Harland Bartholomew and Associates, Inc.
Abstract
Implementation of a GIS for the purpose of collecting and analyzing data on individual trees within an urban forest is a daunting, time consuming task. This is especially true for communities like military installations which vary greatly in size. Often these types of communities require adherence to a specific type of database standard, while utilizing non proprietary software. The purpose of this paper is to: 1) Outline the process utilized to create an Urban Forestry Geographic Information System. 2) Review the challenges in gathering this type of information, and 3) Highlight the steps necessary for migrating this data into ArcCAD/ArcView.
Harland Bartholomew and Associates, Inc., using the Tri-Services Spatial Data Standards Version 1.4 as a guide, developed a PC-based ArcCAD/ArcView Urban Forestry GIS which serves as a model for communities of all sizes.
Urban forestry data was collected with either a Global Positioning System or Pen Computer Based Software. After post-processing, this data was migrated into an ArcCAD/ArcView based Geographic Information System. Planners, resource managers, designers, and maintenance personnel can then access, query, and print this information with ArcView, while updates can be performed using ArcCAD.
Field data was gathered using Trimble Navigation's Pathfinder Pro XL Global Positioning System or Penmap Pen Computer Data Collection Software. Using ArcCAD/ArcView coupled with Microsoft Excel and Access, a fully functional Geographic Information System was implemented.
Introduction
Trees are an important part of any urban environment principally because of the aesthetic and environmental benefits they provide to a community. Managers of urban forests spend a great deal of time budgeting, scheduling, and coordinating work on trees. With the advent of cheaper, more powerful personal computers, tools such as Geographic Information Systems (GIS) are becoming more widely available to aid those tasked with the management of urban forest assets. A Geographic Information System can be defined as a system used for the capture, storage, analysis, and display of spatial data referenced to the earth. This spatial information is in both graphic and non-graphic database form. The graphic portion is a drawing which uses lines, points, and polygons to represent geographic locations. The non-graphic attribute data is stored in database records that can be linked to the graphic elements. Several companies offer specialized software solutions which combine database software with digital mapping, however communities of smaller sizes often do not have the resources available for such a substantial investment. In the case of military installations, specialized software solutions do not lend themselves to compliance with the Tri-Service Spatial Data Standards. The Tri-Service Spatial Data Standards are a database design or schema which was developed from a combination of many schema used by all branches of the military service. A GIS solution utilizing ArcView/ArcCAD offers communities a GIS platform on which can be built a system best suited to their individual needs. In many cases, communities are already utilizing ArcView/ArcCAD for other GIS needs within the community. This means that staff with software experience may be on hand, reducing the need for specialized training. Data created within an ArcView/ArcCAD platform can also easily be migrated into other GIS systems, reducing the chance of system obsolescence.
The United States Air Force Center For Environmental Excellence (AFCEE) has embarked on a program of implementing an ArcView/ArcCAD urban forestry GIS solution for several of its installations. The Air Force envisions benefits for a system of this type to include compliance with the Tri-Service Spatial Data Standard 1.4, compatibility with data sets currently in place, or planned in the future, and ease of use.
The purpose of this paper is to outline the process for creating an urban forestry Geographic Information System, discuss the challenges of gathering urban forestry field data, and to explain the process of migrating field data into the GIS.
Requirements for Creating an Urban Forestry GIS
The process of creating an urban forestry GIS, whether for a military installation, large city, or small municipality is essentially the same. This process includes the examination of five criteria which center around the user of the system. These are: 1) Needs/Requirements of the end user 2) Hardware/Software Requirements 3) Requirements of the GIS 4) Database Design Requirements 5) System Maintenance/Updating Requirements.
The needs and requirements of the end user are usually the first criteria to be examined when creating an urban forestry GIS. It is important to outline and understand exactly what type of information the user of the system wants to put into the system and utilize on a regular basis. Often these needs can be determined by asking a few simple questions, which in turn determine exactly how the system is expected to operate. While not totally inclusive, some typical questions may include:
Many times the specific conditions of the urban forest play a role in shaping the GIS. For example if large numbers of one tree species are planted as wind breaks, alternate methods for capturing this information will be implemented into the GIS design.
