To implement ecosystem management effectively, we must be able to collect, access, and manipulate data in a shared environment across diverse functions, ownerships, and scales.
The purpose of this paper is to provide a synthesis of management experience related to data management, collection, and inventory. The following areas are addressed:
Key Questions: Several fundamental questions and considerations provide the starting point for understanding data management, collection, and inventory needs related to a desired product or project. The organization of these questions reflects an ordered way of thinking about objectives, partners, products, and users that should precede initiation of any data gathering activity.
Case Studies: Case studies outline a range of approaches to data management, collection, and inventory that demonstrate various technologies and methods available. The case studies are presented to assist managers in understanding and selecting approaches suitable to a particular situation, based on an evaluation of successful and unsuccessful examples, barriers, and promising possibilities. Case studies emphasize considerations relative to the sharing of data, from collection and inventory to management and ultimate use, at various scales and with a variety of potential partners.
The case studies also provide an introduction to new technologies that have been used in a production (as opposed to pilot) environment. They are intended to provide enough information to allow evaluation of a technology at a particular scale for a particular need through lessons provided by the individual case studies.
Overview of Approaches and Technologies: This section presents background on individual components of data management, collection, and inventory. This background can help the manager interpret and synthesize the case studies. Information planning provides a discussion of processes and structured methodologies for determining information requirements and managing information over its life cycle. Data sharing describes processes, technologies, standards, and other factors common to successful approaches for sharing information. Data Manipulation with Geographic Information Systems provides an introduction to the technology of manipulating electronic spatial information. Remote sensing discusses the application of satellite and related technologies to data acquisition and inventory. Data collection and inventory describes a range of methods, techniques, and standards for collection and inventory of ground data.
Glossary: Selected list of terms used in this document.
A Complex Environment
Information is now the single most costly investment which agencies make. Decisions in this arena are not easy to make (figure 1). As ecological stewardship work processes change, information needs change. As new information technologies take hold (satellite imagery, GIS, the Internet), they change how resource data is collected, stored, shared, and used. As more and more partners work toward common objectives, it becomes increasingly important to establish appropriate standards for data and protocols. And as the public gets improved access to the "information superhighway", communities of interest expect to be informed and involved participants in land and resource decision-making activities. Add to these forces for change the imperative of organizational downsizing and reduced budgets. The result is both a critical need and real opportunity to improve return on investment through effective data management, collection, and inventory.
Information has five salient factors which must be managed. These factors are: collection, quality, integration, storage, and access. The following "Key Questions" have been assembled to help the manager address these factors. These questions are simple and direct so a manger does not need to be a "rocket scientist" to get to the salient issues and the bottom line answers. These questions can be applied to any ecosystem and/or geographic information system (GIS) type activity.
Questions relating to the nature of the problem:
Budget Constraints---- ---Competing Projects
\ /
\ /
\ /
\ /
\ /
\ /
Re-engineering --------------The Manager---------------Technology Changes
/ \
/ \
/ \
/ \
/ \
/ \
Changing Missions------ ----Reduction in Staff
& Priorities
Figure 1. Manager's decision environment
Data management, collection, and inventory approaches can be characterized along several dimensions. Two that are frequently useful to consider are scale and extent of functional/organizational integration (or coordination). The case studies that follow attempt to fill in the matrix in figure 2. Scale and coordination are highlighted at the beginning of each case study description.
| Multi-Agency Multi-Function | Multi-Agency Single Function |
Single Agency Multi-Function | Single Agency Single Function | |
| Ecoregion | . | . | . | . |
| Subregion | . | . | . | . |
| Landscape | . | . | . | . |
| Land Unit | . | . | . | . |
Figure 2. Project dimensions.
Note: Mention of trade names of software and hardware in this report, including in the case studies, is intended for information only, and is not an endorsement or recommendation by the authors or of their agencies.
OVERVIEW OF APPROACHES AND TECHNOLOGIES
Different approaches to information planning are appropriate for different ecosystem management endeavors. For efforts like fire recovery projects or watershed analyses, a common approach to information planning is the information needs assessment (INA). INAs most often focus on analysis of all the spatial data needs of a specific project. They have breadth but not depth. A different approach is needed when a group of agencies, an agency, or a unit of an agency embarks on building a specific long term subject matter database that will serve the needs of multiple future projects. Examples of such databases are a vegetation or an ecological classification and inventory database. Flexibility and long term support are key considerations . Planning needs for this type of effort are better met by a structured life cycle planning process.
Information needs assessments typically focus on the identification of needs for resource information which have a spatial (GIS) link. Data and data processing needs are defined at a general, but fairly comprehensive, level. The INA process begins with an understanding of management issues and responsibilities and leads to a determination of what data is essential to address the issues and carry out responsibilities. The results provide the basis for development of GIS themes and associated data tables. They lead to setting priorities, assessing costs and personnel, and setting time frames. The process also opens the lines of communication among individuals who will be involved in the collection, management, and use of information for a given project.
The approach to an INA depends upon the nature and scope of the project which is the subject of the INA, the parties involved, and the operating time frames. This section addresses the major elements that should be part of any INA method and is geared to situations where a low to a medium amount of structure and depth is appropriate.
Although the approaches to an INA may be different, the timing of the INA is the same. Conduct the INA as soon as possible after the project goals and overall methods have been defined and personnel have been selected. Following the INA, the results can be used in developing more detailed plans for acquiring data, designing databases, or developing applications.
An INA is conducted in the context of a specific project or organizational unit that needs information. Ensuring that all INA participants share a clear understanding of the purpose of the project or the responsibilities of the organizational unit is an important starting place. Outcomes expected from the work, and therefore the INA, should be clearly understood. Laws, policies, regulations or directives may contain specific requirements for outputs, methods, or particular information elements to be used or produced, or there may be a great deal of latitude for project participants to determine these. There may be set direction on exactly who will be involved and on the time line and budget for information-related tasks, or the INA may be expected to help set the time line and determine budget and personnel needs. The INA should be planned and conducted to produce the expected outcomes. If the project purpose or organizational responsibilities and the expected outcomes are not clear, get clarification before the INA.
Consider how the following factors will influence the INA:
* Level of Detail. There are a number of aspects to consider that can be grouped under level of detail. Spatial data can be more or less detailed in terms of geographic scale and spatial resolution. Often, although not always, a larger geographic area can be addressed with information captured at a smaller scale and/or of coarser resolution. Attribute data are more or less detailed in terms of how finely they are broken down into categories or classes. Both spatial and attribute data can vary in level of accuracy.
*Constraints. Time, money, and personnel are limited and will be constraints on any project or organization, and in turn on the supporting INA. Some of these constraints may be clearly defined while others may be more vague and implied. A legislative body may enact a law directing an agency to conduct a watershed recovery program within two years. This defines the time constraint. Within the legislation there may be no new positions or money budgeted for the process, placing other constraints on the program. It is important to determine what resources of time, personnel, and money will be allocated to the project and to the INA.Looking at the process and results of similar INAs by other projects and organizations may improve the quality and efficiency of the upcoming INA and prevent duplicate efforts. Chances are good that someone has had to meet similar business needs in the past. A similar project or organizational INA may have been done by your organization or others.
It is advisable to have an inventory of existing data available during an INA to aid in discussions. This allows individuals to know what data is currently available versus what data would have to be acquired to meet projected needs. The inventory is especially critical when the analysis and information outputs/products are heavily dependent upon currently available data. This information will be needed for data acquisition planning following the INA.
There can be many organizations and people involved in a project or organizational INA. Key players are the INA coordinator, subject matter experts, information specialists, and decision makers. Decision makers provide clarification of direction, expectations and sideboards, determine the priority of the issues, data and work, and may also act as subject matter specialists. They move in and out of the INA process as needed. It is advisable to have the most complete group that time, money, and personnel constraints permit. Effort should be made to include in the INA process all organizations involved in the issues or in the geographic area.
There are five main steps to be accomplished in the INA process:
Step 1. Identify issues and business needs.
The INA participants will identify major issues and business needs that frame the project. This is a key time for decision maker participation.
Step 2. Identify information needed to address issues/business needs.
For each issue/business need, identify key questions that need to be answered or decisions that need to be made. Next, identify any information products/outputs that could be useful to address the questions or decisions. Products/outputs include reports, maps, graphic tables, and diagrams.
