A. Veldkamp

Predicting land-use change

Land use change modelling, especially if done in a spatially-explicit, integrated and multi-scale manner, is an important technique for the projection of alternative pathways into the future, for conducting experiments that test our understanding of key processes in land use changes. Land-use change models should represent part of the complexity of land use systems. They offer the possibility to test the sensitivity of land use patterns to changes in selected variables. They also allow testing of the stability of linked social and ecological systems, through scenario building. To assess current progress in this field, a workshop on spatially explicit land-use/land-cover models was organised within the scope of the Land-Use and Land Cover Change project (LUCC). The main developments presented in this special issue concern progress in: 1) Modelling of drivers of land-use change; 2) modelling of scale dependency of drivers of land use change; 3) modelling progress in predicting location versus quantity of land-use change; 4) the incorporation of biophysical feedbacks in land-use change models.

From land cover change to land function dynamics: A major challenge to improve land characterization

Land cover change has always had a central role in land change science. This central role is largely the result of the possibilities to map and characterize land cover based on observations and remote sensing. This paper argues that more attention should be given to land use and land functions and linkages between these. Consideration of land functions that provide a wide range of goods and services makes more integrated assessments of land change possible. The increasing attention to multifunctional land use is another incentive to develop methods to assess changes in land functions. A number of methods to quantify and map the spatial extent of land use and land functions are discussed and the implications for modeling are identified based on recent model approaches in land change science. The mixed use of land cover, land use and land function in maps and models leads to inconsistencies in land change assessments. Explicit attention to the non-linear relations between land cover, land use and land function is essential to consistently address land change. New methods to map and quantify land function dynamics will enhance our ability to understand and model land system change and adequately inform policies and planning.

A spatial explicit allocation procedure for modelling the pattern of land use change based upon actual land use

Modelling of land use changes as a function of its biophysical and socio-economic driving forces provides insights into the extent and location of land use changes and its effects. The CLUE modelling framework is a methodology to model near future land use changes based upon actual and past land use conditions. This paper describes how changes in land use are allocated in the model. A statistical analysis of the quantitative relationships between the actual land use distribution and (potential) driving forces or proxies of these forces underlies the allocation procedure. Based upon thus derived multiple regression equations, areas with potential for increase or decrease in cover percentage of a certain land use type are identified. Actual allocation is modified by autonomous developments and competition between land use types. A multi-scale approach is followed to account for the scale dependencies of driving factors of land use change. This approach provides a balance between bottom-up effects as result of local conditions and top-down effects as result of changes at national and regional scales. The modelling approach is illustrated with examples of scenario simulations of land use change in Ecuador.

Spatial autocorrelation in multi-scale land-use models

In several land use models statistical methods are being used to analyse spatial data. Land use drivers that best describe land use patterns quantitatively are often selected through (logistic) regression analysis. A problem using conventional statistical methods, like (logistic) regression, in spatial land use analysis is that these methods assume the data to be statistically independent. But, spatial land use data have the tendency to be dependent, a phenomenon known as spatial autocorrelation. Values over distance are more similar or less similar than expected for randomly associated pairs of observations. In this paper correlograms of the Moran’s I are used to describe spatial autocorrelation for a data set of Ecuador. Positive spatial autocorrelation was detected in both dependent and independent variables, and it is shown that the occurrence of spatial autocorrelation is highly dependent on the aggregation level. The residuals of the original regression model also show positive autocorrelation, which indicates that the standard multiple linear regression model cannot capture all spatial dependency in the land use data. To overcome this, mixed regressive–spatial autoregressive models, which incorporate both regression and spatial autocorrelation, were constructed. These models yield residuals without spatial autocorrelation and have a better goodness-of-fit. The mixed regressive–spatial autoregressive model is statistically sound in the presence of spatially dependent data, in contrast with the standard linear model which is not. By using spatial models a part of the variance is explained by neighbouring values. This is a way to incorporate spatial interactions that cannot be captured by the independent variables. These interactions are caused by unknown spatial processes such as social relations and market effects.

Simulation of changes in the spatial pattern of land use in China

This paper presents a model for simulating country-wide changes in the land use pattern of China. It is based upon an empirical analysis of the spatial distribution of land use types in China which takes into account socioeconomic as well as geophysical variables. The empirical analysis indicates that a reasonably complete description of the land use distribution can be made by including demographic, soil-related, geomorphological and climatic variables. A multi-scale approach is followed to capture top-down as well as bottom-up factors affecting land use allocation. Competition between different land use types determines which changes will actually take place. The most important land use conversions in China, caused by urbanization, desertification and afforestation, are simulated for a scenario based upon a trend analysis of present land use dynamics. The spatially explicit results allow an analysis of the consequences of a decrease in cultivated area and related production capacity. A preliminary analysis shows that the average production capacity of the lost arable lands is somewhat less than the average production capacity of all agricultural lands together.