Hardware and software requirements play an important role in the development of an urban forestry GIS. With continual technological advances in computer hardware, a computer with adequate resources to perform the basic operations required for an urban forestry GIS can be obtain quite reasonably. In may cases, existing computers within the office can be used. Often local area networks (LAN) are in place allowing many users to share data. Requirements of those who may be sharing the urban forestry data need to be considered as well. Naturally a system of reliable data backup is an absolute necessity. Below are some minimum software/hardware specifications for an urban forestry GIS. Please note these specifications will change as new releases of various software programs are developed.
Software requirements for an urban forestry GIS utilizing ArcCAD/ArcView include the
following
software being properly installed and configured.
System Requirements The minimum requirements of the above software must be met . These requirements include:
These requirements are the minimums described by the above software. For better performance the following guidelines may be more useful as a recommended system configuration:
Based upon input from the above criteria, the requirements of a GIS can be formulated. Many times the easiest way to begin this process is to formulate typical questions or queries which need to be answered. Working through a set of questions insures the most important information is readily available for the user to extract. A few typical queries which may be asked of an urban forestry GIS include:
While extracting information through querying the data is one requirement of a GIS, other
important
features include mapping, and analysis of data.
Issues related to mapping include determining what scale of map is most often needed, and what information should be shown in addition to the urban forestry data. Mapping is important because it can also be used to aid foresters in locating trees on which work is to be performed. Scales and areas of interest should be determined in order to accommodate the necessary requirements in the GIS.
Analysis of the information contained within the urban forestry GIS is an important factor to consider when creating an urban forestry GIS. Charting and graphing are two powerful tools which aid in the understanding of trends within an urban forest. Provisions should be made for these capabilities within the GIS. Software products such as Microsoft Excel or Microsoft Access have the capability to effectively tabulate and analyze trends in data. This is especially important to consider when creating a GIS system for the simple reason that often those who will be using this information may have a higher level of familiarity with software which is commonly found on the desktop, requiring less training than a specialized type of software.
The database design requirements for any GIS are often regarded as the most complicated aspect
of system implementation. Considerable planning is needed to bring to fruition a system which
adequately utilizes the necessary information, while allowing for future database expansion
without adversely effecting system performance. Simplicity is often a good rule to follow when designing
a database. Using a minimum number of tables helps reduce the amount of potential maintenance,
aids systems performance, and makes understanding the system much easier. At a minimum,
tables should be created which store attribute information about individual trees. If needed,
additional tables can be added, tailored to the specific needs of the user. Additionally, tables which store
information about tree maintenance, special study areas, or other relevant information can be added.
Sample Urban Forestry GIS Table Diagram
To simplify database tables and keep their size as small as possible, abbreviations are used to represent values within a data field. These abbreviations are known as Domains. For example, within a data field designated to record utility information, the abbreviation OE may be used to represent overhead electric lines. The use of domains helps insure that fields are consistently and efficiently populated.
Below is an example of typical attribute data gathered for individual trees at various United States Air Force Installations. These are provided with the intent of serving as an example of typical urban forestry data gathered in the field.
Tree Identification - Every tree recorded in the database was given a number so that a list could be created in which each tree with its associated attributes may be readily identified. This system also provided an accounting system for the identification of the total number of trees included in the survey.
Block Identification -The first one or two digits of building numbers generally corresponds to area identification on military installations by use of the facilities within that area. These blocks were designated on a map before data collection began, as a means of simplifying tree location and structuring the data base.
Road or Street Name - The name of the closest street was included for each tree. This association aided in the quick location of a specific tree or group of trees for ongoing inspection and maintenance.
Building Number - The real property building number or address of the nearest building was recorded for each tree. This information aided in locating individual trees, especially in areas where the block or street names did not provide enough specific information for locating the tree.
Plant Type - All trees were designated as either Deciduous or Conifer. These terms are defined below.
