The Buffer GIS project was initiated by the Lake Erie Buffer Team to investigate
the ... the template that each county populated with their CREP information.
LAKE ERIE BUFFER PROGRAM
GIS DEVELOPMENT
OHIO LAKE ERIE BUFFER TEAM Prepared By:
Ben Marckel Gina Menuez Megan Sommers Elaine Moebius Steve Davis Ed McConoughey
June 2003
1
Table of Contents
Background………………………………………………………………………3
GIS Project Overview………………………………………………..………….4
Procedures…………………………………………………………………..…...5
Time Requirements……………………………………………………………...7
Observations, Barriers, Successes and Recommendations…………………...7
Technical Issues………………………………………………………………....9
Conclusions…………………………………………………………………….11
Appendix A………………………………………………………………….....13
Appendix B………………………………………………………………….…17
2
I. Background
The Buffer GIS project was initiated by the Lake Erie Buffer Team to investigate the applicability of GIS technology to quantify and display conservation buffer information. The buffer team was interested in the feasibility of developing a watershed wide GIS based buffer data layer.
The goals of this project were to: •
Determine feasibility and practicality of developing county GIS data bases
•
Determine potential for merging county data basis on a watershed scale
•
Define data elements needed for a common county buffer data layer
•
Train field office staffs in use of GIS technology and the procedure to create a buffer data layer
•
Evaluate different methods (field GPS data collection versus map digitizing) of collecting the data needed to populate the data bases and develop the data layers
•
Determine usefulness of the data layers created for use in watershed modeling or targeting projects
•
Determine usefulness of data in visually displaying accomplishments and tracking buffer installation progress
•
Determine time requirements to implement such a project, and collect, manage, and manipulate the data.
This project was funded using Lake Erie Buffer team operating funds. Additional funding was received for this project via a grant from the Great Lakes Commission. USDA NRCS and local soil and water conservation districts provided office space for the employees hired to carry out the project. NRCS provided computer workstations and vehicles for the interns at no cost to the project. The NRCS state GIS specialist provided training and technical support to the four project employees hired, while NRCS District Conservationists provided the day-to-day supervision of the four project interns.
3
II. GIS Project Overview
The project began in early spring of 2002. NRCS agreed to hire the GIS employees (Summer GIS Interns) as temporary employees using the agency’s established personnel system. A vacancy announcement was distributed and personal calls were made to colleges with established GIS degree programs.
One of the first obstacles the project encountered was the hiring of interested, qualified, and employable recruits. It soon became apparent that even though the project offered many opportunities, few individuals were interested in applying for the positions. In order to find sufficient candidates contacts were made with over 20 colleges in 7 different states.
An additional obstacle proved to be attracting candidates that were employable. At least four highly qualified graduate students applied, and each would have made an outstanding candidate. All were pursuing a highly specialized technical program in GIS and remote sensing, with an emphasis on conservation and resource management. However, each was a foreign student, studying in the U.S on a student visa. Agency regulations of the federal government prohibit employing non-citizens and in the aftermath of Sept 11, none of the non-federal partners were comfortable in putting these applicants on their payroll.
Eventually after much hard work, four well-qualified individuals were hired. One intern had recently graduated with a masters degree in remote sensing, the second with a bachelors degree in environmental sciences, and the third had just graduated with a bachelors degree in GIS information sciences. The final intern hired was a third year college student majoring in communication and computer technology. The interns started at various times in late May and early June.
The interns were provided training in GIS techniques and in the conservation buffer incentive programs. Planning meetings were held with the District Conservationists. The group was given the charge of digitizing one complete year’s buffer practices in each of the four pilot counties and entering it into a data base that could be merged and manipulated as a whole. The District Conservationists and interns designed the buffer elements to be recorded and tracked.
4
The interns designed the GIS database, a procedure for populating the data and for merging and manipulating the data. The procedure used varied from county to county based on the condition of the field office record files, map and staff resources available in the field office. Each intern recorded his or her methodology. It was soon determined that there were numerous ways to get from point a to point b.