The INA Coordinator needs to identify potential common needs between issues and seek clarification from the participants on naming and descriptions after validating that similar products/outputs are truly the same.
Step 3. Identify data needed for information products/qutputs.
Breakdown each output/product into the individual data elements that are needed to construct the product. Each output/product will be described with a map sketch and/or report format. Data elements needed for each will be listed along with source, accuracy, and map scale for each data element.
Step 4. Analyze the INA results.
Develop a prioritized list of data sets needed to accomplish the goals of the project or organization. To this end, the information gathered in the INA is compiled, analyzed, displayed and considered.
Step 5. Document the INA process.
Documentation is essential so that the results of the INA can be used to help plan for data acquisition and database development and to be able to track progress through database implementation. It provides justification of decisions, acts as a tool to hold people accountable, and can be useful to others as a resource for future INA projects. It is good to put the results into a report format and distribute it widely within the project or organization to encourage its use.
The INA process emphasizes data issues. The list of priorities developed through the INA process will be the starting point for determining how to bring all the data needed into one place in a usable format. Someone, most likely a team, now needs to evaluate existing data, determine cost estimates for new data, and connections with other data acquisition efforts. Detailed plans will need to be made for developing each data theme or database that does not already exist in digital format. The list of data elements produced during the INA is a starting place for design of a theme or database, but it must be further developed and refined, preferably with interdisciplinary, and possibly inter-organizational, involvement to maximize usefulness of the resulting data. Existing standards should also be considered (see Review of Existing Standards, p.67).
Groups are forming to help insure that data standards meet both the needs of the agencies and the field organizations. Be sure to share information collected for this project with these groups, and validate decisions. This is critical for implementing consistent data and systems in support of ecosystem management.
The trust level of managers in major long term systems development being done on time and at budget is almost nil. History backs up their pessimism. Major development projects more often than not take over twice as much time to complete as estimated. Developers have a history of promising to do too much for too little. The hard facts are that developing long term, flexible, multi-use data stores and systems takes major investment and time. And these are the kind of systems that would be highly valuable to ecosystem management where data is used by multiple functions, is needed for monitoring change and trend over time, and needs to support questions and analysis that are both known and yet to surface. A harder fact is that development cost has historically been less than twenty percent of the total life cycle cost of a system. Support and maintenance have typically grabbed eighty percent of life cycle cost.
A variety of information planning techniques, standards, and tools that fall under the general heading of "life cycle planning" can be used to better estimate costs and time lines, improve information products, and lower support and maintenance costs. The most effective approaches are those which integrate management of data, technology (information systems) and business (work) processes. These are different but closely interrelated types of information management components. Data cannot be managed separately from the business; nor is it independent of technology. Some of the techniques a manager might apply include: life cycle management (LCM); information engineering (IE) or computer aided systems engineering (CASE); decision analysis; business process reengineering (BPR); and rapid application development (RAD), to name a few.
Each management technique provides value in selected applications. Business process reengineering, for example, is a tool for streamlining work. It calls for designing processes in a "right-to-left" fashion by first identifying customers and desired outcomes, then working backward to determine outputs, activities, and, finally, inputs (including data) needed. Information engineering or CASE methods can be helpful in developing integrated information system components to enable sharing and interoperation with others. The aim of rapid applicatin development is to shorten development time and improve quality by bringing the customer into the process earlier to test and provide feedback. And a dose of decision analysis can be critical, for example, to avoid "spending a dollar on a nickel decision."
Following this section are references for further depth on integrated systems planning. The minimum a manager needs to have a feel for are the steps of structured system life cycle management and the roles that need to be implemented. Following these steps and establishing these roles will help control typical problems associated with large projects , including failure to keep users involved, lack of accountability, hidden costs, lack of completion, over commitment of resources, failure to maintain systems, and frequent changes to requirements.
Figure 4 illustrates the life cycle management process. Phases include:
Initiation: The purpose of the initiation phase is to have a clear understanding of the objectives, benefits, scope, personnel, authorities, reporting relationships, deliverables, and resources (money, people, equipment, materials, and facilities) needed to carry the project through its entire life cycle. It is also appropriate to specify conditions under which a project may be "sunset." The primary product of the initiation stage is a contract (or charter) between management and project staff outlining all the above expectations.
Remember that typically by far the largest commitment of staff time and money come during the operations and maintenance phase. It is critical to the success of the project that one not proceed with development without a signed agreement on deliverables and support. Simply stated, "don't start something you can't finish."
Development: Development usually involves: 1) analysis, 2) design, and 3) buy or build the solution. The goal of this phase is to deliver documented, tuned and tested information components, complete and ready for installation. During this phase computer aided systems engineering (CASE) is appropriate. A project work plan with milestones and performance measures should be laid out in as much detail as necessary. Oversite and quality assurance committee(s) should be appointed.
In an integrated information environment, information systems are configured around and assembled from sharable components. Successful system integration, therefore, is dependent upon project deliverables being completed as planned and on time, in order to meet the schedule of other coordinated efforts.
Beware of "scope creep" during the development stage. This is the primary cause of cost and time overruns. It happens as system developers and end users find ways to improve on the originally agreed upon solution. It takes a lot of personal and corporate discipline to document and lockin designs and postpone further enhancements until the next scheduled revision.
Phases of Life Cycle ManagementInitiation Phase
Mission analysis Concept evaluation/prioritization Project charter Establish basis for life cycle O&M budget and resources Development Phase
System analysis System design Lockin design, concurrently begin the next revision System buy or build, document, tune, test Implementation Phase
Migrate/convert/retire Establish feedback/monitoring/maintenance processes Organizational training and acceptance Operations and Maintenance Phase
Ongoing training Respond to feedback on quality of service/ changes needed Provide access and support Archive records (cycle back to earlier phases as needed) |
Figure 4. Phases of Life Cycle Management
Operations and Maintenance: The final phase is operations and maintenance (O&M). Here the component is used for production work. It is important to have smooth running systems with effective performance monitoring and user feedback mechanisms. The goal is to support users in their efforts to efficiently access, use and manage the information they need for their work.
New projects and component life cycles will be initiated as changes are needed that exceed routine operations and maintenance support. Optimally, projects should be scaled to deliver information management products two or three times a year; more frequent changes may be organizationally indigestible, depending on the technology and impact of the change. Less frequent changes may cause a project effort to lose momentum.
A key to successful LCM is to have well defined tasks for each phase, and persons assigned to life cycle management roles prior to implementation. These roles include the following:
| System Owners | Sometimes known as "sponsors," are the organization or person corporately responsible for a particular corporate business/mission and its associated information system components. This responsibility includes: 1) advocating and ensuring effective use of the system, 2) budgeting resources, 3) ensuring Life Cycle Management roles are filled , 4) giving end user support, and 5) collaborating in setting priorities and integration goals. |
| Project Manager | The organization or person assigned responsibility to implement a change that exceeds normal operations and maintenance capability. |
| Customer Support Specialist |
Sometimes known as a "helpdesk," is the point of contact for support or troubleshooting, and elevating problems or change requests to where they can be addressed. |
| Information Component - Steward |
The organization or person responsible for maintaining assigned information system components, such as application modules or data structures. They ensure components accurately serve business requirements and are logically consistent with other integrated components. Stewards have a key role in component configuration and change management. They are expected to define needed changes to components based on stakeholder feedback. |
| Information Area Coordinator |
Provides a focus for communication about all information systems within a broad corporate mission/business area and ensures all information systems are strategically defined and scoped. |
The authors thank Margaret Connelly, Dora DeCoursey, Tim Quinn, all of the USDA Forest Service, Region 6, and Teresa Valentine, Oregon State Service Center for GIS, for their contributions to this section.
References
(editor note: this list needs considerable development)
For detailed information on the INA process, including examples of forms:
A good technical series for information engineering are the Richard Barker books:
Barker, R. 1990. CASE*Method: Entity Relationship Modelling. Addison-Wesley Publishing Company
Barker, R. and Longman, C. 1992. CASE*Method: Functional Modelling. Addison-Wesley Publishing Company
Barker, R. 1990. CASE*Method: Tasks and Deliverables. Addison-Wesley Publishing Company
Clegg, D. and Barker, R. 1994. CASE*Method Fast-Track: A RAD Approach. Addison-Wesley Publishing Company.
