LUC - Land Uptake Control

A GIS-based Approach

Elena De Santis

and Bernardino Romano

Department of Information Engineering, Computer Science and Mathematics (DISIM), Center of Excellence DEWS,

University of L’Aquila, 67100 L’Aquila, Italy

Department of Civil and Environmental Engineering (DICEAA), University of L’Aquila, 67100 L’Aquila, Italy

Keywords:

Land Uptake Control, Geographical Information System.

Abstract:

Land use caused by urbanisation is one of the main causes of political and social conﬂicts and altered envi-

ronmental quality of land. It is a widespread feeling that the zero-balance objective (i.e. new areas can be

urbanized only if already urbanized areas are restored) is feasible, at least in developed country, where the

population is stabilizing around constant values. But it is very important to control the transient, between the

current almost unrestricted situation and the zero-balance regime. Therefore, this paper proposes land uptake

control procedures, based on geographical information systems and remote sensing.

1 INTRODUCTION

In the broader framework of sustainability challenges,

land uptake surely plays an important role. Advanced

technologies inﬂuence urban developments and, at the

same time, they can support required cultural, social,

economic and politic adjustments and can make ef-

fective control procedures, needed to mitigate the im-

pact of human activities on the environment (see e.g.

(Lynette I. Millett and Deborah L. Estrin, 2012)).

According to consolidated scientiﬁc opinions,

land use caused by urbanization is one of the main

causes of political and social conﬂicts (Plotkin,

1987)and altered environmental quality of land (Ellis

and Ramankutty, 2008), (Sala et al., 2000).

The aspects involved, either directly or indirectly,

in urban conversion of land include the following:

Economic and energetic ﬁeld:

• diseconomies in transport.

• waste of energy;

• reduced agricultural produce.

Hydro-geo-pedologic ﬁeld:

• geological destabilization;

• irreversible use of land;

• alteration of underground and surface water-

courses.

Physico-climatic ﬁeld;

• increased thermal reﬂection and climate change;

• reduced emission absorption capacity;

• effects on carbon sequestration;

• spatial propagation of physico-chemical distur-

bances.

Eco-biological ﬁeld

• physical erosion and destruction of habitats;

• ecosystem fragmentation;

• dystrophy of ecological and biological pro-

cesses;

• penalization of the ecosystem services of the en-

vironment;

• reduction in overall ecological “resilience”.

The awareness of the need of land uptake con-

trol is shared in the scientiﬁc and social community

(EEA, 2006). To cite an example, in (RNE, 2008) it

is stated that the German Federal Government’s goal

is to reduce the land being claimed for settlement and

transport use to 30 hectares per day by 2020. When

last assessed, the new use of land for settlement and

transport stood at 113 hectares per day (trend 2003—

2006). The achievement of the 30-hectare goal in the

practice of urban development is in fact primarily a

matter for the L

ander and the municipalities. The

Federal Government is, however, active in providing

support through the programmes for funding urban

development which are geared towards the inner de-

velopment of towns and municipalities, through re-

search programmes and last but not least through leg-

islative measures ((RNE, 2008), (Henger and Bizer,

2010), (Siedentop and Fina, 2010)).

It is a widespread feeling that the zero-balance ob-

jective (i.e. new areas can be urbanized only if already

450

De Santis E. and Romano B..

LUC - Land Uptake Control - A GIS-based Approach.

DOI: 10.5220/0004487004500456

In Proceedings of the 10th International Conference on Informatics in Control, Automation and Robotics (ICINCO-2013), pages 450-456

ISBN: 978-989-8565-70-9

 2013 SCITEPRESS (Science and Technology Publications, Lda.)

urbanized areas are restored) is feasible, at least in de-

veloped country, where the population is stabilizing

around constant values. But it is very important to

control both the transient evolution, between the cur-

rent almost unrestricted situation and the zero-balance

regime, and the transient duration, to allow the adap-

tation to the new paradigm for the current social and

economic model, which is instead strongly dependent

on the land transformability.

In this paper we propose a land uptake control pro-

cedure, that involves the main actors (municipalities,

regional authorities), at different level of responsibil-

ity. This control procedure could be implemented by

deﬁning appropriate normative tools, by establishing

technical staffs with the task of system monitoring

and of data base maintaining, and by establishing an

efﬁcient information management infrastructure.

