- Shopping Bag ( 0 items )
From the Publisher
"This book has many strengths." (APA Journal, Autumn 2002)
Monitoring Land Supply with Geographic Information Systems Theory, Practice, and Parcel-Based Approaches
Monitoring the supply of buildable land and its capacity to accommodate growth within urbanizing regions is an increasingly important component of urban planning and growth management. Recent developments in Geographic Information Systems (GIS) have opened up new opportunities for local and regional government to monitor land supply and capacity. Based on a study sponsored by the Lincoln Institute of Land Policy, this book reviews the state of the art in land monitoring, particularly as it benefits from the introduction of GIS data and analysis capabilities at the level of individual land parcels.
Monitoring Land Supply with Geographic Information Systems addresses:
* Technical and methodological frameworks for data collection and analysis as well as applications to a range of policy concerns
* Case studies of successful land monitoring programs, including Portland, Oregon; Montgomery County, Maryland; and the Puget Sound Regional Council in Washington
* Thematic topics ranging from database design to urban simulation modeling to organizational contexts
* Detailed findings of a national survey of land supply monitoring programs
This guide presents a comprehensive, timely, and critical overview of a fast-emerging field of planning and policy analysis. It provides an invaluable resource to professionals, including land use and economic development planners, GIS analysts, local government officials, and private developers.
"This book has many strengths." (APA Journal, Autumn 2002)
Introduction Anne Vernez Moudon and Michael Hubner
Part I Overview
1. Current Land Monitoring Practices and Use of GIS: Challenges and Opportunities Anne Vernez Moudon and Michael Hubner
2. Elements of a General Framework for Land Supply and Capacity Monitoring Michael Hubner and Anne Vernez Moudon
Part II Case Studies
3. Portland, Oregon: An Inventory Approach and its Implications for Database Design Lewis D. Hopkins and Gerrit J. Knaap Commentaries: Scott A. Bollens, George Rolfe
4. Montgomery County, Maryland: A Pioneer in Land Supply Monitoring David R. Godschalk Commentary: Lewis D. Hopkins
5. Central Puget Sound Region, Washington: Study of Industrial Land Supply and Demand Lori Peckol and Miles Erickson Commentaries: Scott A. Bollens, William Beyers
Part III Thematic Issues
6. Method and Technical Practice in Land Supply and Capacity Monitoring Ric Vrana Commentaries: Frank Westerlund, Marina Alberti
7. Data Sharing and Organizational Issues Anne Vernez Moudon and Michael Hubner Commentary: Zorica Nedovic-Budic
8. Simulating Land Capacity at the Parcel Level Paul Waddell Commentaries: Nancy Tosta, Kenneth J. Dueker
Conclusions Anne Vernez Moudon and Michael Hubner
Appendix A: Survey of Land Supply Monitoring Practice
Appendix B: Selected Case Summaries
Appendix C: Interview Contacts
Appendix D: May 1998 Seminar Participants
Glossary of Terms and Acronyms
Anne Vernez Moudon and Michael Hubner
The monitoring of urban land has taken place in a variety of contexts in this country. As a general activity, it seeks to record changes in land use, to track land development, and to analyze the patterns and conditions that emerge over time. Specifically, land supply and capacity monitoring (LSCM) focuses on the supply of buildable land and on the capacity of that land to accommodate future development. As such, it also serves to assess future potential uses of land, especially in relation to how zoning and other regulations support or constrain urban expansion and concentration.
The land monitoring process, in its most developed form, is a dynamic and comprehensive activity carried out on a continuous basis. Holmberg (1994, 8, with reference to Masser ) characterizes monitoring as '' crucial for any successful handling of the territorial concern . . . [exhibiting] both similarities and differences compared with traditional census, surveying, and mapping, [and comprising] continuous and real-time supervision of a set of variables within a geographic region. '' Furthermore, monitoring extends beyond passive description to include active intervention in response to information generated. These observations describe well a fully implemented system for LSCM.