Other References:
Linden, R. M. 1994. Seamless Government: a Practical Guide to Reengineering in the Public Sector. 294 pp.
Department of Defense. DODHDBK 287 2167 A Tailoring Guide.
Department of Defense. MIL MST 1521B Technical Review/AUdis.
Department of Interior, Bureau of Land Management. 1993. Automation, Information Resources Management, Modernization, and Life Cycle Management. Manual 1261
Inmon, W.H. 1992. Building the Data Warehouse. QED Technical Publishing Group, Boston.
272 pp.
Yourden, E. 1982. Managing the System Life Cycle,.Yourden Press New York
The need for standards and data sharing is perhaps the mostly widely held area of agreement among practicioners of ecosystem management. However, progress in this arena is slow at best. Most of our lessons learned have occurred in the area of sharing data processed by geographic information systems (points, lines, polygons, associated attributes) . There is increasing recognition, however, of the need to make our census and plot data more compatible and accessible as well. This section will discuss barriers to better sharing of data , success factors, and selected elements of information management and project design that are particularly relevant to multi-disciplinary, multi-organizational projects.
Barriers to Better Sharing of Data
Often we are frustrated with the pace of coordination in information management endeavors. There are strong factors affecting progress in this arena:
Organizational culture and tradition. Organizations built on a culture of decentralization and greater empowerment discourage a shift towards more standardization.
Resistance to change. This can particularly be a barrier when so much other non-optional change is impinging upon an organization or individual.
Reduced budgets. Reduced budgets have led to a number of consequences that affect better sharing of data.
Lack of trained personnel in the emerging fields of information management, information engineering, software, and hardware technologies. This may be simply the result of the lack of resources to train or acquire the necessary skills, but may also indicate a lack of understanding of the need for these skills from traditional resource managers. Accompanying this barrier is the communication breakdown that often occurs between resource specialists unfamiliar with the technology and technology specialists unfamiliar with the resources.
Reduced budgets should be a strong motivator for being more efficient. However, if efficiency is measured at very local scales, reduced budgets will not result in accelerated efforts to share development and support costs.
Balancing long term benefits against short term needs. In most cases, short term needs prevail, because of their real or perceived importance. The work involved in developing agreements requires a long term commitment of both people and dollar resources.
Expectation control. Parties or potential parties to an agreement often want finer data than can be obtained at reasonable cost. This is particularly true for landscape and sub-regional efforts. The task may be seen as filling in holes in existing fine level agency data or extending the data beyond its present bounds.
Administrative problems. Hurdles in formulating interorganizational working agreements (memorandums of understanding), system incompatibilities, difficulties in communicating data and work plans, budget approval processes these all become more of a barrier in interagency cooperative efforts because dealing with them is, proportionally, a bigger part of the job. If organizations wish to encourage cooperation at the local level in any aspect of ecosystem management, removal of administrative barriers needs to be a priority.
An important flexibility factor is geo-referencing done to established standards. GIS overlay capabilites are a major integrative factor. They don't work well with data that is poorly tied to the ground. Change detection (essentially sharing data across time) also requires careful geo-referencing of data.
The collection and mapping of basic instead of interpreted data also greatly increases flexibility. An example of an interpreted map is a map that depicts riparian areas without providing a classification or inventory of the basic "data" that is the foundation for making the riparian interpretation. Since the criteria for developing the riparian map is not directly derived from a consistent set of basic data, it may be impossible to recreate the map from different sources or even from the same source. A riparian map tied directly to basic data will anser the same question consistently. If the criteria or definition of an interpretation changes, a new map can be generated easily.
Credibility and flexibility do not come cheap. The maxim "good, fast, cheap-- choose two" holds true.
Document data quality. When queried for factors that make for successful sharing of spatial data, many resource specialists will respond that the data should be GIS data (i.e., in digital form). Often also a requirement for "good enough" accuracy and resolution is expressed. This latter statement brings up the question "good enough for what?." Seldom is data specifically developed to meet all needs (assuming these could even be predicted in a field as changing as ecosystem management). Agencies and organizations that invest heavily in spatial data do so for specific purposes. Commonly there is a lead organization, function, or level that assures its need is met. How to do this and also meet the needs of partners is the challenge.
Accuracy and resolution recommendations are highly dependent on the application to be made of the data. The most precise data are not always the best for a given project; computational requirements, time available to reach a decision, and the precision of other data to be integrated may preclude the need for the most precise data for every use (Mapping Science Committee 1995). The basic minimum requirement for accuracy is to assess accuracy of data and to report it, whatever the accuracy may be. This allows evaluation of the data for a spectrum of uses. It conveys the reliability of the data to a land manager, a judge, or other interested party. Data about data is called "metadata." See Review of Existing Standards (p.67) and Inventory and Appraisal of Existing Data Resources (p.70) for information on metadata requirements.
Define a common reference system. Putting date in digital form alone does not insure that it is sharable, nor does including metadata. Data from different producing organizations is likely to have differences in data definitions as well as in resolution, accuracy, and other data quality components. Upfront standards efforts can minimize these differences over time. An immediate need to integrate, however, may be daunting to the point of overwhelming. Given that analysis will deal with the overlay and edgematching of different layers and coverages, the bare minimum for such efforts to be successful is a common way to reference all data layers to the face of the earth. This is served by developing or acquiring common foundation data to which all other data is spatially referenced. Digital terrain elevation models and digital orthoquads are examples of foundation data. This foundation data is so fundamental that every effort should be made to make it as broadly applicable as possible. The National Spatial Data Infrastructure provides strong guidance in this regard (Mapping Science Committee, 1995).
Define core data. With the foundation data of a shared system in place, the next need is to raise the beams that will support generation of multiple analytical and communication products. This basic data if often called "core" data. Attempts to define core data for ecosystem management need to recognize the influence of scale and the temporal dimension. Figure 5 is an illustration of a recent attempt to define ecological assessment framework data. Any ecological assessment at whatever scale needs particularly to reach early and fast agreement on the environmental factor framework data. These data are semi-permanent . They are of great value not only in overlay and modelling processes for product creation but for extrapolating ecosystem pattern/process relations from sample areas to unsampled areas.
Define standards. Consistent information enhances our ability to dialogue with each other and perform analysis of ecosystems and watersheds across artificial boundaries, through sharing of data, maps, and a common language. A major hurdle in developing consistent information is defining standards. In a field in such a rapid state of development as ecosystem management, it is advisable not to perfect standards but recognize that in information management as in other aspects of ecosystem management, adaptive management will be necessary. Clearly define the scope of the standards effort (subject matter, geographic area, partners) and review carefully exisiting standards for their applicability. Seek input widely but realize that consensus is not necessarily achieveable.
Agreement on consistent standards does not necessarily equal consistent information, however, it is a starting point for placing information in a format that allows for comparing and understanding significant similarities and differences.
The ability to implement agreement on a consistent standard will be variable. The need to maintain and track historical data may mean retaining specific definitions and processes (that either fall outside the agreements or are included to satisfy a specific, limited need) for an indefinite period of time.
The most difficult step following agreement on a common language is translating the language into a workable, userfriendly, shareable coverages and data bases. Minimize time slippage between developing standards and implementing them. Gaps between developing standards and developing coverages or databases jeopardize the effort that has gone into building the standards. A decision to pursue standards should go handinhand with a decision to build, whether the build be incremental or sweeping.
Although frequently the argument made for defining ecosystem management core data is the need to join together to build new data resources, it is equally important to define standard data to prevent drowning in data. The Internet will remove the technological and communication barriers that have kept us from being mired in too much data. Agreement on core data categories and standards will allow standard data sets to be highlighted and segregated from the general flood.
| SCALE | CULTURAL FEATURES | ENVIRONMENTAL FACTORS | SPECIES DISTRIBUTION | EXISTING VEG | ECOLOGICAL STRATIFICATION | POLITICAL STRATIFICATION |
| Bio- geographic | Land Use and Administrative Boundary |
Potential Veg Class/Formation, Climatic Type, Physiography | Range Maps |
Class/ Formation | Biome/Division | Country, State, Province |
| Region | Land Use Capability, Ownership, Managed Areas, Road Density |
Potenial Veg Series or Groups of Series, Climatic Province/Region, Landform, Geology |
Population Centers, Range Maps | Alliances, or Groups of Alliances |
Province, Section | State, Province, County |
| Landscape | Zoning, Roads |
Potential Veg Association, Local Climate, Topography, Soils |
Endemic Species, Populations | Associations, or Groups of Associations |
Subsection, Land Type | Country, Managed Area, District, Population Centers |
| Site | Site Use, Management Prescription |
Potential Veg Association or Phase, Topography, Elevation, Aspect | Species, Population | Association or Subassociation | Land Type, Phase | Specific Ownership |
Figure 5. Information themes at different scales (Jensen and Bourgeron, 1996).