Tree Name - Trees are typically identified by the common name generally associated with the tree and by their botanical name, which is a Latin derivative specific to the species. Because variation is often found in common names, botanical names are an important identification and communication device. Surveys should include the recording of both the common and botanical name.
Plant Name - This was the common name for the tree.
Latin Name - This was the scientific or Latin name of the tree. This name is composed of two parts, with the first word being the genus name, followed by the species.
Cultivar - This was the specific name given to a cultivated variety of tree.
Growing Environment - Growing environment described the general spatial characteristics of the area in which the tree was growing. There were five growing environment designations included in the survey data:
Stem Diameter - The overall height and size of deciduous trees and conifers was recorded as a measurement of the diameter of the trunk taken at breast height (DBH). Breast height is generally accepted as being 4.5 feet above the ground. For the survey the DBH was recorded in one of nine size classifications:
Condition - The condition of each tree was recorded based on the eighth edition of Guide for Plant Appraisal by the International Society of Arboriculture 1992, Table 6-2. The five indicating factors used by this system to evaluate plant condition were: Roots, Trunk, Branches, Smaller Branches and Twigs and Foliage. Evaluation with this system was based on the percentage of these components observed to be alive and in a healthy condition. This evaluation was then ranked into one of the following categories: Excellent, Good, Fair, Poor, and Very Poor. The Excellent category was not used since root-collar inspections and/or climbing trees for evaluation was not performed. The Very Poor category was considered the same as Dead.
Maintenance - The maintenance requirements of each tree in the survey were addressed in this category. Every tree was assigned a maintenance priority using one of the following four basic designations:
Maintenance needs were considered in conjunction with the Risk Potential (see Risk Potential) assigned to a tree. A high risk potential, for example, may necessitate immediate or constant maintenance consideration.
Other Required Maintenance - This data category was captured to report any unique or specific maintenance requirements not included in the Maintenance category. Specific items checked in this category were:
Risk Potential - The risk potential is the potential for injury or damage to persons or property likely to occur from the tree because of its condition or location. This potential was assessed using a range of 0 to 5. This range is defined as:
Utilities Present - Existing utilities were noted when they conflicted with existing trees. Trees which conflicted with buildings were also noted in this category. These conflicts usually affected the risk potential and the maintenance requirements of the tree. Conflicts were recorded using the following abbreviations:
Comments - Any additional comments were captured during the field survey and put into the database. Examples of this included trees planted in memory of a person or events, trees of historical significance and trees of state champion size.
As mentioned earlier, the maintenance and updating of the information contained within the GIS is equally as important as the design of the system itself. Because an urban forest is a dynamic community which is always changing, maintenance and updating are of prime importance. Strategies for maintenance and updating should be addressed during the design of the GIS in order for these activities to occur in a systematic fashion. At a minimum, an urban forestry GIS should be updated one time every year. The best time to accomplish this may be during a regularly occurring slow cycle in urban forestry maintenance work. This will vary depending upon the location, size, and staff responsible for the urban forest. Typical methods for updating include using as-built documents for system input, windshield surveys, or survey by Global Positioning System.
Data Gathering Challenges
Once a database has been designed and developed into a workable solution, data must be gathered to populate the system. Several challenges become apparent as a system for collecting field data is developed. Existing base mapping, existing tree information, and field mapping strategies are the primary areas which should be investigated while developing a system for collecting urban forestry tree data. Often during the early stages of data gathering adjustments are made to the database. These adjustments are necessary because additional unexpected variables are encountered during data collection.
Existing base mapping is one of the most important issues to consider when beginning to collect and develop an urban forestry GIS. It is upon this information that the tree data will be mapped and used by staff in the office and out in the field. This information will often determine the method of data gathering. First it should be determined if a base map exists, and if it is available in paper or digital form. Many times only paper copies of base mapping exist. These can often be scanned and vectorized into a digital base map. If a vectorized base map is beyond the needs of the project, a scanned digital image of the line drawing can be used as a base map. Aerial photography can also be used by scanning the image and using the digital files as a backdrop for data gathered in the field. Digital files created from scanning either line drawings or photography, which do not contain vector geometry are referred to as raster images. These files contain dots or pixels which make up the image, much like the dots on a television screen. Raster images can often be used in place of digital vector mapping with success, if the size of the image file, or the number of image files needed to cover the study area is not unmanageable.