The main idea is that of introducing the output-

feedback paradigm in the environmental planning

process. While closed loop control is a standard tool

in most technological disciplines, it is not so in social

management systems, in general. This is mainly due

to the complexity of the systems, whose behavior is

in most case unpredictable with sufﬁcient accuracy,

to the difﬁcult in obtaining data and sometimes to the

cultural inadequacy of the system managers (politi-

cians, administrative and technical staff). In fact the

seminal book (McLoughlin, 1969) remains an im-

portant reference for a system approach to the urban

and environmental planning, but, at the best of our

knowledge, this conceptual formulation of the prob-

lem has not so far produced an effective information

system, for a dynamic plan management. Nowadays

the things are dramatically changing: the availability

and penetration in the public administration of Ge-

ographical Information Systems (GIS), the availabil-

ity of remote sensing techniques and powerful hard-

ware and communication infrastructures (WEB), to-

gether with advances in systems and control theory

make innovative procedures feasible ((Sharma et al.,

2012),(Thompson and Prokopy, 2009)). Anyway the

complexity of the systems remains unchanged, and

hence a possible approach is that of splitting the over-

all system into different interacting sub-systems, each

characterized by simple models. The level of abstrac-

tion of each model has to be obviously appropriate for

the problem to be solved.

In the system we are considering humans are in

loop, and the social behavior has not been modeled.

Therefore the resulting control system is not auto-

matic, but it can offer an important formal frame-

work to support the decision process. This control

scheme could be integrated in an environmental De-

cision Support System (eDSS) together with all the

monitoring, management and communication tools.

The goal of the procedure we propose is twofold:

minimization of the total amount of land uptake,

while assuring the per capita standard of urbanized

areas, and maximization of the aggregation of these

areas. In fact the urban sprawl is another phenomenon

to be controlled, in order to mitigate the above deﬁned

aspects, with a particular attention to diseconomies in

transport and waste of energy. See (Glaeser, 2011),

for a vibrant discussion on the advantage of a dense

city, with respect to energy consumptions, sharing of

amenities and so on, even at popular level.

Instead of giving hard constraints on the land

uptake amount, our proposal is based on a reward-

penalty approach, by giving a monetary value to the

untransformed land. In this way each municipality

has ﬂexibility in planning the environmental trans-

formations, but has also the awareness that the deci-

sions have to be the result of an optimization prob-

lem, where the land uptake has its cost. We think that

the ﬂexibility and the graduality are the keys to allow

the practical implementation of environmental control

policies, and to reduce possible social conﬂicts.

The idea of planning the future developments, step

by step, on the base of actual measured data, but in the

framework of given objectives, could cause a change

of perspective in urban planning itself, as it is tradi-

tionally considered, at least in those countries (e.g.

Italy) where a process control on the current plan is

not actuated, after the approval procedures. We do

not dwell here on this question, but a parallel analysis

is in progress.

The paper is organized as follows: the land uptake

problem is deﬁned in Section 2. In Section 3, three

possible control schemes are introduced. Section 4

considers the sprawl control problem. In Section 5 a

discussion is offered, with possible extensions of the

approach.

2 LAND UPTAKE

The issue of excessive urban conversion of natural

land has been raised at European level only in recent

years, mainly through the initiatives undertaken by

the European Environmental Agency (EEA, 2006) to

analyze this phenomenon and its consequences. See

Fig.1 and Fig 2, for some revealing data. See also (Pi-

leri and Maggi, 2010) and (Romano and Zullo, 2013a)

for analysis of urbanization in Italy.

Furthermore, it is worth recalling that the Euro-

pean Commission, based on a need to tackle soil

productivity, risks to human health and the environ-

ment, and to provide opportunities for climate mit-

LUC-LandUptakeControl-AGIS-basedApproach

451

Figure 1: Annual grow of land uptake in some municipali-

ties (EEA, 2006).

Figure 2: Data on urbanisation in Western Europe (Romano

and Zullo, 2013b).

igation and adaptation as well as stimulating busi-

ness opportunities for soil remediation, proposed a

Soil Framework Directive in 2006, which amends di-

rective 2004/35/EC (on environmental liability with

regard to the prevention and remedying of environ-

mental damage) and reafﬁrms the status of “non-

renewable resource” for this particular territorial com-

ponent and the need for its conservation. The Euro-

pean Parliament adopted its ﬁrst reading on the pro-

posal in November 2007 by a majority of about two

thirds. At the March 2010 Environment Council, a

minority of Member States blocked further progress

on grounds of subsidiarity, excessive cost and admin-

istrative burden. No further progress has since been

made by the Council. The proposal remains on the

Council’s table.