Set clear lines of authority. Sharing datasets in some cases means having the opportunity to use a data set available for a price. We share the datasets but not the responsibility for building, updating, or otherwise maintaining them. Increasingly coordination efforts, particularly those at regional and local levels, are more multiparticipant. The parties are working together to fill in gaps in coverages or improve those coverages. In these situations there must be clear lines of authority and responsibility (Carter 1992). Where no one agency or organization is given this responsibility, progress at best is very slow.
Make the user a partner in data clean-up. Data is seldom good hot off the press. As data is used, errors are discovered and data improves. With nationwide data there is a considerable risk that errors will not be reported back, but corrected locally only. Making the user a partner in data cleanup through established procedures for update and error reporting helps.
Set up a process for change management. In general, shared data requires a structured method of assessing and managing changes to standards, data, processes, and technology. Change requests need careful management, because what looks like a simple change to make can have a far-reaching impact elsewhere. A general process for change management looks like the following:
Roles in this process need to be defined and marketed. The process needs to be clear and documented.
Make Data Widely Available. For natural resource management to be socially acceptable, decision making must be a public and participatory process (Cornett 1994). Multiple partners using the same readily manipulatable information to generate alternatives is an extremely powerful way to foster trust and collaboration. Although this discussion often turns to the Internet and other technological solutions, it is important to recognize that the work here is primarily nontechnical. The challenge to federal, state, and local governments is one that we are just beginning to see scattered efforts address. Franklin (1993) stated it well:
The National Spatial Data Infrastructure (NSDI) has been established in response to the need for people and organizations to be able to better share in the collection, use, and maintenance of spatial data. Geographic data has always been important to natural resource managers and scientists, but with the rapid expansion and use of geospatial data technologies, there are more producers and users than ever, and much more involvement with data through electronic networking and shared/collaborative decision making. As more and more use is made of these new technologies there has been a growing recognition that geospatial data issues must be addressed in order to reduce total costs and to make cooperative use and sharing of data a reality.
The NSDI is the way in which the United States is organizing and coordinating its geospatial data activities. The NSDI was established by Executive Order 12906 on April 11, 1994, and directs Federal agencies to provide leadership in its development, and partnership with State, local, and tribal governments, academia, the private sector, professional societies, and others. The NSDI involves bringing together the technology, policies, standards, and human resources necessary to acquire, process, store, distribute, and improve the use of geospatial data. It is also more than this, as the NSDI is a way of doing business and an attitude of cooperation that recognizes the connection among agencies' data requirements and common pieces of geography and its characteristics and values. The NSDI reflects an agreement to use electronic technology to help find and share spatial data so that together we can create a unified collection of data to better understand our world and to help us make decisions about it.
The Federal Geographic Data Committee (FGDC) is an interagency committee that promotes coordinated development, use, sharing, and dissemination of geospatial data on a national basis through the NSDI. The FGDC provides the Federal leadership in working with all other partners and in initiating and coordination a variety of work activities crucial to the implementation of the NSDI. Key areas include:
Framework: The National Geospatial Data Framework will provide a base on which to collect, register, and integrate information accurately. The Framework follows the idea that users have a recurring need for a few common themes of data to use as building blocks for a wide variety of applications. As the Framework develops it will provide a linking mechanism to integrate themes of data across communities and geographic areas.
Standards: Activities are focused on two different types of standards:
Geospatial Metadata Standards These are standards for descriptive information about a digital geospatial data set or file. FGDC Geospatial Metadata Standards have been issued and all federal agencies are required to have Metadata for all geospatial data collected as of their date of issuance. These standards provide the means to organize, find, and transfer data and are critical for long term sharing and effective data management (See Inventory and Appraisal of Existing Data Resources, p. 69).
Geospatial Data Clearinghouse: The Geospatial Data Clearinghouse is a distributed, electronically connected network of geospatial data producers, managers, and users. The Clearinghouse provides a mechanism to find and share data and information using Metadata descriptions, Internet access, commercial hardware, and commercial or public domain software tools. Online access to data sets may be provided, but initially it is expected that most data will still be obtained by ordering from the producer. Although the clearinghouse concept and metadata defined by the FGDC are geared toward spatial information, other agencies, such as the National Biological Service, are extending the FGDC metadata standard to document other, less geographically based information.
Partnerships: Partnerships are a key part of all NSDI activities. This is a way of doing business where a variety of parties bring contributions of value based on their strengths and expertise, and agree to share responsibilities and commitment, and to share benefits and control in order to improve the geospatial data delivery system.
Throughout a project, and especially at its inception, team members should be aware of and conduct an evaluation of existing standard data exchange, collection, measurement, and analysis methods. Within one or more scientific communities, standards exist that ensure the collection and preservation of environmental information. Standards facilitate the exchange of information, the comparison of similar measurements, and the analysis of information across disciplines.
Standards Organizations
Formal standards are developed by practitioners from academia, industry, and government under the auspices of a recognized standards organization. Typically, recognized standards organizations support the development, balloting, revision, and publication of standards documents produced by the membership. Once a standard is ready for review, approval is made using a consensus process resolving "negative" votes before it can continue. Because standards are created by individuals who have other job responsibilities and because it is a consensus process, standards activities can be slow. In an age of technological innovation it will be a challenge for standards bodies to keep pace. Nonetheless, existing standards should be sought out and endorsed for use within a project.
Formal standards organizations include:
Relevant standards
ISO has convened an international committee (TC 211) to address the development of a comprehensive geospatial data standard. Through workgroups, documents are being drafted on metadata, quality, a reference model, standard features, among other areas, and will become parts of an international standard. This standard can then be voluntarily used in most countries, although compliance can be made mandatory by certain governmental organizations (Committee for European Normalisation, CEN in Europe and through NIST for U.S. federal organizations). A number of other relevant ISO activities exist:
The American National Standards Institute (ANSI) is a private voluntary consensus standards body headquartered in New York. The X3L1 Workgroup is involved in geographic standards.
The ASTM has several committees that address environmental monitoring, assessment, and spatial data standards:
The Federal Geographic Data Committee (FGDC) is authorized under OMB Circular A16 to coordinate the development of geographic data standards within the U.S., engaging both federal and nonfederal participation. Standards for spatial data exchange and documentation (metadata) have been developed and approved through the FGDC. The Federal Geographic Data Committee has
Thematic Subcommittees that are defining information content for the following categories of spatial information:
The following is a cursory list of relevant standards to ecosystems projects:
No single discipline has the exclusive answer to how the total information should be managed. A careful synthesis will be required within each project to select standards based on its orientation and goals. It should be the responsibility of at least one of the project members to monitor relevant standards for the project. Ideally there should be an individual within a coordinating group overseeing the project at a regional or national level that tracks relevant activities and passes recommendations to field operations.
U.S. Federal agency employees are encouraged to participate in the development and refinement of existing standards through private, voluntaryconsensus standards organizations such as ASTM or ANSI. Awareness of agency representation on FGDC thematic subcommittees is encouraged by federal employees. It is through enduser involvement that standards are developed and enhanced to reflect realistic conditions, methods, and information for others to use.
In the early part of a project, or during a project if it is already underway, an inventory of existing information holdings should be made. This would include a summary of each data file or collection in use or required by the project. The inventory should include basic descriptive information that can be used to find information. This descriptive information (metadata) is of benefit to project members, reviewers, and others using project information. Each data set should include the following metadata:
These elements represent a superset of the mandatory and optional fields of information about a environmental information data set to assist in discovery and reuse of the data within and outside the project. Approximately 200 optional and mandatory metadata elements exist in the FGDC content standard for digital geospatial metadata required of all federal agencies for new data sets that include a geographic reference.