CAD files often represent the most common type of digital base mapping. Several issues related to CAD files are worth discussion in order to ease the transition of the field data into the GIS. The CAD file format is important to determine during the beginning stages of project development. Bentley's Microstation and Autodesk's AutoCAD are the two popular CAD packages on the market today, and can be used with a GIS based on ArcCAD/ArcView. However, many other CAD packages exist. Provisions for converting CAD files into a format which can be used by an ArcCAD/ArcView system should be made. The coordinate system and distance units of the CAD file are of primary importance if data is to be gathered with a GPS unit or other type of digital data gathering system. When transferring data from either type of data gathering system, the coordinate and measurement system will need to be used to export the information to the correct geographic position. If a base map of any type does not exist, data can still be gathered with the intent that it will be used with future mapping as it is acquired. Again, the coordinate system in which the data is collected should be noted to insure proper conversion into the future base mapping when available.
When gathering existing base map information, it is also important to investigate the availability of any existing tree information. Often CAD files may contain existing tree locations which may be imported into a pen based data gathering system, or used as a hard copy mark up in the field. Decisions about whether to use any existing tree information should be based on the accuracy and thoroughness of the data. Some situations may be better suited to using existing tree information supplemented with additional data gathered in the field.
Three primary systems which are often used for gathering urban forestry information are: Global Positioning Systems(GPS), Pen Computer Based System, and Traditional Paper Surveys. Each system has its advantages and disadvantages which will be briefly outlined below.
Global Positioning Systems represent the state of the art in field survey technology. The
Global Positioning System (GPS) is a worldwide radio-navigation system comprised of 24 space
satellites, which orbit the earth at an altitude of 10,900 nautical miles. A GPS uses these satellites as
reference points to calculate positions on the earth. Accuracy can vary from several meters to several
centimeters depending on which system is used. A GPS with sub-meter accuracy is typically used to
capture urban forestry data. The Trimble Home Page located on the world wide web at
http://www.trimble.com contains a detailed explanation about GPS and how it works. These types of
data collection systems are very efficient at collecting large amounts of attribute data quickly.
Personnel who utilize these systems require some amount of training to operate the equipment. Because of
the complex nature of the equipment, it is often helpful to have the GPS equipment set up and
operating before arrival at the survey site. This is also an excellent opportunity for those involved with the
data gathering to familiarize themselves with the system and its operating procedures. Since this
system is based upon the storage of attribute information within a logging device, it is often helpful
to
review the previous days work for possible errors and omissions. This can be accomplished
by printing out the previous days attribute information and mapping. Most GPS system software
provides accommodations for this type of activity. At a minimum, after all data has been collected
for the project, attribute and mapping information should be reviewed one final time before leaving
the site. Other potential problems involved with gathering data with a GPS include data loss due
to equipment failure, and loss of accuracy due to improper equipment setup.
Trimble Pro XL Data Capture System
Pen computer based data collection systems are comprised of pen type computers or laptop
computers, which are taken into the field for data collection. These systems contain copies of digital
base mapping which can be used for inputting attribute information. Pen computer based systems
work best when existing digital files are provided which include tree positions, thus requiring the input
of tree attributes only. Again this type of system will require some amount of software training
for those involved with field work. A pen computer based survey system allows users to check
attribution at any time during the data collection process. This reduces the need for daily checks on the
data collected in the field. When using data collection systems such as pen computers or laptops,
data backup is extremely important, and should be planned into the data gathering activities.
