Journal of Land Use Science Vol. 3, No. 4, December 2008, 199–201
EDITORIAL Land use or land cover? There is nothing constant in this world but inconsistency Jonathan Swift
Today most land cover data include elements of land use and vice versa. Historically, mappings of land were concerned with land use and manually recorded socio-economic activities and land management practices. With the increased availability of remotely sensed imagery since the 1970s and the ability to process such data using computers, the principal concern of such mappings has been to record the phenomenon of land cover and land use. As Veldkamp and Verburg (2004) note, many land use and/or land cover modelling approaches have often treated land use and land cover as if they were interchangeable concepts. However, the conflation of the two concepts in most geographical information derived from remotely sensed data is problematic for the research community who require either land cover for environmental models or land use for policy making. This special issue of the Journal of Land Use Science contains four papers describing issues related to land use and land cover. Each of them tackles a different aspect of the land cover/land use paradigm. Comber (this volume) presents a philosophical rationale for the separation of land use and land cover. He argues that their frequent confusion is problematic for many data integration and modelling activities which require either ‘land use’ or ‘land cover’ as their inputs. For example, the common approach to overcome this confusion is to translate land cover classes to land use classes. However, Comber notes that the direct translation of cover to use is not straightforward due to their different temporal attributes: land cover tends to be static over short periods of time, whereas there may be multiple land uses at any given place. The paper proposes an approach for the separation of land cover and land use using ‘data primitives’ – dimensions that describe at the most fundamental level the land use and land cover processes. The method presented by Comber identifies different aspects of land use and land cover, for example the economic value and social value in urban contexts and food production in rural areas. Each class is scored in each dimension, and by combining the scores the degree of land cover or land use is (subjectively) determined. The method allows the concepts of land cover and land use to be separated relative to the task in hand, and the outputs show the degree of land use, the degree of land cover and the locations where the concepts of use and cover are confused. Bakker and Veldkamp (this volume) propose the land use–land cover ratio as a method for characterising different landscapes. They also question the assumption of a one-to-one ratio between land use and land cover found in many modelling activities as multiple uses can be related to one cover type. Much of the mismatch between land use and land cover can be attributed to differences between the total land area dedicated to the production or provision of a land-based commodity (observed land cover) and the actually used or harvested area (reported as land use). Bakker and Veldkamp introduce the concepts of ‘primary land uses’ as the production or provision ISSN 1747-423X print/ISSN 1747-4248 online # 2008 Taylor & Francis DOI: 10.1080/17474230802465140 http://www.informaworld.com
200
Editorial
of a certain quantity of a certain commodity and ‘secondary land uses’ which do not directly affect or control the land cover as does primary land use and can co-exist with primary land uses and with each other. This multifunctionality of land is seen when at least one secondary land use is combined with a primary land use. The land use–land cover ratio proposed by Bakker and Veldkamp incorporates a secondary land use function as relevant adjunct to characterising land cover and quantifies the multifunctional nature of much land cover, allowing trade-offs between primary and secondary use to be modelled. This approach accounts for multiple land uses when converting land use to land cover by incorporating the land use requirements using a land use–land cover ratio. Hein et al. (this volume) analyse the economic impacts of land use change for woodlands in Zambia. They present a framework for assessing the impacts of land use change using an environmental economics approach and an overview of three methods for analyzing the impacts of land use change on ecosystem services and functions. The framework analyses the economic costs of land degradation and the benefits of sustainable land management factors that are frequently ignored or under-estimated in decision making. The approach comprises three complementary types of assessment: partial valuation, total valuation and impact assessment. The relevance of the work relates to the range and value of goods and services provided by agricultural and natural ecosystems, many of which directly or indirectly contribute to human welfare and, as such, have economic value. However, their value is currently unrecognised in many planning and policy arenas because the benefits are often difficult to specify as they vary in nature, space and time, they are often invisible and not traded in markets and they are perceived differently by the stakeholders that pay the costs and the beneficiaries of the particular service (e.g. downstream water users benefiting from the regulation of water flows). Hein et al. presents