Inventory of existing information resources should include holdings of relevant environmental and spatial data managed by other agencies. To assure that data are not collected multiple times, it is a good investment of time to identify data coordination mechanisms in your locale. In most states there is a mapping advisory committee composed of state, federal, and local representatives interested in the collaborative collection of data through combined efforts or contract mechanisms. Geographic Information Systems (GIS) councils exist in many parts of the country to coordinate spatial data activities across multiple disciplines. This is often a reasonable place to begin to establish or expand networks of potential cooperator agencies.
The authors thank John Moeller, Executive Director of the Federal Geographic Data Committee (FGDC), and Doug Nebert, U. S. Geological Survey, for their contributions to this section.
References
Cornett, Z. J. 1994. GIS as a catalyst for effective public involvement in ecosystem management decisonmaking. In Remote Sensing and GIS in Ecosystem Management, ed. V. Alaric Sample, pp. 337345. Island Press, Washington, D.C.
FGDC. 1994. Content standards for spatial metadata. Washington, Federal Geographic Data Committee. FGDC Secretariat, U.S. Geological Survey, Reston, VA. (also available from ftp://fgdc.er.usgs.gov/pub/gdc/).
FGDC. 1995. Development of a National Digital Geospatial Data Framework. Washington, Federal Geographic Data Committee. ( also available from ftp://fgdc.er.usgs.gov/pub/gdc/).
FGDC. 1994. Guidelines for implementing the National Geospatial Data Clearinghouse (Guidelines for Federal Agencies). Washington, Federal Geographic Data Committee, June 3, 1994. ( also available from ftp://fgdc.er.usgs.gov/pub/gdc/).
Franklin, J. F. 1994. Developing information essential to policy, planning, and management decisionmaking: the promise of GIS. In Remote Sensing and GIS in Ecosystem Management, ed. V. Alaric Sample, pp. 1824. Island Press, Washington, D.C.
Green, K. 1992. Spatial Imagery and GIS. Journal of Forestry, 90 (11): 3236.
IRICC. 1996 (draft). Information Needs Assessment: Interorganizational Resource Information Coordinating Council (contact: Margaret Connelly, USDA Forest Service, Portland, OR).
Jensen, M. E. and Bourgeron, P. S. 1996. (draft) Information Themes.
Mapping Science Committee, Board on Earth Sciences and Resources, Commission on Geosciences, Environment, and Resources. 1995. A data foundation for the National Spatial Data Infrastructure. Approved by the National Research Council, Washington, D.C. National Academy Press, Washington, D.C.
Mapping Science Committee, Board on Earth Sciences and Resources, Commission on Geosciences, Environment, and Resources. 1994. Promoting the National Spatial Data Infrastructure Through Partnerships. Approved by the National Research Council, Washington, D.C. National Academy Press, Washington, D.C.
Data Manipulation with Geographic Information Systems
Geographic Information Systems (GIS) have become an integral component in information management, collection, and inventory. The perceptions of what a GIS is and what it does vary greatly from manager to manager on depending on their exposure to the subject. The following short discussion will help provide the manager an overview of the salient aspects of GIS. The discussion contains the definition of GIS, a condensed history, the scope and rationale for a GIS, and the questions a GIS can answer.
A Geographic Information System (GIS) is an organized collection of computer hardware, software, geographic data and personnel designed to efficiently capture, store, update, manipulate, analyze, and display all forms of geographically referenced information.
Many people believe the advent and use of GIS is a recent event. However, the use of GIS principles and techniques have been used for many years by federal, state and local governments as well as private industry. Below is a short historical synopsis of the development of GIS and its principles and techniques:
1960s
-Map overlays were widely used in the 1960's. This was an analog
system by which point, line, and polygonal data were registered
together on a common base map through the use of manual overlays
for unit area analysis.
1970s
-Digital data came into widespread use in the 1970s due to advances
in computer technology. Computer technology allowed the conversion
of point, line, and polygonal data to a digital form. The digital
data was used mainly for map preparation, plot generation, and
data storage and retrieval. The use of overlays was still prevalent.
-Image processing also began in the 1970s. The LANDSAT series of satellites with mutispectral sensors provided a digital record of the reflective values of the surface of the earth. Computer algorithms were developed to analyze and classify these digital reflective values for land cover identification and mapping.
1980s
-Civil agencies developed some of the first GIS software packages
to be widely used in the civilian sector. Two examples of these
GIS software packages were MOSS and GRASS. These packages were
able to capture, store, update, manipulate, analyze, and display
all forms of geographically referenced information.
1990s
-Commercial Off The Shelf (COTS) GIS software packages were developed in the mid 1980s and are being heavily used and marketed in the 1990s. Arc/Info is a prime example of COTS GIS software which is now being used by civil agencies such as the Forest Service, Bureau of Land Management, US Fish and Wildlife Service, and the US Geological Survey to name a few.
GIS is not just composed of hardware and software; it is much larger than that. A successful GIS is a carefully crafted mosaic of hardware, software, mission essential data, trained people and logical and documented procedures. All of these components are absolutely essential and if any are faulty or missing then the GIS will fail.
GIS is rapidly becoming a standard and essential tool in today's business practices. The following is the short list for the rationale for GIS. The following short list is based on the assumption a GIS is implemented and maintained in a professional manner:
A GIS can assimilate a large amount of spatial information, which can be viewed from a variety of perspectives, to assist in making the best decisions in a timely manner.
A GIS can answer a number of questions:
There are a few things a manager should remember about what a GIS is not and they are as follows:
1.) A GIS is not simply a computer system for making maps even though much of its output is in map form.
2.) A GIS does not improve data accuracy or quality, it only assimilates, processes, presents data in many forms and perspectives.
3.) A GIS does not make decisions; it only assists in making better decisions.
As previously stated there are numerous books and articles about GIS which any manager can reference for more information. Hopefully, this short discussion has provided the manager with an overview of the salient aspects of GIS.
References
GIS Sourcebook -- This sourcebook started out as a National Park Service GIS manual and is being transformed into a Departmental guidance manual for use by all Bureaus in the Department of the Interior. It is an exhaustive resource providing operational guidance to GIS users dealing with data issues, hardware and software choices, and associated technologies (i.e. GPS, CADD). This sourcebook is available on-line at http://www.nbs.gov/sourcebook. It is maintained by the National Biological Service (NBS).
The ability to integrate remote sensing imagery in a GIS, used in combination with ancillary data, including Cartographic Feature Files (CFFs), digital elevation models (DEMs), digital orthophotoquads (DOQs), and other resource data layers provides a solid geospatial data foundation to support ecosystem managers.
Advanced Very High Resolution Radiometer (AVHRR). AVHRR data is obtained through the EROS Data Center (EDC) from the Department of Commerce, NOAA weather satellites. It has relatively course spatial resolution of 1.1k x 1.1k pixel size with five spectral bands. Broad area coverage is obtained with a swath width of 2,700 kilometers. Temporal resolution is excellent, with repeat coverage over a given geographic area every 12 hours. Major ecosystem management applications for this data include: vegetation discrimination, vegetation biomass, snow/ice discrimination, vegetation/crop stress, and geothermal mapping. This imagery is currently used by the Forest Service, working in association with the US Geological Survey (USGS), on development of a global vegetation data set and with the United Nations on verification for a global vegetation characterization data base.
Landsat MultiSpectral Scanner (MSS). MSS imagery has a spatial resolution of 80m, spectral resolution in 4 bands, revisits the same geographic location every 16 days and covers an area of 185 x 170km per scene. The Landsat program gathered digital MSS data from 1972 through 1992. The result is a twenty year time span of data that can support evaluations of change in landscapes or land cover over a longer time period than any available earth observation system.
MSS data is available through the EDC working in cooperation with the NASA Landsat Pathfinder Program. Pathfinder efforts are focused on evaluation of global change using available remote sensing technologies. The North American Landscape Characterization (NALC) Project, a component of the Landsat Pathfinder Program, is developing an archive of MSS data and production of three date georeferenced data sets, called triplicates, acquired in 1970's, 1980's and early 1990's. The NALC images are geometrically rectified, georeferenced, and placed into a UTM map projection. Pixels are resampled into a 60m x 60m size format for compatibility with the 30m x 30m Landsat Thematic Mapper (TM) data resolution.