The 2006 EEA report states that: all available ev-

idence demonstrates conclusively that urban sprawl

has accompanied the growth of urban areas across

Europe over the past 50 years. This is shown from

a recent European perspective. The areas with the

most visible impacts of urban sprawl are in coun-

tries or regions with high population density and eco-

nomic activity (Belgium, the Netherlands, southern

and western Germany, northern Italy, the Paris re-

gion) and/or rapid economic growth (Ireland, Portu-

gal, eastern Germany, the Madrid region). Sprawl is

particularly evident where countries or regions have

beneﬁted from EU regional policies.

When we speak about land uptake, more precisely

we consider ”urbanized soil”, in accord to the follow-

ing classiﬁcation (Romano and Zullo, 2013a):

• Built-up Land: surfaces covered by buildings and

identiﬁable through the ground projection of the

perimeter of the latter. Natural soil has been re-

moved completely in order to build foundations

• Urbanized Soil: land used for urban functions,

involving the replacement or retention of natural

soil: it includes built-up land and land used for

ancillary settlement functions, such as public and

private gardens, sports facilities, unpaved roads

and other service areas, either permeable or im-

permeable to water

• Artiﬁcial Land: areas where the natural layer has

been replaced by other materials, either permeable

or impermeable to water, to allow different uses:

it includes parts of built-up land, but also streets,

squares and parking lots (paved or permeable), as

well as sports ﬁelds or excavation areas

• Sealed Soil: surfaces covered by layers of im-

permeable material preventing the absorption of

surface water. It includes built-up land and land

used for other purposes that require paving, such

as streets, squares and parking lots and all those

cases where the natural soil layers are removed

entirely and replaced by other materials that im-

prove the stability and indeformability of sur-

faces.

3 LUC TECHNIQUES

We consider the case of a region, divided into a num-

ber of municipalities. We assume that the control ob-

jective is deﬁned at the regional level, but the deci-

sions are taken at the municipal level. The regional

authorities have the responsibility of system monitor-

ing and control.

Let i ∈

{

1..N

}

be the index for the municipali-

ties. The integer k denotes a time step. The physi-

cal time between k and k + 1 mainly depends on the

adjourning rate of geographical data (typically one or

more years between two successive remote sensing

actions). For simplicity we will call ”time” the in-

dex k. Let x

(k) be the amount of urbanized areas in

municipality i at time k, u

(k) the amount of new ar-

eas, which will be urbanized between time k and time

ICINCO2013-10thInternationalConferenceonInformaticsinControl,AutomationandRobotics

452

k + 1, as planned by municipality i at the beginning

of time period k. Finally, let δ

(k) a ”disturbance”

such that u

(k)+δ

(k) is the actual urbanized areas be-

tween time k and time k+1, as measured at the end of

kth period of time. The disturbance δ

(k) models the

deviations with respect to the planned behavior (e.g.

unauthorized building). Obviously δ

(k) can be neg-

ative, and in this case it models delays in the execu-

tion of planned intervention or the fact that some pre-

viously urbanized areas have been de-urbanized (e.g.

caves re-naturalization). The value u

(k) + δ

(k) can

be negative, too. The value x

(k) comes as the result

of a measurement at time k.

In the ﬁrst model we assume that the central-

ized controller deﬁnes for each municipality the up-

per bound of the urbanized areas, with a deﬁned time

horizon T . At time T a new upper bound will be de-

ﬁned and so on. This upper bound could be computed

on the base e.g. of demography, within clusters of

homogeneous territorial units (municipalities, in our

cases). See (Romano and Zullo, 2013a) for a compu-

tation of the per capita land uptake, called hereafter

urbanization standard or simply standard, denoted by

σ. The reward or the penalty is computed step by step

on the base of the deviation of the actual urbanized

value with respect to the upper bound. The goal is that

of reaching a ratio between urbanized areas and pop-

ulation which is close, and possibly below, the stan-

dard. Therefore the bound is an important starting

point in the control process, and the decisions about it

have to be shared in the involved social communities.