During the initial Lake Erie CREP Project meeting, the team developed a standard list of essential data to be reported for each conservation buffer practice. This data was later put into a data layer with the help of Rick Fout, GIS Specialist of the Deleware Soil and Water Conservation District. This data layer acted as the template that each county populated with their CREP information. Each county created their individual data layer using this template. This procedure was used to ensure that the information would be standard and universal among the four counties. The importance of identical project construction among all four counties was emphasized, the ultimate goal being to create an easily blended final project.
III. Procedures Developed
A. Standard Data Layer Development The project utilized ArcView software (Version 3.2a) to track Fiscal year 2001 buffer practices that were installed under the Conservation Reserve Enhancement Program (CREP)
Buffer practices digitized included:
1. Filter Strips (CP21) 2. Riparian Buffers (CP22) 3. Field Windbreaks (CP5A) 4. Wetland Restoration (CP23)
The following data elements were tracked in the database developed:
1. FSA Farm Number
7. Section
2. FSA Tract Number
8. Latitude
3. Total Contract Acres
9. Longitude
4. FSA Contract Acres
10. Parcel ID Number
5. County
11. Owner First Name
6. Township
12. Owner Last Name 5
13. Owner Address
21. CREP Contract Length
14. Owner City
22. Total Contract Length
15. Owner State
23. Program Type
16. Owner Zip
24. Practice Code
17. Contract Start Date
25. Length of Protected Stream (ft)
18. Contract End Date
26. Average Width (ft)
19. Contract Years
27. Wetland Acreage (If Applicable)
20. CREP End
28. Buffer Acreage (If Applicable)
¾ Data Elements Included & Tracked: 1. Contract Fields 2. Conservation Practice Code 3. Field Acreage 4. Field Identification Number
Additional Data Layer Consideration & Information:
In addition to the orthophoto mosaics (aerial imagery), additional data layer themes were obtained from various sources, to assist in the data layer creation. Such sources included local and state government agencies, the Ohio Department of Natural Resources Web site and the University of Toledo Web site. Examples of specific data layers (shape files) include: 1. Roads, Highways & Streets 2. Hydrology (Streams, Rivers, Main Drainage Channels, etc.) 3. Soils Profile 4. Parcel Identification Numbers 5. Acreage Calculation Script 6. ArcView Garmin Script
The interns developed step by step procedures to collect and populate the databases. The procedures varied based on the record situation from office to office and on the amount of data that had been developed locally before the interns started. Each of these procedures are documented in the Appendix.
6
IV. Time Requirements to Collect & Populate Data The Buffer Team project consisted of approximately 432 total work hours per intern, over a 12-week period. Data collection and hard-copy organization combined for the initial 72-hours of work for the project. Digitizing and table creation comprised the majority of project hours, equaling approximately 267 hours. Modifications and corrections, the final stage of the project, totaled 60 hours. Approximately 43 hours were spent in NRCS training sessions and meetings.
Over the course of the Buffer Team project, approximately 40% of the 420 hours were utilized for completing the main task. This percentage equals approximately 171 hours. If project standards/procedures were set prior to the start of the project less time would be required. The development and establishment of procedures and trouble-shooting required many hours of attention. Approximately 43 hours were spent in NRCS training sessions/meetings and compiling the project report. However, the amount of time spent on the project will be affected primarily on the number of records needed to be created and populated. This number varied among counties participating in the project.
Access to previously created shapefiles may have saved many hours of digitizing, even though many of the polygons required editing. Some time was spent working on other overages for practices not included in the project such as CRP grasses, CRP Trees, and EQIP. A considerable amount of time was spent on editing polygons and populating data tables.
V. Observations, Barriers, Successes and Recommendations
Operations Issues
1. Use of orthophoto coverage was an issue. USDA supplies UTM coordinate photos. However many SWCD's have access to local state plane coordinate photo’s which are newer, sharper and a higher quality aerial photo. Field offices preferred the local photos for better quality, however not all offices had access to that format. Thus a conversion step was needed to merge data from the two different references, which added a time consuming and
7
complicated step to the process of using multi county merged data. It is important that a consensus be established in terms of ortho-photo coverage use so that projections are similar and do not need to be re-projected for a successful merger of data.
2. Collecting the data needed to populate the database is labor intensive and time consuming. It took each intern most of the summer to go back and do the data collection for one county. It would not be feasible or cost effective to go back and post collect this data on a large scale. The efficiency in utilizing the data will have to come from automating the GIS effort as part of the conservation planning process from start to finish.