Landsat Thematic Mapper (TM). TM imagery has 30m spatial resolution with 7 spectral bands. The Landsat 5 satellite covers the same geographic location every 16 days, each scene covering 185 x 170km. The Landsat Program is the longest running program in the collection of multispectral, digital data of the earth's land surface from space. The temporal extent of the collection, the characteristics and quality of Landsat data, and the ability to collect new data directly comparable to that in the archive, make Landsat data a unique resource to address a broad range of issues in ecosystem management.
An interagency consortium called the MultiResolution Land Characteristics Monitoring System (MRLC) has been established to purchase TM imagery to cover the United States (not including Alaska). This group has selected the appropriate scenes and developed image processing, georeferencing, terrain correction, clustering and classification strategies. This federal consortium includes: EPA, EMAP; US Fish and Wildlife Service, GAP Program; USGS EDC and National Water Quality Assessment Program (NAWQA); and, NOAA, Coast Watch Change Analysis Program (CCAP).
Airborne video . High resolution video cameras and digital cameras provide a powerful tool for ecosystem management. A camera can be mounted in a small airplane and imagery can be viewed as soon as the flight is complete. Video imagery linked with GPS provided location information for each frame, allowing user to quickly locate specific areas on maps or images. Video images can be digitized and used with other GIS layers.
Digital orthophotos . An orthophoto is an airphoto that has been rectified to remove distortions due to tilt or height displacements. Simply put, an orthophoto has all the qualities of a map but retains the qualities of an aerial photo. Since the early 1970's government resource management agencies such as the Forest Service have been producing analog hard copy orthophotos. They are valuable as a supplement or alternative to standard maps. In the early 90's digital orthophotos became available to resource management agencies. The strength of digital orthophotos is for feature mapping rather than resource classification. They are used as backdrops to other data, and are particularly useful in on screen digitizing processes. Due to their relative newness, the potential of digital orthophotos is just being explored.
Users of digital imagery range from novice, who simply use imagery as a backdrop for overlaying GIS information - to expert users, who do classification of landscapes for input into large area assessments or forest planning efforts. Thus a wide range of users' needs must be met in the form of tech transfer and training to fulfill the ecosystem management mission.
The key to maximizing the benefits of using remote sensing to achieve the goals of ecosystem management will hinge upon (1) personnel understanding how the derived information can be integrated into their information management systems (such as GIS), (2) personnel being trained in the use of these tools for mapping, etc., and (3) personnel having access to remote sensing data.
Access to remote sensing data should rapidly increase for several reasons. These reasons include new cooperative agreements among agencies to share data acquisition costs and improved information access on the information super highway. Currently, many potential users of remote sensing are either not aware of how to access the data or believe the cost of acquiring the data is prohibitively expensive. Improved access to the data, both physically and financially, will likely increase the number of users of remote sensing data.
New data procurements, such as the one recently made by the Forest Service, will provide AVHRR, Landsat MultiSpectral Scanner (MSS) triplicates, and Landsat TM imagery for all lands in the continental United States. This low cost purchase will provide imagery at very little cost to the entire Forest Service (except Alaska) at three different scales. Purchases such as these provide inexpensive and easy access to data and will surely spur interest in using digital imagery for applications such as RPA assessments and forest planning.
Access to the Internet now gives users the capability to preview satellite imagery acquired from several platforms including Landsat. This will allow users to assess the condition of the imagery (e.g., cloud cover) and will ensure that the user is confident in the quality of the imagery before purchasing. Internet access will also provide an easy method for ordering and transferring digital imagery and thus streamline a process that was recently time-consuming and inefficient at best.
Currently plans for several new satellite sensors are in the making, and are scheduled to be launched within the next decade. If these plans come to fruition, users of remote sensing data will have several new options. These sensors will allow for different perspectives of the landscape due to differences in spectral, spatial, and temporal resolutions of the imagery they will provide.
Data Collection and Inventory
Generally speaking we collect data to answer specific resource
questions And, acquiring this data, whether efficiently or not,
generally incorporates a rich texture of methods, objectives,
and assumptions in achieving the goal of providing information
that is of such value that the lack of such information would
lead to policies or actions that would degrade the environment
and/or communal standard of living. The frustration and hope
of achieving a more integrated approach to natural resource inventory
and monitoring were set forth in former Forest Service Chief Max
Peterson's keynote address to the Inventory Integration Workshop
in October 1984:
Generally speaking, approaches to data collection have less to do with a particular discipline and more to do with the physical parameters of source, content, and scale of the data to be collected. That is, any contrived functional category of data related to any issue such as recreation, vegetation classification, wildlife habitat, cultural or socioethnographic perceptions has within it a complex fabric of source, scale and content characteristics which may be of interest in developing resource management policy or plans. In fact, the very policies or plans that result from the analysis of the collected data have content and scale ramifications. Analysis applied to a scale for which the data is not matched or is inappropriately integrated can be disastrous indeed.
Data collection approaches can be seen as a continuum, ranging from purely functional with a single variable of interest at a single point in time to fully integrated with multiple variables and scales over multiple time frames. For purposes of our discussion we will place this continuum into three discrete categories:
o Linked functional
o Integrated/Interdisciplinary
Historically, data collection efforts tended to start out at the functional level because managers or researchers were looking to answers to specific, narrowly focused questions. This method kept individual process costs down and maintained process control, thus reasonably assuring timely results. Unfortunately, as the resource questions have become more complex and funds more restricted, these efforts have a tendency to, at best, stall at level 2, the linked functional, and at worst remain at level 1. There are many reasons for this, cost being most often cited, both in terms of personnel and administration. But many are simply institutional and bureaucratic in nature. If the longterm benefits are unknown, and can not be reasonably predicted, the momentum to reconfigure two or more functional surveys into one interdisciplinary process may be justifiably lacking. On the other hand in cases where there is clear benefit, approaches to breaking down the institutional barriers should be sought.
The concept of integrated data gathering is important and should focus on:
o development of consistent, compatible information,
o reasonable development, retrieval, and analysis costs.
Sharing is facilitated by using common protocols and standards. These may differ at different information scales and thus pose problems in sharing data and information. Approaches must seek to standardize terminology by beginning to identify layouts of output products, then defining what variables are needed to produce those products. Existing standards and information should be used whenever possible to develop a framework for organization, compilation, and presentation. One way to ensure that the most appropriate approach is used is to view the opportunities for integration from a more global perspective. Table 1. provides a simplified representation of how scale and objective can be related.
Key ingredients of integrated resource inventory and monitoring processes are standard outputs, inputs, definitions, and data base links. Additionally, the influence of time and space must be considered. Lund (1984) included the time and space elements in his delineation of four basic types of integration:
1) Multi-location integration refers to inventory and monitoring strategies that compare many sites with the intent of implementing an action on at least of them or to pool or aggregate resource data from all the sites for a common purpose or objective.
2) Multilevel (scale) integration recognizes that inventory and monitoring data may be required to support six decision levels including international, national, organization/agency, administrative unit, activity, and project. Uses 'core' variables to link various scales of data which requires linking of management direction to ensure outcomes.
3) Multiresource (interdisciplinary) integration refers to inventory and monitoring strategies that are designed for a variety of users from their inception and thus have included a multiuser needs assessment. These strategies will frequently use a common sampling frame, but may require a linked approach to address budget and efficiency constraints.
4) Temporal integration refers to a need for trend data where collection strategies must provide for consistent methods of measuring key variables through time. Information needed for monitoring and evaluation must be considered during the initial design process.
Census A census refers to a complete enumeration of the population of interest. If economical, this method is highly desirable because it minimizes the risk that you do not have a representative view of the population of interest. This approach is most useful when the population is small or the data so critical that the expense is justified.
Mapping surface features of a landscape, benthic
(underwater) terrain, or object using remote sensing methods
including photography, videography, satellite imagery, radar,
sonar, laser imaging. The photography and videography may be
from either aerial or ground perspective. While these methods
can be expensive, cost may be reduced by grouping variables of
interest into coarser classes.