Urban and regional planners bear the chief responsibility for land monitoring. Although not a widespread activity, LSCM has emerged as a distinct area of planning in metropolitan regions where the demand for land is high, thus supporting the political will to restrict the effective supply of land available for urbanization through policies and regulations that limit development, particularly at the urban fringe. As such, LSCM stems primarily from efforts to manage urban growth.
Concerns about the supply of urban land have become increasingly urgent over the past decade. During this period urban growth has been characterized by rates of land consumption substantially higher than increases in population. Interest in the supply of urban land springs from two separate yet related, considerations. One is the conservation of open space--agricultural and natural areas--within metropolitan regions, particularly at the urban fringe. The other relates to the high costs of providing infrastructure (especially transportation infrastructure) required to serve sprawling metropolitan development. As these issues have come to the center stage of local political agendas, many jurisdictions have passed laws to manage growth and to monitor closely changes in urban land use and available supply (De Grove 1992; Stein 1993). Generally, growth management mandates aim for orderly and efficient growth patterns, not only to preserve land outside the urbanized area and to reduce infrastructure costs, but also to mitigate the negative impacts of growth on traffic flow and air and water quality and to distribute the differential changes in local tax bases that result from uneven development (Diamond and Noonan 1996). As an accompaniment to growth containment policies, LSCM serves to assess the adequacy of the land supply (limited as it is by regulation) to accommodate future population and employment increases, while achieving the objectives of these policies.
Although LSCM has become an important component of urban planning within the context of growth management, it has yet to emerge as a bona fide specialization within planning study. Research on the subject is limited, and no comprehensive theory has been articulated to guide its practice. Few academic researchers in planning claim land supply monitoring to be their area of scholarly activity. Godschalk's and Bollens' work in the mid-1980s (Godschalk et al. 1986; Bollens and Godschalk 1987) still stands alone in its comprehensive coverage of issues and methods related to LSCM. As Bollens noted in summarizing the field for the American Planning Association (Bollens 1998), various researchers have addressed separate issues that relate to specific aspects of land supply and capacity--such as urban sprawl, housing and land prices, costs of infrastructure, regulatory impacts, and program evaluation. Little work has focused directly on the technical, methodological, organizational, or substantive aspects of LSCM as a distinct planning activity. This gap may be partly explained by a lack of '' critical mass'' of well-developed LSCM programs. Another explanation may lie in the intrinsic complexity of LSCM, which must address an increasingly wide range of interacting variables, including environmental conditions, capital facilities and infrastructure planning, regional economies and land markets, and real estate and development practices.
At this point, therefore, land supply monitoring in this country is defined more by a set of practices than by a cohesive theory or body of empirical evidence. A handful of jurisdictions pioneered LSCM in the 1970s and 1980s, among them the Metropolitan Council (Minneapolis-St. Paul), Minnesota; Montgomery County, Maryland; Portland Metro, Oregon; Lane Council of Governments (Eugene-Springfield), Oregon; and the San Diego Association of Governments, California. (The work of these jurisdictions is updated in Appendix B.) These and other pioneers have essentially determined the state of the practice. Although jurisdictions that conduct LSCM attempt to keep abreast of what others are doing relative to land monitoring, there are no formal ties between them--for example, there is no national group or organization that keeps tabs on ongoing developments in the field. As a result, the various approaches to LSCM still await critical review and systematic evaluation and comparison.
This chapter summarizes the characteristics of current practice in land supply and capacity monitoring in the United States. The summary relies on research of more than four dozen jurisdictions across the country, a subset of which are analyzed in detail in Appendix A. In this chapter the general scope of practice is reviewed first, followed by a discussion of the recent advances in GIS and their implications for LSCM. (The actual methodological dimensions of LSCM are addressed in Chapter 2.) The chapter concludes with a summary of the challenges and opportunities facing the field now and through the coming decade.