Three‐dimensional landscape process modelling: the effect of DEM resolution

Many landscape models have been developed over the past decades; however, relatively little is known about handling the effects of changing spatial and temporal resolutions. Therefore, resolution effects remain a factor of uncertainty in many hydrological and geomorphological modelling approaches. In this paper we present an experimental multi-scale study of landscape process modelling. Emphasis was laid on quantifying the effect of changing the spatial resolution upon modelling the processes of erosion and sedimentation. A simple single process model was constructed and equal boundary conditions were created. Using artificial digital elevation models (DEMs) eliminated effects of landscape representation. The only variable factors were DEM resolution and the method of flow routing, both steepest descent and multiple flow directions. Our experiments revealed an important dependency of modelled erosion and sedimentation rates on these main variables. The general trend is an increase of erosion predictions with coarser resolutions. An artificial mathematical overestimation of erosion and a realistic natural modelling effect of underestimating resedimentation cause this. Increasing the spatial extent eliminates the artificial effect while at the same time the realistic effect is enhanced. Both effects can be quantified and are expected to increase within natural landscapes. The modelling of landscape processes will benefit from integrating these types of results at different resolutions

Modelling Land-Use and Land-Cover Change

The decade since the initiation of the Land-Use/Cover Change (LUCC) project in 1995 (see Chap. 1) has witnessed considerable advances in the field of modeling of land-use/cover change. The science plan of the project indicated that the major task would be the development of a new generation of land-use/cover change models capable of simulating the major socioeconomic and biophysical driving forces of land-use and land-cover change. In addition, these models were supposed to be able to handle interactions at several spatial and temporal scales. Recent publications indicate that the LUCC science community has successfully met this challenge and a wide range of advanced models, aiming at different scales and research questions, is now available (Briassoulis 2000; Agarwal et al. 2001; Veldkamp and Lambin 2001; Parker et al. 2003; Nagendra et al. 2004; Veldkamp and Verburg 2004; Verburg et al. 2004b; Verburg and Veldkamp 2005). One of the most important observations that can be made examining the range of available land-use/ cover change models is the wide variety of approaches and concepts underlying the models. This chapter intends to describe the variety of modeling approaches, discuss the strengths and weaknesses of current approaches and indicate the remaining challenges for the land-use science community. Not being able to discuss all individual models and approaches, we will focus on broad distinctions between approaches and discuss how modelers have dealt with a number of important aspects of the functioning of the landuse system. A land-use system is understood here as a type of land use with interrelated determining factors with strong functional relations with each other (see Fig. 1.2). These factors include a wide range of land-use influencing factors than can be biophysical, economic, social, cultural, political, or institutional. The discussion of modeling approaches in this chapter is illustrated with examples of models and results from selected research projects.

Projecting land use transitions at forest fringes in the Philippines at two spatial scales

This paper presents two applications of a spatially explicit model of land use change at two spatial scales: a nation-wide application for the Philippines at relatively coarse resolution and an application with high spatial resolution for one island of the Philippines: Sibuyan island, Romblon province. The model is based on integrated analysis of socio-economic and biophysical factors that determine the allocation of land use change in combination with the simulation of the temporal dynamics (path-dependence and reversibility of changes), spatial policies and land requirements. Different scenarios of near-future developments in land use pattern are simulated illustrating the effects of implementing spatial policies. Results from the coarse scale model with national extent mainly serve to identify the overall pattern of land use change and ‘hot zones’ of deforestation. The detailed application provides more insight in the pattern of land use change and its consequences for ecological processes. The use of the results for environmental assessments is illustrated by calculating spatial indices to assess the impact of land use change on forest fragmentation. It is concluded that spatially explicit modeling of land use change yields important information for environmental management and land use planning. The applications illustrate that the scale of analysis is an important determinant of the model configuration, the interpretation of the results and the potential use by stakeholders. There is no single, optimal, scale for land use change assessments. Each scale enables different types of analysis and assessment: applications at multiple scales therefore give complementary information needed for environmental management.

Evaluating impact of spatial scales on land use pattern analysis in Central America

The complexity of the relations between land use patterns and their spatial determinants causes the scale of analysis to influence the results. Often, focus is on one aspect of this scale effect, the spatial resolution. This study emphasises the influence of a varying spatial extent on the analysis of land use patterns in six countries in Central America. Statistical techniques are used to determine the relationship between six land uses and a number of potential determining factors, varying both resolution and extent. Results indicate that the effect of spatial resolution, by aggregating a basic grid to larger units, is small in comparison with other similar studies. The effect of a varying extent, by keeping either national boundaries or analysing the entire region at once, on the other hand, is substantial. An unrealistic redistribution of all major land use types, including a large-scale reforestation, is predicted using statistical analysis with the entire region as extent. When expanding the extent to a unit larger than a country, implicit assumptions concerning market mechanisms and national policies are adopted that do not correspond to the actual situation. Despite the existence of the Central American Common Market, it cannot be assumed that any agricultural land use will expand to satisfy an increasing demand in another country. Findings strongly suggest that any modelling effort at regional or global level should incorporate a thorough analysis of the effects of spatial scale on land use change predictions.

Multi-scale modelling of land use change dynamics in Ecuador

Abstract A spatially explicit multi-scale approach to land use change modelling is explained and demonstrated. It is based on the empirical description of the biogeophysical and socio-economic drivers of land use at different aggregation levels, using a system analytical approach for the characterisation of agro-ecosystems. Sub-national changes in land use following changes in the national demand for agricultural commodities are modelled on the basis of complex interactions in time and space and the competition between alternative land uses. In a multi-scale allocation procedure, land use changes are calculated for cells of a geographical grid with a maximum resolution of 5×5 min. The model was applied to Ecuador. Land use change allocation in the model was validated with historical data. A hypothetical future base-line scenario of increasing demands for agricultural commodities was used to demonstrate how dynamics of land use are modelled. The results may be used to detect so-called ‘hot-spots’. These are dynamic areas where impacts of land use change on the natural resource base can be expected. The model offers scope for comparison of different scenarios in which alternative development paths can be defined, for example, with respect to changes in food demands, technology levels, infrastructure, soil suitability or the protection of natural areas.