3. There was significant spin-off from the project in terms of generating interest in the field office staff in the capabilities and utilities of the GIS tools. The interns helped train the remaining field office staff to become familiar with their use.
4. The most valuable use of the GIS technology was to produce accurate maps and drawings for conservation plans, payment measurements and conservation practice application. Use of the technology greatly enhances the quality of conservation technical assistance and was probably more valuable than the data bases created.
5. Most field offices are short of staff time and overwhelmed with programs and paperwork. Often they are faced with the choice between getting the conservation on the land and keeping the files and paperwork up to date and tidy. As a result, records are in various degrees of completeness and organization, which would complicate the effort to apply this process across all counties in the watershed. Field offices would not have the staff time to assume this additional workload without additional staff resources. Unfortunately, when the interns left so did the GIS data collection efforts in most (not all) of the pilot counties.
6. This effort will only be feasible on a large scale when USDA moves to a complete electronic and digital based system of technical assistance. When and if farm data (farm and tract numbers, field and farm boundaries, soils maps) is completely digitized, and an efficient service center system of accessing and manipulating the conservation planning data provided, then the tools will be in place to more effectively utilize this technology.
8
7. Effective large-scale application of this technology will require automating these processes into day-to-day conservation planning tools. Doing this will require more IRM support and staffing than is currently available to field offices.
8. Utility of the data will depend on the desired use. It was found that the databases were not very useful to produce maps at the county or beyond level. That is because the buffer practices occupy such small slivers on the landscape; they showed up as pencil lines on the maps at this scale. Usefulness for watershed modeling purposes would be more valuable but depends on having a complete database populated for the area modeled.
One challenge with modeling work is to differentiate between buffers that are newly installed through conservation programs and those, which were developed through the years with out program benefits. So far no remote sensing technique is available that is effective in picking up these small areas.
Technical Issues
1. Collection of tabular data can vary as long as the attribute template is standardized. All file information may be organized prior to attribute table population or one may choose to simply work through contract files and enter data from these files.
2. Customer Toolkit is a program with amazing capabilities, however the interns were not fluent in toolkit procedures to a degree that would have been a benefit to the project.
3. File management is very important. A lack of organization with files resulted in confusion at times. Make sure to manage files in a systematic way to reduce time spent finding files. If altering previously created files for the project make sure to make copies of these files so that the original file is not altered.
4.
It is helpful to keep a written record of the files and the practices they represent. The “Manage Data Sources” option under the “File” menu is helpful to rename files. This operation will allow you to rename all files associated with a shape file (i.e. .dbf file, .shx file). Click on the name of the shapefile to be renamed. Click the “rename” button and type in the new name and pick a directory where the file should be stored. TIP: Make sure that the
9
file you wish to rename is active in the project. ArcView will not allow you to perform the “rename” function if it is active in the view or table.
5. Make sure map units are set accordingly in the “View Properties” so that measurements can be done. Example: To work within the UTM projection the map units were set to “meters” and the distance units were set to “feet”. This is done so that measurement operations can be executed.
6. When merging layers make sure to check for overlapping polygons. If you merge two layers with overlapping polygons, errors will occur in the attribute table.
7. If joining tables you must save the “join” as a new shapefile to make it permanent.
8. If no overlaps occur between polygons a “Union” can be initiated.
9.
To save time populating fields with a common input feature such as county name, a calculation can be executed to populate this cell for all records. Begin editing the theme table; highlight the field you wish to perform a calculation in. Select “Field”, then “Calculate”. Type in the string or number you wish to have calculated into the field and select “OK”. (Note: If you are entering in a string, make sure to place quotation marks around the string before you select “OK” This is true even if your string includes a combination of numerical and alphabetical characters. If you do not include quotation marks your calculation will not be processed. You do not need to use quotation marks when entering in information for a numbered field.)
10. As an example, a landowner with five buffer practices would have five individual records to report each separate practice. This ensures the differences in acreage, practice types, location and contract dates are captured for the larger analysis of the buffer project.
11. The database table created in ArcView can be converted to Microsoft Access. This enables personnel on an office computer network to work on the database as a shared file. In this case, any employee can access and update records as new CREP cases develop. By the continual update of the database through a networked system, the link between it and the ArcView project is guaranteed to keep the ArcView project up-to-date.