HIERARCHICAL RELATIONS BETWEEN ASSESSMENT SCALES, TYPES,
AND VARIOUS ECOSYSTEM DELINEATIONS
| ASSESSMENT | BIOPHYSICAL ENVIRONMENTS | EXISTING CONDITIONS | |||
| SCALE | TYPE | TERRESTRIAL UNITS | AQUATIC UNITS | VEGETATION UNITS | SOCIAL UNITS |
| Broad | Global | Domain | Zoogeographic Region | Class | Continent |
| Continental | Division | Zoogeographic Subregion | Subclass | Nation | |
| Regional | Province | River Basin | Group | State | |
| Mid | Sub Region | Section/ Subsection | Sub-basin | Formation | County |
| Fine | Landscape | Landtype Association | Watershed | Series | Community |
| Land Unit | Landtype, Landtype Phase | Valley Section , Stream Reach |
Association | Neighborhood | |
| Site | Ecological Site | Channel Unit | Group | Household | |
Table 1. How scale and objective can be related.
Nonstatistical sample subjective, often costeffective precursor to statistical sampling. Cunia (1982) lists four situations in which this form of sampling may be preferred: 1) variations between elements of the population are large and sampling is expensive; 2) the needs for information about a population are immediate and a decision must be made before a wellexecuted statistical sample can be executed; 3) funding is short or unavailable and the only alternative is to use existing information and extrapolate to the population of interest; and 4) approximate knowledge of some of the population parameters are needed to design an efficient statistical sample.
Statistical sample the collection of data, representative of the population of interes, in a scientifically acceptable manner. The sample is generally probabalistic in nature with every unit of the population having a positive probability of selection and these probabilities are known and are independent of the person taking the sample. The basic tenants of this method were best detailed by theoretical statistician A.N. Kiaer of Norway in 1899 (Seng, 1951) when he observed that the two most important elements of successful sampling were proper representation and rational selection of sample units. He further stated:
o The representative method of investigation is applicable, not only in social and economic fields, but also in agriculture and forestry.
o To obtain a representative selection of sample units, it is necesary to group different communities under investigation. Thus in social studies, the towns and country communes should be differentiated, and be further separated by size (large, medium, or small) and location (coastal or inland). He pointed out that this principle of grouping (or stratifying) homogenous parts of a country must be applied with care to obtain a representative sample.
o If possible, two or more different methods of selecting a representative sample should be used so that the results of the inquiry are credible, and proof of the usefulness of methods can be attained.
o It is important to study and develop the practical and theoretical aspects of the method so that proper limits can be set to representative statistics.
Still sage advice nearly 100 years later.
Approches and methods should be in accordance with inventory objectives using commonly accepted and agreed upon standards with well defined quality assurance procedures. It is important to note that many sampling tools may be used within the same inventory approach (table 2).
Implementation
Selecting appropriate sampling approaches, methods, and tools will depend primarily on the initiative of the dominant user and their willingness to view the larger landscape in the decision process. Accepting a functional approach to fulfilling a need for inventoty and monitoring information is neither good nor bad if the investigator has evaluated the alternatives to take a braoder approach and found them less efficient. This will be especially true of one time event or limited scope issues.
When evaluating existing data and approaches be careful not to be blinded by the original intent of the data collection. Many historic vegetative inventories, for instance, were primarily focused on timber. And, while the dbh of trees in a stand are the bread and butter of harvest planners, they also carry a great deal of information for plant ecologists seeking data on vegetative structure and seral stage. Other functional studies may additionally be linked through overstoryunderstory relationship research. These types of linkedfunctional efforts, while not optimal, begin to shine new light on opprotunities to develop more useful integrated inventories.
The following are considerations for collecting inventory and monitoring data when functional data or approaches exist:
o does the study require acquisition of new data?
o is there a narrow or broad need for the data?
o what are the periodicty and scale issues related to the data?
Functional inventories will continue to play a vital role in data collection for research. The important issue is how to use this information to establish its relative value and on this basis determine the priority of including it in an integrated inventory. One convenient 'node' that can be included in functional surveys is geographic reference information collected in a standard protocol such as those provided by the Federal Geographic Data Committee (FGDC).
Successful inventory and monitoring planning will rely on coordination among the various researchers, managers, and users of the data and information. Information needs, objectives, and data collection designs must be the product of a communal effort that realizes that not all inconsistencies and contentious issues will be resolved. There is no silver bullet, but a more global approach in the planning process goes a long way to alleviating future problems and provides for more efficient inventory and monitoring strategies.
Sampling tools commonly used for various sampling methods
| Sampling Tool | Census | Mapping | Non-statistical | Statistical |
| Telephone survey | [to be filled in] | |||
| Mail questionnaire | ||||
| Personal Interviews | ||||
|
Dimensional plots
- (circular, rectangular, etc) | ||||
|
Point sampling
- Horizontal and vertical | ||||
| Transect sampling | ||||
| Digital photography | ||||
| Videography | ||||
| Satellite imagery | ||||
| Radio telemetry | ||||
| Radar/Sonar | ||||
|
Profile/content analysis (soils) | ||||
|
Volume/content/flow sampling (air & water) | ||||
| Banding/tagging | ||||
| Discipline | Census | Mapping | Non-statistical | Statistical |
| Aquatic fauna | [to be filled in] | |||
| Aquatic flora | ||||
| Water quality | ||||
| Terrestrial fauna | ||||
| Terrestrial flora | ||||
| Soils/Geology | ||||
| Avian studies | ||||
| Air quality | ||||
| Recreation | ||||
|
Archeology/ Cultural Studies | ||||
Table 2. Sampling tools and methods
The following references can serve as a starting point for developing a sampling strategy. This is by no means an exhaustive list but by virtue of these references and their associated bibliographies the researcher or manager can begin to blaze a trail of critical questions and answers to develop an effective sampling strategy. The list is subdivided by broad topic areas for ease of use.
Draper, N. R.and H. Smith. 1982. Applied regression analysis. 2nd ed. New York: John Wiley & Sons, Inc. 407p.
Houseman, E. E. 1975. Area frame sampling in agriculture. Statistical Reporting Service No. 20. Washington, DC: U.S. Department of Agriculture, Statistical Reporting Service. 79p.
Freese, F. 1967. Elementary statistical methods for foresters. Agric. Handbk. 317. Washington, DC: U.S. Department of Agriculture, Forest Service. 87p.
Husch, B., C. I. Miller, and T. W. Beers. 1972. Forest Mensuration. 2nd ed. New York: Ronald Press. 410p.
LaBau, V.J.and T. Cunia T.(eds.) 1990. Stateoftheart methodology of forest inventory: A synposium proceedings. Gen. Tech. Rep. PNW263. Portland OR: U.S. Department of Agriculture, Forest Service, Pacific Northwest For. Exp. Sta. 592p.
Lund, H.G.; Thomas, C.E. 1989. A primer on stand and forest inventory designs. Gen. Tech. Rep. WO54. Washington, DC: U.S. Department of Agriculture, Forest Service. 96p.
Lund, H.G. 1986. A primer on integrating resource inventories. Gen. Tech. Rep. WO48. Washington, DC: U.S. Department of Agriculture, Forest Service. 64p.
Lund, H.G.; Thomas, C.E.(Technical coordinators) 1995. A primer on evaluation and use of natural resource information for corporate data bases. Gen. Tech. Rep. WO62. Washington, DC: U.S. Department of Agriculture, Forest Service. 168p.
Hansen, Mark H.; Frieswyk, Thomas; Glover, Joseph F.; Kelly, John F. 1992. Eastwide Forest Inventory Data Base: User's Manual. USDA Forest Service. North Central Forest Experiment Station, General Technical Report NC151. 48p.
Woudenberg, Sharon W.; Farrenkopf, Thomas O. 1995. Westwide Forest Inventory Data Base: User's Manual. USDA Forest Service. Intermountain Forest Experiment Station, General Technical Report INT317. 67p.
U.S. Department of Agriculture. 1992. Forest Service resource inventories: An overview. FIERR Staff Document. Washington, DC: U.S. Department of Agriculture, Forest Service. 39p.
Haugen, G. (coord.). 1994. Section 7 fish habitat monitoring protocol for the Upper Columbia River basin. PNW/INT/Northern Regions. U.S. Department of Agriculture, Forest Service. 61p.
Verner, J., M. L. Morrison and C. J. Ralph,.(eds.) 1984. Wildlife 2000: Modeling habitat relationships of terrestrial vertabrates. Madison, WI: Univ. of Wisconsin Press. 470p.
