Presenter/Author Information

Marie Castellazzi
I. Brown
L. Poggio
Alessandro Gimona

Keywords

vector, spatio-temporal, land use, stochastic model, uncertainty

Start Date

1-7-2012 12:00 AM

Description

Ecosystem services (ES) assessments are relying upon scenarios to explore the impacts of potential land use changes [UK NEA 2011], and to explore the possible land use pathways leading to desirable ES targets (using a normative approach). The scenario framework acknowledges the high uncertainty inherent to land use changes [Goodwin and Wright 2010] and a stochastic approach to land use modelling would provide some useful support. Moreover, for ES assessment, land use modelling needs to consider that i) some services are dependent upon fine land use mosaics (e.g. pollination, water quality), ii) decision-making is linked to spatial units, and iii) stakeholders should be involved in scenario development [MEA 2006]. To support these requirements, a vector representation of the landscape linked to decision’s units (e.g. parcels) is deemed relevant; the recent availability of datasets such as LCM2007 [Morton et al. 2011] can support land use modelling at fine spatial scale on national extent. This paper discusses the relevance of the LandSFACTS modelling tool [Castellazzi et al. 2010a, Castellazzi et al. 2010b] to meet the above requirements. The model provides potential land use allocations meeting all user-defined spatio-temporal constraints on land uses, for every spatial unit and for every time step. It is based on vector landscape, with each polygon being part of nested or overlapping groups representing decision units (land managers or administrative boundaries). The model is limited by its regular time steps, fixed polygon boundaries, and categorical definition of land uses. However, within those limits, its stochastic and rule-based process allows the exploration of the variability in spatial configurations, and thus provides the means to quantify spatial uncertainties. The spatial variability analyses can provide some useful support to the identification of bottlenecks; i.e. regions where constraints or policy might have higher difficulty to be implemented due to the spatial characteristics of the environment. Scenario examples at multiple scales from Scotland-wide to a sub-catchment are presented to discuss the use and limitations of the modelling approach for representing land use complexity for ES assessments.

Share

COinS
 
Jul 1st, 12:00 AM

Modelling land use change and its spatial variability for ecosystem services assessments

Ecosystem services (ES) assessments are relying upon scenarios to explore the impacts of potential land use changes [UK NEA 2011], and to explore the possible land use pathways leading to desirable ES targets (using a normative approach). The scenario framework acknowledges the high uncertainty inherent to land use changes [Goodwin and Wright 2010] and a stochastic approach to land use modelling would provide some useful support. Moreover, for ES assessment, land use modelling needs to consider that i) some services are dependent upon fine land use mosaics (e.g. pollination, water quality), ii) decision-making is linked to spatial units, and iii) stakeholders should be involved in scenario development [MEA 2006]. To support these requirements, a vector representation of the landscape linked to decision’s units (e.g. parcels) is deemed relevant; the recent availability of datasets such as LCM2007 [Morton et al. 2011] can support land use modelling at fine spatial scale on national extent. This paper discusses the relevance of the LandSFACTS modelling tool [Castellazzi et al. 2010a, Castellazzi et al. 2010b] to meet the above requirements. The model provides potential land use allocations meeting all user-defined spatio-temporal constraints on land uses, for every spatial unit and for every time step. It is based on vector landscape, with each polygon being part of nested or overlapping groups representing decision units (land managers or administrative boundaries). The model is limited by its regular time steps, fixed polygon boundaries, and categorical definition of land uses. However, within those limits, its stochastic and rule-based process allows the exploration of the variability in spatial configurations, and thus provides the means to quantify spatial uncertainties. The spatial variability analyses can provide some useful support to the identification of bottlenecks; i.e. regions where constraints or policy might have higher difficulty to be implemented due to the spatial characteristics of the environment. Scenario examples at multiple scales from Scotland-wide to a sub-catchment are presented to discuss the use and limitations of the modelling approach for representing land use complexity for ES assessments.