The term land supply monitoring suggests an activity whereby the supply of land is tracked as it is occupied by urban activities, and as changes take place in the use of urbanized and urbanizing land over time. It also suggests a systematic assessment of potential uses of the land supply for the purposes of informing land policy. In the reality of practice, LSCMis rarely implemented in a comprehensive way. Few, if any, jurisdictions have managed to establish monitoring systems as ongoing inventories of the entire land supply. Nor have they used LSCM as a tool to measure on a regular basis the effectiveness of established policies and regulations and the possible impacts of new regulations, policies, or land development practices.
A survey of jurisdictions across the country shows that few have access to databases that are comprehensive or detailed enough to address the entire range of urban land uses. In addition, few systematically monitor trends by capturing longitudinal data on land use and development. Hence, most monitoring approaches involve partial inventories carried out as '' snapshots'' of the land supply, as opposed to continuous processes of data collection and analysis. Furthermore, these snapshots take place periodically and not always at regular intervals.
The limited adoption of LSCM among jurisdictions nationally has many explanations, of which several stand out. First, LSCM is often prohibitively expensive to carry out, especially in terms of database development and maintenance. Second, most jurisdictions have not implemented growth management policies, which has resulted in uneven institutional support and requirements for LSCM. And third, in jurisdictions where growth management is in place, the contentiousness of local politics may create an unstable political climate for sustained LSCM system development.
Characteristics of Practice
Our survey of jurisdictions (see Appendixes A and B) highlights six general characteristics of current LSCM practice. First, it is led by the public sector and typically carried out by local governments--municipal, county, and metropolitan jurisdictions. At the same time, land monitoring activities are usually closely watched by the private sector. Second, it is regional in scope, reflecting the metropolitan-wide reach of land markets, major infrastructure systems, and growth control policies. Third, it addresses primarily medium-to long-term supply and capacity prospects, largely to relate appropriately to comprehensive plans and to regional growth forecasts, which usually extend over periods of 10 to 20 years. Being closely linked to long-range planning, LSCM is most successfully performed as a centralized or coordinated activity led by a metropolitan planning organization (MPO) or similar metropolitan-wide body. In this context, LSCM has been successfully employed in evaluating the implementation of local and regional plans. It has contributed to keeping issues of urban growth, land consumption, and infrastructure development current, and to raising peripherally the public awareness of the environmental and fiscal costs of sprawl.
Fourth, LSCM involves the collection, maintenance, and analysis of large, complex data sets, with elements ranging from aerial imagery to infrastructure networks to parcel-specific land records. As a multifaceted and highly data-dependent activity, LSCM generally requires significant coordination to complement extensive investments in specialized labor and technology by local government. '' Corporate'' data centers are common, functioning as the equivalent of line departments responsible for generating, maintaining, and process-ing data for other departments within a contributing jurisdiction or jurisdictions. Data are captured and stored for various units of land, corresponding to data type and administrative function (e. g., parcels for assessments, districts for utility servicing). Analyses of the data range widely in the level of aggregation with which they address supply and capacity. Although parcel-level analyses are increasingly common, many efforts aggregate data to a zonal level--such as census tracts, Transportation Analysis Zones (TAZs), Forecast Analysis Zones (FAZs), or zoning districts--in order to evaluate the potential for land to be developed. They yield results that are fairly aggregated and, as discussed later, largely inadequate to meet the current needs of growth management policy and urban land and systems management.
Fifth, most land monitoring programs have been sporadic rather than continuous efforts, whereby land supply and capacity are analyzed for an established time horizon but often not tracked over time to support a good understanding of trends. Although a comprehensive, one-time inventory can provide a baseline against which to measure trends and subsequent changes, capturing longitudinal data over any significantly long period has proven difficult. Moreover, changes in technology, inconsistent financial support for data collection and maintenance, and shifting mandates for long-range planning have all contributed to the lack of continuity in monitoring efforts nationally.