10
CONCLUSIONS
1. It was surprising how difficult it was to find GIS trained students interested in the intern positions.
2. The interns were successful in creating a county buffer GIS data layer that could be merged across county boundaries. The process is documented in the appendix in this report for others to repeat. One challenge that had to be resolved was that each county, as a result of local programs or ways of doing things, wanted to maintain slightly different data in the “common” database that was developed.
3. The GIS data layer turned out to be a limited utility in making maps at the county or watershed scale… buffer practices themselves make up such a narrow footprint on the landscape they show up as hairlines on a watershed or county scale map.
4. The GIS database would have utility in integrating data into a watershed model. However, a major unresolved challenge is how to account for existing/natural buffers on the landscape. These cannot at the present time be detected by available remote sensing techniques and field-collecting data on the natural areas is tremendously time consuming. Thus the databases which can be developed are limited to just buffer areas installed under programs. In addition, application of a database in a larger watershed would require all entities to maintain such database and keep it up to date. 5. Field collecting the data involved a tremendous time commitment. It is beyond the current staffing capability of the field offices to maintain this data. When the interns left the field offices, in most cases (not all) the database was no longer maintained. The major promise in use of this technology will be realized when and only when USDA goes totally digital and integrates its county FSA maps, soil maps, landowner, parcel and tract information into one common system. When that happens then the entire process could be automated to collect this information. Automating that process is beyond the current IRM capability of expertise that exists at the field office level.
11
6. One of the major successes of the project was to use the GIS technology to make accurate maps for landowners in the conservation planning process. Practice location, acreage measurement, and visual plan images were greatly improved using this technology. The presence of the interns in the offices provided the impetus to move existing staffs ahead in this endeavor. The interns were very valuable in providing training to field office staffs in the use and application of the technology. As a result we are seeing much more use of the technology than we would have without the project. This in itself probably provided a greater benefit than the actual database that was developed.
12
APPENDIX A Step by Step Procedures Developed to Using Digitized Map Data
Determining the data necessary to represent the various conservation practices is the first stage of the project. Since the needs of each specific project differed based on individual goals and resources and information available, the following different methodologies were exercised by the participating counties during the Lake Erie Buffer Team GIS initiative.
Pre-Digitizing Methodology A (Utilized in Wood and Defiance County)
Having no prior conservation practices digitized into the ArcView data layer at the start of the project, the initial goal in these counties focused on obtaining necessary data elements from office files within the Soil and Water Conservation office and Farm Service Agency. In order to provide an organized layout of all tabular data, the specific conservation practice information was applied into hard copy spreadsheets that represented the database table in the data layer. The spreadsheets were organized in a 1” binder with each practice listed according to the last name of the producer/landowner.
Pre-Digitizing Methodology B (Utilized in Huron County)
Mike Patterson, the District Conservationist for the Huron County NRCS had created several ArcView conservation coverages prior to the start of the Buffer team project. The creation of a standardized attribute table template for all layers was a main concern along with finding the corresponding file for each polygon in the existing coverages.
Data collection consisted of searching and pulling all files for a specific practice. Using the aerial photography from the conservation plan and a plat map to decipher roads, the corresponding polygon was found on the screen, selected and the corresponding attribute record was edited. On most occasions polygons needed editing due to inaccurate acreage amounts. This was done by using the vertex-editing tool. Some splitting of polygons was also done using the Polygon Split tool. This process separated a single polygon into two polygons, each with their own record in the attribute table.
13
Pre-Digitizing Methodology C (Utilized in Fulton County)
The GIS project was built using the MrSID 1997 orthophoto image provided by the Fulton County Regional Planning office in the State Plane 1983 projection. The 2000 centerline data was overlaid on the orthophoto for location reference using the County, State and U.S. roadway system. The parcel data layer, also supplied by the Regional Planning office was referenced for land parcel numbers for the data template. The GIS was built using these three data layers in the State Plane NAD83 projection. The data management file system was created utilizing the guidelines established by the USDA Service Center manual, Manual for Managing Geospatial Datasets in Service Centers, June 2002.