The second model considers possible interactions

among the municipalities, given the bounds and the

time horizon: in this model a municipality can buy

the right of urban transformation from another mu-

nicipality. This one is in our opinion a very impor-

tant mechanism, because it introduce the concept that

avoiding unnecessary transformations gives quantiﬁ-

able advantages. On the other side transforming be-

yond the initially given bound has an additional cost,

and hence the decision about it has to be carefully

taken.

Finally this last model is reformulated in a reced-

ing horizon framework (see e.g. (Camacho and Bor-

dons, 1995) and (Soeterboek, 1992) for introductory

readings on this theme), where the bounds on the ur-

banized areas are dynamically adjourned, taking into

account the predicted population after H steps.

In this paper we assume that the standard σ is time

invarying, but in order to meet the zero-balance objec-

tive it could be possible to consider a weighted stan-

dard, by means of a decreasing function: e.g. σ(k) =

σ(0)exp(−λk)+

σ(1 − exp(−λk)), where σ(0) is the

current standard, λ > 0 and

σ is the minimal standard,

which is compatible with a predicted social and eco-

nomic scenario. Determining such a minimal stan-

dard is not an easy task: some research effort is in

progress on this point, but we are not aware of any

established result.

3.1 Single Municipality Model

Given the horizon T , let x

the upper bound for ur-

banized areas in municipality i, computed by con-

sidering the standard and the expected population

at time T . The state of the system is x(k) =



(k) x

(k) . . . x

(k)



. The state equation

(k + 1) = x

(k) + u

(k) + δ

(k) (1)

k = 0, ..., T − 1, i = 1..N

and let the input be

(k) = F

(k)



− x

(k)



, if x

≥ x

(k)

(k) = 0, if x

< x

(k)

with F

(k) ∈ [0, 1]. The condition x

< x

(k) means

that the current degree of urbanization is above the

standard, with respect to the future estimated popu-

lation. This could be due or to a decreasing trend in

population or to the fact that the urbanization is above

the standard with respect to the current population and

to the expected population. Consider moreover the

cost function

= c

∑

k=1



− x

(k)



with c > 0. The value c



− x

(k)



is a ”reward” for

being below the bound, or a penalty, for being above

the bound.

We assume that the centralized controller knows

(k), i = 1...N, k = 0...T , and x

(see Fig. 3).

Step by step, the value c



− x

(k)



if positive has

to be considered as a reward, if negative is a penalty

to be paid. The value F

(k) in equation (1) is a pa-

rameter each municipality can decide by its own, to

distribute in time the transformability power: the idea

is that of gradually reducing the transformed areas, in

order to make this process feasible and socially sus-

tainable. Therefore it is a tool to plan the future ac-

tions, on the base of the deviation of the measured

state with respect to the bound.

3.2 Multi-municipalities Model

In this case we assume that the municipalities can buy

the transformability right from other municipalities,

LUC-LandUptakeControl-AGIS-basedApproach

453

Figure 3: Single-municipality model. R is the central con-

troller (regional authority). Mi is the ith municipality.

with a price determined by the market. Therefore we

need another state variable, x

(k), which is the resid-

ual amount of urbanizable area at time k, which is ini-

tialized at time 0 with the value x

, calculated as the

preceding section. The centralized controller knows

(k) (from the measurements) and x

(k) (communi-

cated by the municipality i), i = 1...N, k = 0...T . The

state equations become:

(k + 1) = x(k) + u

(k) + δ

(k) (2)

(k + 1) = x

(k) +

∑

j=1

i j

(k)x

(k) −

∑

j=1

i j

(k)x

(k)

(k) = x

k = 0, ..., T − 1, i = 1..N

where

i j

(k) ≥ 0, η

i j

(k) ≥ 0

(k) = 0, η

(k) = 0

0 ≤

∑

j=1

i j

(k) ≤ 1

i j

(k) = ρ

(k)

and it is reasonable to assume that

i j

(k)η

i j

(k) = 0

i.e. it is not possible for municipality i to buy and to

sell the urban transformation right, at the same time,

with an interaction with the same municipality j. See

Fig. 4 .

The input is

(k) = F

(k)



(k) − x

(k)



, if x

(k) ≥ x

(k)

(k) = 0, if x

(k) < x

(k)

and

= c

∑

k=1



(k) − x

(k)



Figure 4: Multi-municipalities model. Dashed arrows de-

note interactions among municipalities.