Rosgen, David. 1994. A classification of natural rivers. San Francisco, CA: Article Clearing House Info. Store, Elsevier Science B.V. Catena 22 (1994):169199.
U.S. Department of Agriculture; U.S. Department of Interior. 1993. Biological monitoring ot lakes and reservoirs. AEMC Tech. Bull. No. 2:93. Logan, UT: National Aquatic Ecosystem Monitoring Center, Utah State Univ. 26p.
U.S. Geological Survey, Intergovernmental Task Force on Monitoring Water Quality. 1995. The strategy for improving waterquality monitoring in the United States. Reston, VA: U.S. Department of Interior, U.S. Geological Survey, Office of Water Data Coord. 117p.
National Research Council. 1994. Rangeland health: New methods to classify and monitor rangelands. Comm. on Rangeland Classification, Board on Agric. Wash., DC: National Academy Press. 180p.
Uresk, D.W. 1990. Using multivariate techniques to quantitatively estimate ecological stages in a mixed grass prairie. Journal of Range Mgmt. 43(4):282285.
Yaun, S.; Maiorano, B. M. Yaun, S. M. Kocis, and G. T. Hoshide. 1995. Techniques and equipment for gathering visitor use data on recreation sites. 95232838MTDC. Missoula, MT: U.S. Department of Agriculture, Forest Service, Missoula Tech. and Dev. Cntr. 77p.
Hall, F.C.; Bryant, L.; Clausnitzer, R.; GeierHayes, K.; Keane, R.; Kertis, J.; Shlisky, A.; Steele, R. 1995. Definitions and codes for seral status and structure of vegetation. USDA Forest Service. Pacific Northwest Research Station, General Technical Report PNW363. 39p.
Heywood, V.H. (ed.) 1995. Global biodiversity assessment. London, England: Cambridge Univ. Press. 1140p.
Lund, H.G.(ed) 1993. Integrated ecological and resource inventories. Proceedings national workshop, Phoenix, AZ, April 1216,1993. WOWSA4. Washington, DC: U.S. Department of Agriculture, Forest Service, Watersheds and Air Staff. 177p.
Schreuder, H.T., T. G. Gregoire,and B. W. Geoffrey. 1993. Sampling methods for multiresource forest inventory. New York: John Wiley & Sons, Inc. 446p.
U.S. Department of Agriculture. 1984. Proceedings of the Inventory Integration Workshop, Portland, OR, October 1519,1984. Washington, DC: U.S. Department of Agriculture, Forest Service, Range and Timber Mgmt. Staff. 165p.
U.S. Department of Agriculture. 1992. Taking an ecological approach to management. Proceedings national workshop, Salt Lake City, April 2730,1993. WOWSA3. Washington, DC: U.S. Department of Agriculture, Forest Service, Watersheds and Air Staff. 241p.
U.S. Department of Agriculture. 1994. A federal guide for watershed analysis. Ver. 2.2. Portland, OR: Regional Ecosystem Office. 26p.
U.S. Department of Agriculture; U.S. Department of Interior. 1995. Ecosystem analysis at the watershed scale: Federal guide for pilot watershed analysis. Ver. 1.2. Portland, OR: Regional Ecosystem Office. 202p.
Accuracy - The degree to which a measured value is known to approximate a standard, true, or accepted value. Contrast with precision.
Attribute Data - Attributes are associated with features and describe their characteristics, measurements, and other facts or observations about them. They are usually analogous to a data element or a column in a data base table.
Basic Data - Data that is not interpreted. Vegetation would be basic data from which old growth could be interpreted.
Business Process Re-engineering (BPR) - The activity by which an organization defines its goals and the work processes by which it attempts to achieve them and redesigns and re-implements those processes. This can result in very much improved computer systems, which remove redundant activities and streamline and optimize the use of an organization's resources.
Coincident Line - Any polygon edge or line on a thematic layer that shares a common location with any other polygon edge or line on another layer. For lines to be coincident, an identical set of coordinates must define them on all layers. Any polygon edge or linear feature on a thematic layer that coincides with one or more other polygon edges or linear features from any other layer(s). A line is considered coincident when a complete match of X,Y coordinate for X,Y coordinate exists for all coordinate pairs defining the feature.
CASE - Computer-Aided Systems Engineering is the combination of graphical, dictionary, generator, project management and other software tools to assist computer development staff in engineering and maintaining high-quality systems for their end users, within the framework of a structured methodology. Sometimes referred to as Computer-Aided Software Engineering.
Database - A collection of related information about a subject or theme organized in a useful manner that provides a base or foundation for procedures such as retrieving information, drawing conclusions, and making decisions.
Geographic Information System (GIS) - An information system that facilitates the capture, storage, retrieval, analysis, and display of geographically referenced data.
Global Positioning System (GPS) - Method of using signals received from satellites to determine a position on the earths' surface.
Integrated Inventory - A map or layer that provides common units in the landscape that support the resource information needed to meet the Agency's mission. For instance, an integrated, or common, vegetation unit is designed to meet the objectives of the majority of functional areas that collect data about, and use, vegetation information. See also Inventory.
Inventory - A detailed list or survey that describes quantity, quality, and/or location of items within the survey, e.g., ecological types, plant communities, or channel units. Classification and identification precede inventory and/or mapping. Some items can be inventoried without mapping, e.g., the number of nest trees in an area. Other items, such as ecological units, must first be mapped to be inventoried. After they are mapped, location and acres of specific units can be determined. Usually an inventory is conducted by sampling the geographic area in questions and statistically using that sample to describe the population.
Layer - Map of a subset of all the features that exist within the geographic limits of the map. The subset may be chosen for a variety of reasons, the most common being thematic. A thematic layer includes features belonging to a particular theme or subject, like transportation, vegetation, boundaries, etc. Layers may be hard copy or electronic (soft copy). Layers must be in the same coordinate system, at the same scale, and properly registered. Layers prepared in this fashion may be overlaid, enabling geographic analysis between layers.
Map Unit - A map unit consists of a classification or name, and a description, applied to a polygon. Map units are attributionally unique but may classify and describe one or many polygons. The map unit description usually represents a summary of information about a number of similar polygons, rather than describing a particular polygon. The map unit name is derived by applying rules and conventions to the descriptive information.
Mapping - Mapping, in its most generic sense, is the process of using points, lines, or polygons to represent spatial features on a map. Ecological mapping is the science of delineating repeating patterns of relatively distinct units on the landscape.
Metadata - Information or documentation about data. A data dictionary contains metadata (rather than the physical data itself).
Point - An object which has no dimension, but has specified geometric location by a set of coordinates. Though all geographic phenomena have dimension, their expression on a map as a point symbol is determined by scale. Examples of geographic phenomena usually symbolized as points on large-scale maps are wells, weather stations, and navigational lights. In contrast, an airport which might be shown as a polygon outline of actual runways on a large-scale map, might be shown as a point symbol on a small-scale map.
Polygon - Any two-dimensional figure having measurable area. Examples of geographic phenomena represented by polygons are: homogeneous vegetation stand, ranger district, research natural area, tract of privately owned land.
Precision - A quality associated with the refinement of instruments, measurements and calculations, indicated by how closely grouped are a set of reported values of the same measurements, or how repeated observations conform to themselves. Contrast with accuracy.
Remote Sensing - The science and art of obtaining data and information about an object or representation of that object without coming into physical contact with the object. Usually refers to collection of information about the earth's surface using photographic or electromagnetic sensors aboard aircraft or satellites.
Rapid Application Development (RAD) - An approach to information systems development which relies heavily on interactions with the people who will be using the system: for example in workshops to establish needs and by iterative prototypes to demonstrate development direction.
Spatial - Of, or existing in space. Also refers to a map that contains the spatial distribution of natural resource information organized through the use of classifications into points, lines, or polygons.
Spatial Data - Data defining the geographic location of features. Maps and aerial photographs present a graphic portrayal of spatial data. Electronic spatial data has two forms, vector and raster.
Theme - The subject matter of a map or data layer containing information regarding related phenomena. For example, the theme hydrography might include river and other stream locations, lake and reservoir boundaries, and springs. Though the data types and map symbols may vary, they are considered to be within a common theme.