The 2001 CREP files were located and the CREP data was collected from each physical file and the data immediately entered into an EXCEL spreadsheet. The EXCEL file was converted to Microsoft Access database file, because Access has greater flexibility and utility in data query. The database will continue to grow with the addition of new CREP customers entering into the conservation program. The entire spreadsheet was later converted to a dbase IV file and added as a table in ArcView.
Universal Digitizing Procedure To create new polygons, the shapefile was made active by clicking on the appropriate theme. Under the “Theme” menu, “Start Editing” was selected. The polygon tool was used to create a new polygon, edits were saved and the “Stop Editing” command was selected. A polygon was created for each farm field included in the conservation plan contract. Excess acreage may exist between two fields and may skew total acreage amounts for a polygon (Ex. a wooded area between fields not included in the contract acres). Fields were given separate records; therefore one contract may have several corresponding records. This was done to preserve the accuracy of each field's acreage that contributed to the total contract acres. A script found in the Toolkit Customer program calculated acreage. This script had been installed prior to the beginning of the project. A “CalcAc” field was automatically generated and each polygon had an acreage measurement calculated in the appropriate cell for this field.
14
All tabular information was saved in a database file (.dbf) format. New data fields were added according to criteria set in the conservation template. All edits were saved. Data was entered in for all files for recent contract years. Any missing information not included in the customer file folder was acquired from records in the FSA office. The layer for riparian buffers (CP-22) also included windbreak (CP-5A) data because only one CP-5A practice had been contracted for 2001.
Polygon lengths can be found by using the measurement tool. Average widths of conservation practices are calculated by multiplying the calculated acres of a polygon by 43,560 (ft), then dividing by the length of the polygon (ft). Stream length was found by measuring the length of the stream that flowed along the side of the polygon. The measuring tool was used and the length was recorded in feet. Some table management was necessary to make sure the table was clear of errors or spaces. Table editing included performing queries to move information from one field to another.
Buffer acreage was also left unpopulated and wetland acreage was entered in only for wetland records. The wetland records also contained no length of stream information because an accurate length measurement could not be estimated from contract and/or ArcView file information.
Additional Notes: For Methodology B, township and section information were derived from the county plat map directories. No parcel ID’s were included in the tabular data, some NRCS offices do not have access to ArcView coverage containing this information.
For each shapefile a query was then performed to select 2001 contract records and a new shapefile was created from the selected records. This procedure was done for each conservation practice. The three themes were grouped together to create a single conservation coverage for the 2001 Fiscal Year. The final step in creating the 2001 conservation practice coverage was to convert the shapefile from UTM to State Plane coordinates. This was done using the Projection Utility Wizard in ArcView, found under the “File” menu. The “current projection” could then be set by scrolling down the drop-down menu to “NAD_1983_UTM_Zone_17N” and the units were set to “Meter”. When asked if the coordinate information system should be saved,
15
“Yes” was selected. The new project was then set to “NAD_1983_Ohio_North” and the units to “Foot_US”. A directory and file name was chosen. After the Summary page appeared, pressing “Finish” and waiting for the Utility Wizard to finish processing completed the reprojection.
16
APPENDIX B Step by Step Procedures Developed to Using GPS Collected Data
Methodology I (Huron and Wood Counties)
Each of the four field offices was given a GarminMAP76 unit and receiver for the collection of field data. A one-day training session was held at the Wood County NRCS office in Bowling Green, Ohio. Robert Parkinson, GIS Specialist for the NRCS State office, conducted the session and instructed participants on how to set-up the GPS unit for data collection, step-by-step data collection procedures, and how to download data from the GPS unit into ArcView using the DNR Garmin extension.
Setting up the unit according to the projection requirements of orthophoto projections is important when overlaying information accurately. Huron County uses a digital orthophoto mosaic in UTM coordinates so the GPS unit was set to collect data in UTM. When creating a track or waypoint out in the field it is important to save all information in a systematic manner that is easily recalled. Files can be renamed alphabetically and/or numerically by using the rocker button.
Saving GPS information and transferring it into ArcView with the Garmin Extension is a relatively simple process. The GarminMAP76 was turned on and the set-up was changed to “Simulator” mode. The Interface tab in the Set-up menu was changed from “RTCM In/NMEA Out” to “Garmin”. The GPS unit was connected to the computer using the appropriate cord. It is important to make sure that the GPS unit is set up appropriately and DNR Garmin extension is loaded into ArcView. The “View” menu was highlighted and “Properties” selected. Map units were changed to “meters” and the distance units to “feet”.
