proposal reduces
the disruptive impact on the existing system and data. Although
very simple, the application example in Section 5 indicates the fea-
sibility of the 3D cadastre for land development under current
administrative procedures. An obvious merit of this approach is
the prevention of future conflicts. Fewer conflicts and contraven-
tions will arise with clearer and more explicit spatial land use.
The application of the 3D cadastre will prevent the misunderstand-
ings inherent in a multi-vocal text description or requirements,
which tend to mask problems that will emerge with time.
However, the full implementation of the 3D cadastre requires
some organizational change. Our experience shows that a gap ex-
ists in the current administrative framework. The departments of
architectural design and land planning are both involved in devel-
opment. The current, well-matched 2D-based routines will cause
some problems in applying 3D representations; for example, some
extra labor will be required to adapt to them. However, our case
study has shown that this gap can be bridged if an effective tool
(e.g., the 3D cadastral system) is provided. In addition, the robust-
ness of 3D geometric computation is also a key issue for the 3D
cadastre, since some important functions rely on it. Otherwise, it
will restrict the application of the 3D cadastre in the administra-
tion of land space. Furthermore, more complex situations—under-
ground water and mining—must be taken into consideration. It is
worthwhile exploring a more effective methodology to deal with
 R. Guo et al. / Computers, Environment and Urban Systems 40 (2013) 46–55
complex geometry. These issues are the subjects of our future
study.
Overall, social evolution will bring about a proliferation of prop-
erty rights, restrictions, and responsibilities (RRR), a trend that de-
mands their explicit clarification and refined management. These
RRR are typically embodied in property objects, which are embed-
ded in 3D space. Therefore, a cadastre based on 3D/4D technology
will be the future of land administration (Bennett, Rajabifard,
Kalantari & Wallace, 2010).
Acknowledgements
The work was supported by Special Fund for Land Resource
Scientific Research in the Public Interest (No. 201111009). The
authors would like to acknowledge anonymous reviewers for their
constructive comments to this article.
References
Agdas, V., & Stubkjær, E. (2011). Design research for cadastral systems. Computers,
Environment and Urban Systems, 35, 77–87.
Arens, C., Stoter, J., & van Oosterom, P. (2005). Modelling 3D spatial objects in a geo-
DBMS using a 3D primitive. Computers & Geosciences, 31(2), 165–177.
Benhamu, M., & Doytsher, Y. (2003). Toward a spatial 3D cadastre in Israel.
Computers, Environment and Urban Systems, 27(4), 359–374.
Bennett, R., Wallace, J., & Williamson, I (2006). Managing rights, restrictions and
responsibilities affecting land. In Combined 5th trans tasman survey conference
and 2nd Queensland spatial industry conference, Cairns, Australia.
Bennett, R., Rajabifard, A., Kalantari, M., & Wallace, J. (2010). Cadastral futures:
building a new vision for the nature and role of cadastres. FIG congress 2010:
Facing the challenges – Building the capacity, Sydney, Australia.
Carlson, E. (1987). Three-dimensional conceptual modeling of subsurface
structures. In Proceedings of the Auto-Carto 8, (pp. 336–345). ACSM/ASPRS,
Falls Church, Virginia.
Chong, S. (2006). Towards a 3D cadastre in Malaysia: An implementation evaluation.
Master’s Thesis, Delft University of Technology, The Netherlands.
Choon, T. A. N. L., Hussin Sr, K. B., & Oon, E. K. H. (2010). 3D property situation in
Malaysia-initiatives towards 3D cadastre. Building, 11–16.
Cohen-Steiner, D., Colin De Verdiere, E., & Yvinec, M. (2004). Conforming delaunay
triangulations in 3D. Computational Geometry, 28(2–3), 217–233.
Doner, F., Thompson, R., Stoter, J., Lemmen, C., Ploeger, H., van Oosterom, P., et al.
(2010). 4D cadastres: First analysis of legal, organizational, and technical
impact – With a case study on utility networks. Land Use Policy, 27, 1068–1081.
Foley, J. A., DeFries, R., Asner, G. P., Barford, C., Bonan, G., Carpenter, S. R., et al.
(2005). Global consequences of land use. Science, 309(5734), 570–574.
Guo, R., & Ying, S. (2010). 3D cadastre analysis and data delivery. Journal of China
Land Science, 24(12), 45–51 (in Chinese).
Hassan, M., & Rahman, A. (2011). Unique identifier for 3D cadastre objects
registration. The 2nd international workshop on 3D cadastres, 2011, Delft, the
Netherlands.
Hassan, M., Ahmad-Nasruddin, M., Yaakop, I., & Abdul-Rahman, A. (2008). An
integrated 3D Cadastre–Malaysia as an example. The international archives of the
photogrammetry (pp. 121–126), Remote Sensing and Spatial Information
Sciences, vol. 37.
He, B., Li, L., Guo, R., & Shi, Y. (2011). 3D Topological reconstruction of
heterogeneous buildings considering exterior topology. Geomatics and
Information Science of Wuhan University, 36(5), 579–583 (Chinese).
Kalantari, M., Rajabifard, A., Wallace, J., & Williamson, I. (2008). Spatially referenced
legal property objects. Land Use Policy, 25(2), 173–181.
Kelly, T., & Wonka, P. (2011). Interactive architectural modeling with procedural
extrusions. ACM Transactions on Graphics, 30(2), 14:11–14:15.