Penmap Pen Computer Based Data Capture System
Traditional paper based surveys involve using paper maps and recording information about trees on paper. This represents the oldest method of data collection. Its accuracy depends entirely on that of the base map's but it does provide a permanent record of the field work which can be referred to if needed. This method of data capture can often be very time efficient if conditions in the field are right. For example, an urban forest with a small amount of species and size diversity, mapped with existing paper maps, might need a few attributes which can easily be gathered on paper maps or forms
Regardless of which type of data collection system is used, it is very important to insure all people involved with the capture of field data understand the procedures, definitions, and practices of data collection. The data capture form, whether paper or on a data collection unit, should be reviewed in detail before data capture begins. All definitions, abbreviations, and other standards should be agreed to and understood. If possible these procedures should be documented for future reference so the results of the survey can be better understood back in the office.
Data Migration Process
Once the capture of field data is complete, the process of migrating the urban forestry data to the GIS begins. In order for the GIS to be meaningful to the users, it is necessary to insure that the data gathered is of the highest quality possible, consistent with the database design, and error free. Post processing efforts to insure this occurs include, data analysis, data quality assurance, and data standardization. Once these steps have been completed, the final data can be moved into the GIS and the system can be tested.
The data migration process begins by moving the field data from the form in which it was gathered in the field, into some type of software which allows further processing. Spreadsheet software is best suited to post processing activities, of which Microsoft Excel and Lotus 1-2-3 are two examples. Once the field data is in a spreadsheet type of program the data can be analyzed. This analysis might include the charting and graphing of various attributes collected in the field. The primary goal of analyzing the field data is to insure that the intent of the survey was accomplished, and that the conditions which exist in the field will be accurately represented by the data.
Most importantly, the field data should be checked for errors and omissions. In order for a system of the highest quality to be delivered, the data must be thoroughly checked for errors before the GIS is created. Errors in field data can be broadly classified as being of one of two types: spatial errors (Trees not located in the correct geographic location), and attribution errors (Information about the trees is missing or incorrect).
Spatial errors can be checked using an ArcView project containing existing base mapping. The field data can be added as a table within the ArcView project once it is exported from the spreadsheet program. Using the "Add Event Theme" command within ArcView, the field data can be overlaid with the existing mapping and checked for spatial errors. Inconsistencies with coordinate systems will also be apparent during this time.
Attribute errors can be discovered by using sort, search, and subtotal commands from within the spreadsheet program. Common attribution errors may include missing information, inconsistent abbreviations (domains), misspellings, and wrong Latin or common plant names.
After all corrections have been made to the data, it is ready to move into the GIS. If a system
using just ArcView will be implemented, the final corrected data should added to the ArcView project
file as discussed above. With an ArcCAD system, the corrected data should again be added to
the ArcView project file, then exported as a shape file which can be utilized by ArcCAD. Once
the necessary coverages have been made using ArcCAD, other tables can be added to the system. At
this point the original queries can be run on the system to insure that it is functioning as intended.
Any fine tuning may be accomplished at this time to correct unforeseen problems. After
successful system operation, the data should be backed up in the event of hardware/software problems.
CD-ROMs can be created for a reasonable cost, and are a permanent method of storing project data.
Completed GIS Project Using ArcView
Conclusions
Because of the important role trees play in the urban environment, the maintenance of the urban forest is becoming an issue demanding greater attention. Tools such as a Geographic Information System are helping the urban forester better manage the assets under his care. With proper planning a GIS can be implemented which is tailored to serve the needs of the individual community while still being user friendly and affordable. With a better understanding of the requirements for an urban forest GIS, communities and their urban foresters can have the ability to implement better, more cost effective maintenance programs which will benefit future generations.
References
Urban Forestry Management Plan for Vandenburg Air Force Base, Harland Bartholomew & Associates, Inc., St. Louis, MO., June 4, 1996
Guide for Plant Appraisal, International Society of Arboriculture, Eighth Edition, Savoy IL., 1992
GPS: A Guide to the Next Utility, Jeff Hurn, Trimble Navigation, Sunnyvale, CA., 1989
L. Andrew Franke Landscape Architect Certified Arborist Harland Bartholomew & Associates, Inc. 400 Woods Mill Road South Suite 330 Chesterfield, Missouri 63017-3427 Tel: (314) 434-2900 Fax: (314) 576-2702 E-Mail: andrew_franke@parsons.com