Sixth and last, our research reveals the following primary substantive orientations within LSCM practice:
LSCM has, as may be expected, been subject to the contentious climate of growth management politics. In debates over '' planning for'' versus '' planning with'' growth, critics of the former approach generally assert that regulations limiting development or containing urban growth excessively restrict both the amount and location of buildable lands. Home builders and other development interests also question the timing of land availability, particularly short-term supplies of buildable single-family lots and large sites for multifamily development and nonresidential uses.
Land monitoring programs focused on long-range supplies of land over entire urban regions tend to overlook factors that affect short-term land availability--even though they may be the target of private-sector criticism. Over the short term, regional inventories typically fail to capture locational variations in market demand and may have difficulties addressing institutional factors such as permit approval rates and staged extensions of infrastructure. Moreover, their treatment of physical and environmental constraints often is too general or too spatially coarse to address specific barriers to development. Finally, regional inventories typically lack detailed consideration of various economic variables, such as the demand for large parcels (particularly for industrial uses), market availability, land prices, and the location preferences of households and firms. Negotiated within the political arena of growth management, these criticisms have had an impact on LSCM, spurring efforts to assess land supply in increasing detail and incorporating an increasingly complex array of constraints and market considerations.
Recent advances in land information technology, and specifically the implementation of parcel-based geographic information systems (PBGIS) by local governments, hold some promise that future land monitoring will be able to respond more directly and comprehensively to many of these concerns in addition to meeting the land policy analysis needs of public decision makers. The following sections review the state of the art in geographic information systems (GIS) and discuss the implications of the new technology for LSCM.
Land record keeping has been the subject of elaborate bookkeeping systems since the beginning of urbanization. Historically, advances in writing and printing technology were quickly applied to improve the state of land records. Computers have naturally followed suit as a welcome aid to maintaining land information. Even in computerized form, however, land information was, until recently, stored and retrieved in two largely separate formats: land records in tabular form and maps depicting the land parcel (or other areas of land) showing its location and physical extent. The latest trends in using GIS for managing land information, and especially parcel-based GIS are having a significant impact on land monitoring. Following are a review of recent developments in land information technology used by local government and a discussion of the potential of PBGIS to further enhance LSCM.
GIS in Local Government: A Decade of Expansion
The use of computerized land information systems (LIS) for land supply monitoring from the mid-1970s through the mid-1980s was reviewed in Godschalk et al. (1986). Drawing on their survey of the practice, the authors described a prototype automated land supply information system (ALSIS), consisting of a set of computerized databases for tracking both parcel-based land records (a '' parcel file'') and in-process development and development applications (a '' project file''). Using case studies, the authors illustrated the implementation of various specific components of such a system in a range of local and regional government settings.
At the time of this study, geographic information systems (GIS) technology was in its infancy as a tool for local land planning and management; only 6 of 24 jurisdictions surveyed by Godschalk et al. (1986) had developed computer-mapping capabilities. Since these observations were made, more than ten years ago, GIS has advanced considerably and its use by local and other governments has mushroomed, extending from daily routine tasks to sophisticated special-purpose analyses. As part of this trend, planning departments have become primary users of GIS. The rate of GIS usage is especially high among jurisdictions that practice some form of land supply monitoring. Of the more than three dozen jurisdictions surveyed for this work that actively practiced LSCM, nearly all utilized GIS as a tool for capturing, managing, or analyzing land data, and more than half did so with a parcel layer as a major component (see Appendix A).
Various technological changes have fostered the adoption of GIS by local governments, enhanced its application for multiple purposes, and increased the affordability of system implementation. First, distributed computing environments, incorporating local area networks and network file systems, have increased flexibility and reduced redundancy related to both data storage and software. Second, the introduction of personal computers (PCs) and personal workstations powerful enough to process large databases, have enabled the storage and manipulation of highly detailed and extensive spatial data layers, such as parcels. Third, relatively low-cost, user-friendly, and increasingly powerful desktop software has allowed a broad range of local government staff to utilize GIS on their own computers (Public Technology Inc. et al. 1991; Korte 1997; Orman 1997).