a framework specifically targeted at analyzing the economic impacts of land use change that addresses these issues. In the framework, specific attention is paid to how ecosystem services and ecosystem dynamics can be linked to economic values. The framework supports three main approaches (partial valuation, total valuation and impact assessment), and each of them is illustrated with a case study. Wadsworth et al. (this volume) present an approach for integrating thematically inconsistent data sets using quantified conceptual overlaps. The problem the authors address is endemic in geographical information and natural resource assessment. The real world is infinitely complex and any data collection such as a field mapping survey involves processes such as abstraction, classification, aggregation and simplification. The end result is that natural resource inventories of the same phenomenon at different times or by different people often vary for a variety of reasons that have nothing to do with changes or differences in the feature being mapped, but reflect advances in technology (e.g. satellite remote sensing) or science (e.g. plate tectonics). The result of changing technological, scientific and policy objectives is that repeated natural resource inventories rarely employ the same methods as in previous surveys (Comber, Fisher and Wadsworth 2003) and scientists often have to use inconsistent data in their models. Wadsworth et al. assess land cover and land use change in central Siberia using several different digital data sets derived from satellite images, acquired in different years which differ in the number and types of classes, spatial resolution, acquisition date and sensor. The characteristics of the different data sets make it difficult to interpret change maps as either land cover/land use change or data inconsistency. The method presented by Wadsworth et al. combines quantified conceptual overlaps (Ahlqvist 2004) with semantic statistical approaches (Comber, Fisher and Wadsworth 2004a,b) and characterises each of the classes in each of the data sets using five physical ‘domains’ related to land cover. This approach is similar to the data primitives in Comber, but scores the upper and lower bounds of each class in each domain. The papers presented in this special issue examine issues related to data inconsistency:
Journal of Land Use Science
201
Internal data inconsistency, where the concepts of land use and land cover are rolled together. This is now so prevalent that graduate students see this as being the norm and the land use/land cover epithet predominates much of the literature despite the logical incompatibility of the two concepts (Bakker and Veldkamp, this volume; Comber, this volume). External data inconsistency, where the different data sets record the same features in different ways due to changes in technology and scientific understanding (Wadsworth et al, this volume). Resource assessment inconsistency, where the true value of land activities are not accommodated by current methods for understanding the impacts of land use change. This is due to the lack of linkages between actual ecosystem services and actual ecosystem dynamics to measured economic values (Hein et al, this volume). Different data sets will always differ from each other as they are created in response to different local and national needs. However, any given land cover data set or any given land use data set contains elements of the other despite the fact that they are fundamentally different processes with different temporal characteristics, with different relations to the processes of entity classification (where each object is allocated to one class) and with different relations to physical attributes and to socio-economic activities. The conflation of land cover and land use has been a convenient product of the availability of the remotely sensed data and the ability to automatically process it: it is not grounded in good scientific practice as documented by Fisher, Comber and Wadsworth (2005). This editorial is a plea for data producers to shift from data-driven land cover/land use applications to ones that are driven by the task in hand. A.J. Comber Department of Geography, Leicester University, Leicester, UK
[email protected]
References Ahlqvist, O. (2004), ‘‘A Parameterized Representation of Uncertain Conceptual Spaces,’’ Transactions in GIS, 8, 493–514. Comber, A.J., Fisher, P.F., and Wadsworth, R.A. (2003), ‘‘Actor Network Theory: A Suitable Framework to Understand How Land Cover Mapping Projects Develop?’’ Land Use Policy, 20, 299–309. Comber, A.J., Fisher, P.F., and Wadsworth, R.A. (2004a), ‘‘Assessment of a Semantic Statistical Approach to Detecting Land Cover Change Using Inconsistent Data Sets,’’ Photogrammetric Engineering and Remote Sensing, 70, 931–938. Comber, A.J., Fisher P.F., and Wadsworth R.A. (2004b), ‘‘A Semantic Statistical Approach to Negotiation Heterogeneous Ontologies,’’ International Journal of Geographic Information Science, 18, 691–708. Fisher, P.F., Comber, A.J., and Wadsworth, R.A. (2005), ‘‘Land Use and Land Cover: Contradiction or Complement,’’ in Re-Presenting GIS, eds. P. Fisher and D. Unwin, Chichester: Wiley, pp. 85–98. Veldkamp, A., and Verburg, P.H. (2004), ‘‘Modelling Land Use Change and Environmental Impact,’’ Journal of Environmental Management, 72, 1–4.
The author has requested enhancement of the downloaded file. All in-text references underlined in blue are linked to publicatio