Lewis, R., & Sequin, C. (1998). Generation of 3D building models from 2D
architectural plans. Computer-Aided Design, 30(10), 765–779.
Li, L., Zhao, Z., Guo, R., & He, B. (2012). 3D Topological Construction For Spatial
Physical Object. Geomatics and Information Science of Wuhan University, 37(6),
719–723 (Chinese).
55
Lu, P. F. (2011). People’s daily overseas edition (31, March, 2011), <http://
paper.people.com.cn/rmrbhwb/html/2011-03/31/content_783172.htm?div=-1>
Accessed 12.20.11.
Lu, T., Tai, C., Bao, L., Su, F., Cai, S., et al. (2005). 3D reconstruction of detailed
buildings from architectural drawings. Computer-Aided Design and Applications,
2(1–4), 527–536.
Mendez, E., Schall, G., Havemann, S., Fellner, D., Schmalstieg, D., & Junghanns, S.
(2008). Generating semantic 3D models of underground infrastructure. IEEE
Computer Graphics and Applications, 28(3), 393–402.
Miller, G. L., Talmor, D., Teng, S. H., Walkington, N., & Wang, H. (1996). Control
volume meshes using sphere packing: Generation, refinement and coarsening.
In Proceedings of 5th international meshing roundtable, 1996 (pp. 47–61).
Pittsburgh, PA.
Molenaar, M. (1990). A formal data structure for 3D vector maps. In Proceedings of
EGIS’90. Amsterdam, 1990 (pp. 770–781). The Netherlands.
Paulsson, J., Paasch, J. M. (2011). 3D property research – A survey of the occurrence
of legal topics in publications. The 2nd international workshop on 3D Cadastres,
2011, Delft, the Netherlands.
Penninga, F., Oosterom, P., & Kazar, B. (2006). A tetrahedronized irregular network
based dbms approach for 3d topographic data modeling. Progress in Spatial Data
Handling, 2005 (pp. 581–598).
Penninga, F., & Oosterom, P. (2008). A simplicial complex-based DBMS approach to
3D topographic data modelling. International Journal of Geographical Information
Science, 22(7), 751–779.
Shenzhen Special Zone Daily (2, February, 2005). (2005a). The first underground
space was independently leased for commercial use. <http://www.southcn.
com/news/dishi/shenzhen/ttxw/200502020202.htm> Accessed 12.20.11.
Shenzhen Special Zone Daily (11, August, 2005). (2005b). Provisional measures on
shenzhen underground space development and utilization, <http://sztqb.
sznews.com/html/2008-08/11/content_294721.htm> Accessed 2011.12.20.
Si, H., Gartner, K. (2005). Meshing piecewise linear complexes by constrained
Delaunay tetrahedralizations. In Proceedings of the 14th international meshing
roundtable, 2005, (pp. 147–163). Springer.
Sorensen, E. M. (2011). 3 Dimensional property rights in Denmark, 3D property
design is working – Visualization not. The 2nd international workshop on 3D
cadastres, 2011 (pp. 521–530). Delft, The Netherlands.
Stoter, J. E. (2004). 3D cadastre. PhD’s thesis. Delft University of Technology, the
Netherlands.
Stoter, J., Ploeger, H., Louwman, W., Oosterom, P. v., & Wünsch, B. (2011).
Registration of 3D situations in land administration in the Netherlands. The
2nd international workshop on 3D cadastre, 2011, (pp. 149–166). Delft, the
Netherlands.
Stoter, J., & Oosterom, P. (2005). Technological aspects of a full 3D cadastral
registration. International Journal of Geographical Information Science, 19(6),
669–696.
Stubkjær, E. (2004). Cadastral modeling: Grasping the objectives. In Proceedings of
joint FIG commission 7 and COST action G9 workshop on standardization in the
cadastral domain, Bamberg, Germany.
Tanaka, M., Anthony, L., Kaneeda, T., & Hirooka, J. (2004). A single solution method
for converting 2D assembly drawings to 3D part drawings. Computer-Aided
Design, 36(8), 723–734.
Thompson, R. J. (2007). Towards a rigorous logic for spatial data representation. PhD’s
Thesis, Delft University of Technology, the Netherlands.
Tilman, D., Fargione, J., Wolff, B., D’Antonio, C., Dobson, A., Howarth, R., et al. (2001).
Forecasting agriculturally driven global environmental change. Science,
292(5515), 281.
Turner, B. L., Lambin, E. F., & Reenberg, A. (2007). The emergence of land change
science for global environmental change and sustainability. Proceedings of the
National Academy of Sciences, 104(52), 20666–20671.
van der Molen, P. (2003). Institutional aspects of 3D cadastres. Computers,
Environment and Urban Systems, 27(4), 383–394.
Wenninger, M. J. (1974). Polyhedron models. Cambridge: Cambridge University
Press.
Yin, X., Wonka, P., & Razdan, A. (2009). Generating 3D building models from
architectural drawings: A survey. IEEE Computer Graphics and Applications,
29(1), 20–30.
Ying, S., Li, L., & Guo, R. (2011). Building 3D cadastral system based on 2D survey
plans with SketchUp. Geo-spatial Information Science, 14(2), 129–136.
Zlatanova, S. (2000). 3D GIS for urban development. PhD’s Thesis, Graz University of
Technology, Austria.
Zlatanova, S., Rahman, A. A., & Shi, W. (2004). Topological models and frameworks
for 3D spatial objects. Computers & Geosciences, 30(4), 419–428.