Prospecting the Evolution of Winegrowing Region through MAS Modelling Giovanni FUSCO, Matteo CAGLIONI Université de Nice Sophia Antipolis / CNRS UMR 7300 ESPACE ICCSA 2014, International Conference on Computational Science and Its Applications June 30 - July 3, 2014, University of Minho, Guimaraes, Portugal.
23
Embed
Prospecting the Evolution of Winegrowing Region through MAS Modelling - ICCSA2014
French rural landscapes are deeply marked by the presence of winegrowing. Urban sprawl around French cities is nowadays exerting growing pressure on the vineyards in several regions. This is the case of the winegrowing region of Bandol, on the Mediterranean coast.
Local authorities and winegrowers are increasingly aware of the risks associated with uncontrolled continuation of present trends. But interactions among wine-growing and new urbanization are today even more complicated and encompass several factors as land ownership, social networks among wine-growers, entrepreneurial demography, production constraints, capital transfers from land development to wine-growing and road accessibility. Prospecting the evolution of winegrowing in the Bandol region in face of urban pressure is thus a challenging task, which can greatly benefit from geosimulation techniques.
Spatial strategic foresight is exploratory and does not determine a certain future for the Bandol wine-growing system. The starting point is the simulation of a trend scenario, where urban pressure, wine-growing economic parameters and land-use constraints take the most plausible values according to expert knowledge. Alternative scenarios, partially departing from the trend, are later elaborated by modifying coherently the parameter set of the geo-simulations.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Prospecting the Evolution of Winegrowing Region through MAS Modelling
Giovanni FUSCO, Matteo CAGLIONI
Université de Nice Sophia Antipolis / CNRS UMR 7300 ESPACE
ICCSA 2014, International Conference on Computational Science and Its ApplicationsJune 30 - July 3, 2014, University of Minho, Guimaraes, Portugal.
Content Overview
1. Introduction
2. Methodology
3. Geosimulation
4. Conclusion
AOP perimeter: 1650 ha planted in 2010 (RGA), 95% in 4 municipalities
A wine-growing region between Marseilles and Toulon
1941: Obtention of the AOP Label
N
The Bandol wine-growing region
N
0 5km
The socio-economic context
• The wine-producing world : wine-growers / wine-makers, the Bandol Association (ODG), the INAO
• The urban world : land developers, households, municipal governments, metropolitan government
• Growing concern from ODG and INAO on the consequences of municipal/metropolitan governments plans on the future of wine-growing
… social demand for a prospective research on wine-growing landscape in the Bandol region
Exploring the future of vineyard landscapes
• A research action within the PATERMED project (ANR Systerra): exploring the future of vineyard landscapes in the Bandol region
• Urban / Rural interaction: urban pressure on agricultural land + financial input from urban development into winegrowing
• Importance of agent strategies and interactions resulting in (and being shaped by) spatial structures
Modelling urban / rural interaction
• Modeling choices : MAS approach + Sensitivity Analysis for scenario building
• Which elements and process are to be included in the model? What implementation within a MAS approach?
• What outputs are relevant in a spatially informed strategic foresight?
Two-phase modelling process:1) Model development and exploration for a small virtual world (INPUT 2012, Cagliari)
2) Model implementation and validation with real-world data(ICCSA-GEOANMOD 2014)
Input - Output Model Structure
MODELMODEL
cadastral map
land use
Land use plan
wine growers’ structure
future cadastral map
future land use
future wine growers’ structure
NetLogo
exogenousvariables:-housing demand-urban land price-price of wine-interest rate-…
Wine-grower agents:lifecycle and social relationships
friendshiphostility
At the end of the active life (65years), how many children want to take over parent’s activity?
0Sell property orSharecropping
1Property
transmission
2+Property division
age capital
Different contexts for land sale
1.From winegrowers at the
end of their career without children who want to take
over the activity
2.From winegrowers without
residual financial capital and negative revenues
3.From winegrowers with parcel(s) under urban pressure and urbanizable
The winegrower agent sell (all) their parcels and exits the system
(eventually sharecropping agreement)
The winegrower agent sells one or more parcels and can buy
other parcels further away
To other winegrower(s) To real estate promoters
Winegrower’s economic function
= α S - t K - f ( S , dc ) - f ( c )Bandol produce capital cost logistic costs fix costs
(including cost of living)
K capitalS surface [ha]dc distance from the logistic centerα value of net winegrowing produce per hat interest rate
independentcooperative
independentcooperative
independentcooperative
∆Cumulatedcapital
Flow diagrams of processes in the modelProcess of parcel purchase Process of parcel sale
Process of property transmission
Process of land development through individual housingProcess of land development through subdivision
The Bandol wine-growing region in 2010
Land-Use in 2010 :Built up areas, vineyards and agriculture in the centre, natural areas in the periphery
The Bandol wine-growing region in 2010
Land-Ownership : 53 independent winemaking domains and 120 big cooperative winegrowers already control most of the agricultural land within the Bandol perimeter
Trend Scenario after 40 years
Strong urban growth,stable Bandol wine prices
• Leapfrogging of urban growth in agricultural and natural land
• Vineyards are quantitatively stable, but move away outwards to colonise agricultural and natural land
• Increase of small cooperative wine-growers and big domains
Increased land development after 40 years
Increased urban pressure in the form of new subdivision
Dwindling agriculture
• Discontinuous urban fabric overtakes agricultural land and non-Bandol vineyards
• Bandol winegrowing is boosted by capital transfers from land developers, but migrates to more peripheral areas
• Land prices within the Bandol perimeter stay high, domains prosper
Winegrowing crisis after 40 years
Reduced retail price of Bandol wine and attractiveness of winegrowing for the new generations
• Vineyards shrink, other crops expand and move outwards
• Strong contraction of natural areas
• Land prices diminish
• Winegrowers and domains don’t diminish, as they diversify in other crops
Scenarios overview
6 scenarios obtained through coherent combination of 3 factors in the MAS simulations:
• In most scenarios, winegrowing is able to adapt to external pressure, thanks to financial transfers from land development and to the complementary strategies of wine-growers /makers.
• Vineyards migrate outwards
• Even in crisis scenario winegrowing shrinks only marginally.
Conclusion: what the model can do
• Dialectics between agent micro-behaviours and emergence of meso- and macro- spatial structures
• Role of basic social and economical interaction within vineyards development
• Role of land ownership structure and demographic variables within vineyards and urban development
Under the constraint of a set of hypotheses, the model produces spatial forms for possible futures of the winegrowing economy of the Bandol region.
• Impacts of urban pressure, land-use plans and exogenous economic variables on vineyards landscapes
Conclusion: what the model cannot do
• Land-use plans and urban dynamics are exogenous
• Very limited modeling of interaction with different agricultural activities
• Topological / accessibility properties of road networks are not modeled
• Landscape modeling is limited to land-use quantification (no qualitative aspects like perceptions, heritage conservation, nor 3D rendering of landscapes)
Conclusion: the role of a MAS model in Spatial Strategic Foresight
• Not a prediction model, but a tool to explore the possible futures of the landscape under coherent sets of hypotheses
• Integration of socio-economic and spatial processes, beyond the black box of CA modeling
• But : danger of over-complexifying process modeling
• A precious tool to understand the role of key variables and policies in a spatially informed strategic foresight
Within the limits of lean modeling, MAS become useful companions for scenario building (mix of expertise on external parameter and simulation of agents’ interactions)