A wide range of benefits accrue from GIS adoption and make it an attractive investment for local governments. These benefits include (1) increased effectiveness of urban services delivery (e. g., utilities management, emergency services), (2) cost savings through the automation or streamlining of routine tasks (especially custom mapping and on-demand data retrieval), (3) increased employee productivity, (4) reduction of redundancy between and within departments (especially related to the creation and updating of spatially referenced data), (5) revenues gained from selling digital GIS and map products, (6) improvements in the speed and accuracy of responses to queries, and (7) multiple benefits stemming from improved decision making at all levels (Public Technology Inc. et al. 1991; Korte 1997; Klosterman 1997).
A recent national survey of cities (population greater than 25,000) and counties (population greater than 50,000) bears out the dimensions of the spread of GIS to local jurisdictions and to a wide range of departments and functions within them (Kollin et al. 1998). The survey results show that in 1996, 77 percent of respondents (85 percent for cities of more than 100,000 population) reported using GIS. Among GIS users, the average number of departments using GIS was 2.34 (2.85 for large cities). The most common uses of GIS included comprehensive planning, zoning and subdivision review, transportation planning, and utilities and storm water management. The types of GIS data most often utilized included road networks, political and administrative boundaries, hydrologic features, land use and zoning maps, and cadastral/ land records (all more than 80 percent usage in 1997). Interestingly, surveyed jurisdictions with or without GIS most often identified the impacts of '' growth and development'' as a top '' natural resources'' concern. However, among these respondents, only 42 percent indicated that they actually used GIS to address growth and development problems. This last finding suggests that GIS, in its current stage of implementation within local government, presents a largely untapped resource to address problems of urban growth, in part through the monitoring of land supply and capacity.
As GIS technology has spread, so has the creation of and access to spatial and spatially referenced data. One of the most significant developments for land monitoring has been the addition of parcel layers by many county and municipal GIS programs. In a separate study of 29 local jurisdictions with operational GIS in the southeastern part of the United States, Nedovic-Budic (1993) found that parcels constituted the most frequently mapped feature. Often spearheaded by an assessment or engineering department, the construction of a digital parcel base map can incorporate older computerized land information systems (LIS) data (which are based primarily on the parcel record) and provide new tools for the display and analysis of land-related data in a spatially explicit manner. Such land records '' modernization'' is now in progress widely throughout the nation, and parcels are quickly becoming standard elements in the GIS of most sizable local jurisdictions (Huxhold 1997; Kollin et al. 1998). The following sections address specific implications of this trend for the application of GIS to the monitoring and analysis of land use, supply, and capacity.
Parcel-based GIS: Representation and Implementation
To understand the potential of parcel-based GIS for land monitoring, it is useful to consider the several different ways that parcels may be represented spatially in a GIS--as points, polygons, or grid cells--and the appropriate application of each approach. The simplest representation is a parcel-point layer, composed of geocoded parcel locations that correspond to either parcel centroids (xy coordinates) or to '' address ranges'' along a street network. The parcel-point approach is widely used in the private sector for demographic and market analysis, for site searches, and for the marketing of development and real estate products (such as with computerized Multiple Listing Service data) (Castle 1993, 1997; Thrall 1997). Within the public sector, parcel-point GIS has been and continues to be used for a variety of planning analyses and land monitoring applications, especially by jurisdictions in which a fully digitized parcel coverage is not yet available. However, parcel points present significant limitations for accurate land monitoring, partly because of the lack of precision in geocoding, especially in address matching (Drummond 1997), and because of the absence of a spatially extensive representation of the boundaries and area of the parcel. Parcel points thus limit the ability to perform accurate overlays and to analyze land supply within a detailed geographic context (i. e., the parcel itself and its surroundings).