3.3 Multi-municipalities Model with

Receding Horizon

Consider a modiﬁed version of multi-municipalities

model, with the additional input h

(k), and with un-

bounded time horizon

(k + 1) = x(k) + u

(k) + δ

(k) (3)

(k + 1) = x

(k) +

∑

j=1

i j

(k)x

(k)+

−

∑

j=1

i j

(k)x

(k) + h

(k)

(k) = x

k = 0, 1, 2..., i = 1..N

Let σ be the standard and p

(k) the value of pop-

ulation at time k + H, as predicted at time k. Such a

value will be in general a function of the population

records of the municipality i, in the time interval 0...k,

denoted by p

[0,k]

. In symbols,

(k) = f

( p

[0,k]

)

The value h

(k) depends on p

(k) and on p

(k −1),

as deﬁned in the next Table.4.

Let ∆

(k) = x

(k) − x

(k), ∆

(k) = p

(k) −

(k − 1). Then

if ∆

(k) < 0 and ∆

(k) < 0 h

(k) = 0

if ∆

(k) ≥ 0 and ∆

(k) < 0 h

(k) = −γ

if ∆

(k) < 0 and ∆

(k) ≥ 0 h = σ∆

(k)

if ∆

(k) ≥ 0 and ∆

(k) ≥ 0 h = σ∆

(k)

(4)

where

γ = min



−σ∆

(k), ∆

(k)



See Fig. 5 for a pictorial description.

ICINCO2013-10thInternationalConferenceonInformaticsinControl,AutomationandRobotics

454

Figure 5: Multi-municipalities model with receding hori-

zon.

The input is

(k) = F

(k)



(k) − x

(k)



, if x

(k) ≥ x

(k)

(k) = 0, if x

(k) < x

(k)

and

= c

∑

j=1



( j) − x

( j)



4 SPRAWL CONTROL

The cost function deﬁned in the above models can

take into account also the problem of sprawl control.

Let σ

(k) some sprawl index (e.g. as deﬁned in (Ro-

mano, 2004)), of municipality i at time k. Let s be

a positive real, so that s ∗



(k − 1) − σ

(k)



, if pos-

itive, is a reward because the sprawl index decreased

in time period k, with respect to k − 1. Then the cost

function is deﬁned as:

∑

j=1



c ∗



( j) − x

( j)



+ s ∗



(k − 1) − σ

(k)



Moreover the function J could take into account also

other phenomena, related with the sprawl, and shape-

dependent. We do not dwell here on this point.

5 DISCUSSION

In this paper we described a land uptake control pro-

cedure. It is a ﬁrst attempt to introduce the output-

feedback paradigm in the urban planning process. In-

stead of giving hard constraints on the land uptake

amount, our proposal is based on a reward-penalty

approach, by means of a suitable cost function, that

gives a monetary value to the untransformed land.

We considered three different models: the ﬁrst one

is the simpler, but it is rather strict, since the bound

of the transformable land is given at the beginning

for each municipality, and it remain ﬁxed, for all the

ﬁxed, given a priori time horizon. The second model

consider the possibility of interaction among the mu-

nicipalities. In this case the bounds evolve in time,

and there is an adaptation mechanism, with respect to

the different needs of municipalities, while the over-

all regional urbanization bound is met. The last model

introduces the adaptation of the bounds also with re-

spect to the population dynamics, in the framework of

receding horizon philosophy. In this case the proce-

dure has the advantage of remaining effective for an

unbounded time horizon. Moreover, changes in the

bounds due to population variation are gradually con-

sidered, ad hence too hard variations are avoided (as

could happen for the ﬁrst and the second model, when

at time T the parameters are reset, to start a new con-

trol period) with obvious advantages.

The models are very simple, but some work has to

be done for the parameters identiﬁcation. Urbaniza-

tion standard are to be evaluated and shared in the sci-

entiﬁc and social communities. The time duration of

each time period has to be deﬁned, in relation with the

availability of adjourned geographical data, time hori-

zons have to be deﬁned, with respect to time constants

of the involved processes, and ﬁnally the parameters

in the cost function has to be carefully designed, in or-

der to produce the desired control effects. The simu-

lation of the models, with the construction of possible

scenarios could be useful.