To access the Garmin extension, “DNR Garmin” was selected from the Menu bar and the “Open Garmin GPS” chosen. The “Track” or “Waypoint” menu was then selected. When downloading tracks, for example, from the GPS unit all tracks in the log will be downloaded unless specified using the “Advanced” option. “Download” was selected and files were transferred form the GPS into ArcView. Under the File menu the “Save As” command was
17
selected and “GPS Text File” highlighted. A directory and a file name were entered and saved. To save the data as a shapefile, the “File” menu was selected, then “Load” and “Waypoints (or) Tracks from”. The text file created in the previous step was chosen. Then “File”, “Save As”, “ArcView Shapefile” was highlighted. The shapefile was then saved as a point, line or polygon. A directory and name for the shapefile was saved. The new shapefile appeared in the view legend. The checkbox was selected to make the theme active. New shapefiles can be created by selecting desired points or tracks and highlighting the “Theme” menu. Select the “Create New Shapefile” command. Name and save the file. When the “Add shapefile theme to view?” window appears, select “Yes”. Your new shapefile will be added to the view legend. Select the checkbox to make the theme active.
Work with the tabular data according to your project needs. When on the View page select the shapefile you would like to edit. Press the “Open Theme Table” button to access the attribute table. Clean up records and fields by editing and make sure to save all edits.
Methodology II (Fulton County)
After all CREP data was entered, the database was queried on a per township basis to determine the largest contingency of buffer practices in a per township location. The township data was alphabetized per owner to locate the physical files within the FSA and NRCS office files. With this determination, work was begun in German Township, Fulton County. The focus began in this township and progresses on a per township basis, determined by number of buffer practices.
After all the data has been collected and properly arranged in the file system, work can then begin constructing the conservation buffer polygons of the GIS project. The initial base map layers of the MrSID, Fulton County orthophoto, road coverage and parcel data layers were opened to begin the project, and all layers were projected in State Plane North American Datum 1983.
The physical files for the target township were placed within easy reach and work began in the ArcView GIS on a per township plan. The location field was determined by using the photocopy of the orthoquad in the FSA file. The address could not be depended upon for
18
location, as many times the owner, operator and field location address can all be different. To find the specific orthoquad for the conservation buffer site, reference to the Districts County aerial photography was useful in quickly locating the site on the MrSID ortholayer in ArcView. Most, if not all county offices have this basic reference system of aerial flight patterns flown from west to east over the county, working from the north to the southern county lines for aerial coverage of the entire county. This procedure needs to be done on the buffers already in place and planned at this time. In the future, data collection will be accomplished by using Global Positioning System (GPS). Fulton County personnel are now using the GPS as a tool to record the new conservation buffer sites.
The street data layer was valuable in correlating the county roads, state and federal highway systems with MrSID for visual buffer site location according to the FSA files. The total acreage and the actual field measurements were recorded on the FSA ortho file copy.
B. Advantages & Disadvantages of Digitizing & GPS Collection One main advantage for using a digitizing method rather than a GPS collection method is the ability to remain in the office and not spend time traveling to a site to apply GPS procedures. However, when depending on digitized data alone, it is probable that accuracy will be jeopardized. The GPS unit allows one to go out and actually view the lay of the land and document any variations that exist between the mosaic and the earth’s surface. The GPS unit uses coordinate readings from satellites to determine positioning. When GPS data is uploaded into ArcView, latitude and longitude information is stored in the text and tabular data. This is an added benefit of using GPS data verses digitizing. On the other hand, one disadvantage of using the GPS unit is the possibility for technical failure and more loss of time attempting to remedy a malfunction.
Using the digitizing method without a field visit could result in oversight of the land changes from outdated orthophotos, or from early original CREP files. When creating new digitized work from old land files, the potential for error is present. This can result due to changes in landscape, such as new building construction or pond installment. Changes to the landscape may not alter the established buffer strip area. However, impact to the land and water surfaces resulting from such changes could effect placement considerations for future buffer practice considerations.
19