For the multiple purposes of land information management in local government, parcel-point coverages are often utilized as a stopgap solution before the completion of a true parcel-based GIS (see Dueker and DeLacey (1990)on the County Geographic Index, and the case of Snohomish County in Appendix B). A true parcel-based GIS incorporates a fully digitized parcel layer composed of parcel boundaries, allowing parcels themselves to be represented as discrete areas (polygons). As such, the parcel layer offers many advantages over parcel points, including enhanced display and map production and the ability to consider precise site-level details necessary or desirable for a variety of analytical operations. The conversion of parcel maps to GIS format is generally accomplished either through digitization of lines on paper maps, or as constructed directly from deeds and survey information using coordinate geometry and aided by alignment with planimentric or orthophotography layers (Donahue 1994). The parcel layer may also be derived partly from previously developed digital computer-aided design (CAD) drawings (commonly used by public works and engineering departments since the early 1980s).
Whether employing a GIS representation of parcels as points or as polygons, the creation of a parcel layer entails the assignment of a unique parcel identifier, usually referred to as a parcel identification number (PIN), used to link the spatial features to individual land records and other parcel-specific data (Huxhold 1991). As a secondary, but potentially useful, parcel identifier, a standardized Master Address File can link parcels to site-level data produced by the numerous local government departments that identify parcels by address rather than PIN in their record-keeping systems (Donahue 1994).
Point or polygon representations of parcels are features of vector GIS. Parcel polygons may be converted to grid cell format or raster GIS. Recent advances have increased the ease of data conversion from vector to raster and vice versa. Although some detail may be lost through this process, grid cells of 20 to 50 feet are usually adequate to capture the underlying parcels, with at least one cell corresponding to each parcel polygon. The advantages of working with parcels in grid cell format include: (1) reducing the computational burden of large-area parcel-level calculations and transformations, especially those relying on multiple geographic layers, (2) improving the ability to consider weighted variables in screening for site suitability (Dueker and DeLacey 1990), and (3) allowing analysis of the potential use of parcels by considering their small-scale geographic context--such as with '' neighborhood'' calculations that assign values to cells based on the values of nearby cells. (See Portland Metro and Lane Council of Governments case summaries in Appendix B for examples of raster GIS analysis.)
Within urban and regional planning, most land monitoring applications of GIS without a parcel coverage (or parcel layer) have been, and continue to be, based on zonal geography. 1 Most frequently, zones correspond to administrative or statistical areas--census tracts, block groups, and blocks; planning areas; and Transportation Analysis Zones (TAZs) and Forecast Analysis Zones (FAZs)--allowing planners to relate land areas to sociodemographic and economic statistics that are collected at or aggregated to specific zonal units. These practices derive from the early use of census data (when it was available only in tabular form) and the tradition of social area analysis. Zones may also relate explicitly to land use and development, including regulatory units (e. g., zoning districts, planned land use areas) and areas representing contiguous land uses (typically classified from aerial photography). Although parcel-level land records can be aggregated to any of the various zone types for subarea analysis of supply and capacity, the relatively coarse spatial resolution of zones, as well as their abstract delineation, imposes limitations on the utility of analysis results generated with zonal GIS.
The coarseness of zonal geography (TAZs and FAZs that encompass several thousand acres are common in suburban areas) limits the degree to which the patterning and clustering of development and development potential can be detected. PBGIS, representing parcels that range from fractions of acres to multiple acres, even in suburban areas, offer a more precise basis, in terms of scale and level of disaggregation, than zonal GIS for identifying and analyzing spatial patterns. In addition, because the parcel corresponds directly to many of the administrative and private-market development decisions that accompany incremental land use change, it has been recognized as the preferred unit for land monitoring (Bollens and Godschalk 1987; Enger 1992; Vrana and Dueker 1996; Bollens 1998).
Typically, however, additional land supply data are captured and represented in units other than parcels. Over the past decade, theory and practice of GIS for land management and planning have expanded the concept of a multipurpose cadastre (stressing the representation of nearly all land information as attributes of parcel units) to embrace the concept of a multipurpose land information system (MPLIS) (Ventura 1991). An MPLIS incorporates various '' spatially registered layers of institutionally independent data, '' including parcels as a primary spatial layer (Dueker and DeLacey 1990, 488- 489).