Used with permission from the Maine Office of Information Technology
In the last ten years, there has been great interest in creating highly-accurate ortho-rectified aerial photos for use in geographic information systems (GIS) in Maine, not only at the federal and state levels, but also at the municipal level. Many towns have developed such data on their own, but lack both the financial and technological resources to share these data, nor are they able to take advantage of the large amounts of data collected by state and federal agencies. As a result, several terabytes of digital aerial photos exist in the state, but the means of sharing them does not. Considering the millions of dollars spent in developing such data, it makes sense to invest a small amount more to publish the data and help it reach its true potential.
The Maine Office of GIS (MEGIS), in cooperation with the Maine Library of Geographic Information (GeoLibrary), has developed an open web mapping service (WMS) platform specifically to meet this need. MEGIS has developed a "production pipeline" comprised of scripts and open-source software that greatly decreases the time it takes to convert imagery to a WMS. The platform provides the service using OpenGIS Consortium (OGC) standards, and relies almost exclusively on open-source software such as the Geospatial Data Abstraction Library (GDAL), Python, and MapServer. The process consists of two parts - preparation of the data to create the WMS, and then a web serving platform to host it.
The purpose of this project is to partner with any organization holding publicly-available digital aerial ortho-photos for Maine in order to allow these data to be made available to any user via WMS. Such a service can easily be consumed in most GIS software including Google Earth, and can also be easily integrated into any web-mapping application. Maine's approach to this solution has spanned all levels of government in Maine, including hosting data from the federal government U.S. Geological Survey (USGS) and U.S. Department of Agriculture (USDA), state government (MEGIS and Maine Department of Marine Resources), and local government (Southern Maine Regional Planning Commission, the Greater Portland Council of Governments, and the towns of Augusta, Manchester, York, Kittery, and Hampden). This service is still early in its development and we foresee doubling the data holdings in the next six months. Collaboration was key to this service's development, MEGIS collaborated with USGS, USDA, the University of Southern Maine, and Maine GIS stakeholders via the Maine GeoLibrary. The GeoLibrary Board provided the funds to purchase the server hardware, while MEGIS provides the expertise to convert raw imagery into WMS. The other partner organizations provide their raw data in return for having free access to it via WMS.
While WMS and providing image services have been around for a few years, this project is unique in that it is the first attempt, as far as we are able to ascertain, to provide a free and openly-available platform for sharing aerial ortho-photos. Indeed, even big federal agencies such as USGS and USDA do not have platforms available to provide all of their Maine imagery via web services. The benefits of these services will be tremendous. State agencies and federal agencies will be able to tap into any imagery product available in Maine, and small municipalities will be able to provide their data outside their town offices at no expense. This solution also provides several opportunities to protect and conserve natural resources. For example, the
Maine Department of Environmental Protection (MEDEP) uses high-resolution aerial photos to determine environmental compliance. The ability to access municipal imagery allows them to see the state of projects on a specific date, and such data is admissible as evidence in court, if a matter advances to that level. MEDEP also uses high-resolution data to assess the health of urban watersheds. Other natural resource agencies use these data for habitat conservation and land use planning, to conserve natural resources. Another example is that having these data available allows government workers to assess many areas without having to make a gasoline-consuming trip out to the site. Additionally, fostering this collaboration brings all the stakeholders together for better future planning of aerial photo projects, and will reduce the number of plane flights required to capture the same data. Finally, another innovation in this project is the wide use of open technologies. The web service uses the WMS protocol, an OGC approved open GIS standard. The images are processed using mostly the open-source GDAL package, and many of them are compressed using the open JPEG-2000 compression algorithm. Scripts to process the imagery are written in the open-source Python language and typically auto-generated. Finally, the WMS itself is served out via the open-source MapServer software from the University of Minnesota. This allows us to share the methodology and process with any other organization that would like to duplicate it. It eliminates the need for expensive commercial software.
This project, which began with its first offering in October 2008, increases government transparency and citizen engagement because it puts in the hands of the public the same data that are being used to make governmental decisions in a GIS context. By utilizing an open standard for the web service, the project ensures that any citizen with access to Google Earth or any other GIS package (such as free and open-source QGIS, MapWindow, or gvSIG) can see the same data.
The greatest immediate measurable results are the usage of the service so far. Although it is still in its infancy, in terms of the amount of data available and the size of the audience aware of the service (which to date has only been advertised on the GeoLibrary web site and via email), the service is already seeing 2,000 - 3,000 hits per day. Although we have posted only two municipal products so far (Augusta area and York County data), we have been inundated with requests by other municipalities and already have a long backlog of data to prepare. Another immediately measurable result is significant cost savings. Earlier attempts at providing web imagery services were costing the state $110,000 per year, prohibiting the inclusion of municipal data. Using WMS and open-source software, that cost is slashed to $6,000 per year. Users will see increased efficiencies by being able to utilize a number of different imagery products for their particular geographic area, without having to hunt down the source and then acquiring the data. Finally, by using open-source software we can easily transfer the process to any other organization that wants to use it without incurring expensive licensing costs.
In order to ensure a high-quality service and performance levels, we followed best practices for developing the initial offering. The service is available from two servers, with a third for developing new services. The production servers are located in different buildings and utilize different network links, thus providing high availability in the event one building was lost. The solution is easily scalable by simply adding more servers. Finally, by using lightweight software we are able to maximize the performance of the service, rather than using "heavy" commercial software which requires greater hardware resource for the same results. One result of early development of this service has been to test the methodology and learn from some mistakes. These lessons have included using the lightest software needed to do the job (we initially tested MapServer against two other commercial competitors), using JPEG-2000 compression whenever possible instead of proprietary compression, and working directly with towns to get data.
Go online to try out the service.