In all the models we considered, the regional au-

thority has the task of maintaining the information on

the state of the system (or a ”land register”), super-

vising the process and computing the cost functions

for each municipality. But its control role could be

emphasized, if one consider the possibility of incen-

tivizing exchanges of transformation rights between

municipalities, as in the second and third model, when

there is some overall interest in the transaction. Only

to give an example, suppose there is municipality i

that wants to buy from municipality j some more ad-

ditional transformable area, needed for a new factory

settlement. Municipality j could agree to sell this

right, because, for geographic proximity the factory

will be an opportunity also for its citizens. On the

other side municipality j could own natural resources,

which furnish ecosystem services to municipality i, as

well as to others municipalities. Hence this kind of

transaction could be incentivized by regional author-

LUC-LandUptakeControl-AGIS-basedApproach

455

ity.

In the framework we deﬁned, a number of addi-

tional problems can be considered. For example in

the third model the current urbanization bound is ad-

journed with respect to the expected future popula-

tion. But the population level could in some case de-

pend on the actual planned new urbanization. Hence

there is a modelling problem to be solved. As an opti-

mization example problem, we can consider the case

in which a municipality wants to maximize its util-

ity, by deciding the amount of transformation rights

to sell (or to buy) in each time period, and their price,

known the state of the system and the population

trend.

Finally, at this stage, the LUC procedure is just a

proposal, based on technical ideas and analysis. In

order to become effective, normative and legal frame-

work should be faced. This task is beyond the scope

of this paper.

REFERENCES

Camacho, E. and Bordons, C. (1995). Model predictive

control in the process industry. Springer Verlag, New

York.

EEA (2006). European commission, urban sprawl in eu-

rope, the ignored challenge. Directorate general Joint

Researche Center, Copenaghen.

Ellis, C. and Ramankutty, N. (2008). Putting people in the

map: anthropogenic biomes of the world. Frontier in

Ecology and the Environment: 6.

Glaeser, E. (2011). Triumph of the City: How Our Greatest

Invention Makes Us Richer, Smarter, Greener, Health-

ier, and Happier. The Penguin Press.

Henger, R. and Bizer, K. (2010). Tradable planning per-

mits for land-use control in germany. Land Use Pol-

icy, 27(3):843–852.

Lynette I. Millett and Deborah L. Estrin (2012). Computing

Research for Sustainability. The National Academies

Press.

McLoughlin, J. (1969). Urban and regional planning: A

systems approach. Faber and Faber.

Pileri, P. and Maggi, M. (2010). Sustainable planning? ﬁrst

results in land uptakes in rural, natural and protected

areas: the lombardia case study (italy). Journal of

Land Use Science, 5(2):105–122.

Plotkin, S. (1987). Property, policy and politics: towards a

theory of urban land-use conﬂict. IJURR, 11(3):382–

404.

RNE (2008). Germany federal government, for a sustain-

able Germany. Progress Report 2008 on the National

Strategy for Sustainable Development.

Romano, B. (2004). Environmental fragmentation ten-

dency: the sprawl index. In ERSA 2004 Congress,

Porto, Portugal.

Romano, B. and Zullo, F. (2013a). Land urbanization in

central italy: 50 years of evolution. Journal of Land

Use Science. DOI:10.1080/1747423X.2012.754963.

Romano, B. and Zullo, F. (2013b). Models of urban land

use in europe assessment tools and criticalities. In-

ternational Journal of Agricultural and Environmen-

tal Information Systems. To appear.

Sala, E., Chapin, F., Armesto, J., Berlow, E., Bloom-

ﬁeld, J., Dirzo, R., Huber-Sanwald, E., Huenneke,

L., Jackson, R., Kinzig, A., Leemans, R., Lodge, D.,

Mooney, H., Oesterheld, M., Poff, N., Sykes, M.,

Walker, B., Walker, M., and Wall, D. (2000). Global

biodiversity scenarios for the year 2100. Science,

287(5459):1770–1774.

Sharma, L., Pandey, P. C., and Nathawat, M. (2012). As-

sessment of land consumption rate with urban dynam-

ics change using geospatial techniques. Journal of

Land Use Science, 7(2):135–148.

Siedentop, S. and Fina, S. (2010). Monitoring urban sprawl

in Germany: towards a gis-based measurement and

assessment approach. Journal of Land Use Science,

5(2):73–104.

Soeterboek, R. (1992). Predictive control- a uniﬁed ap-

proach. Prentice Hall International.

Thompson, A. and Prokopy, L. S. (2009). Tracking urban

sprawl: Using spatial data to inform farmland preser-

vation policy. Land Use Policy, 26(2):194–202.

ICINCO2013-10thInternationalConferenceonInformaticsinControl,AutomationandRobotics

456