As users of this nation’s public or private trails, Sharetrails.org/BRC members likely rely on a variety of mapping technologies for navigation. Being able to visualize where you are, and where you’re going is important. Maps help us gauge distance and direction, and determine elevation or ground conditions. Personally, I feel that having a good map when venturing outdoors is as important as taking water and a first aid kit. I always carry a GPS, but having a paper map as a backup provides another level of comfort during an outing. It’s familiar, provides a tactile experience, and its fun to know where the heck you are on that piece of paper!
In the past couple decades, the transition from paper to digital maps has been accelerated. We can now carry more “maps” on our GPS or cell phone than would fit in a 100 tightly stuffed backpacks 20 years ago! Additionally, digital maps are dynamic, in that we can change the base from aerial images to a topography, layers can be added, turned on and turned off, and routing between destinations can be accomplished in a couple clicks or swipes on your device. These mapping tools and their information empower us in a variety of ways and can lead to a better understanding of various types of terrain and how we move across it. But, have you ever thought about the technologies that make all of this possible?
If you are my age, you probably learned to get a feel for your surroundings with the USGS Topographic Quadrangle Map Series or a state-based gazetteer, like a DeLorme map book. The government agencies and private-sector firms that created those cartographic products are now driven by geographic information systems (GIS), global positioning systems (GPS), and remote sensing technologies. Collectively, those high tech geospatial tools make it possible for users to access vast amounts of mapping data from a GPS unit or phone.
Entities obtain spatial data through the use of survey-grade GPS units and GPS-equipped vehicles while satellites, airplanes, and unmanned aerial vehicles (UAV) are capturing land cover, aerial photos, elevation data, and even thermal imagery. For example, the acquisition of highly-accurate elevation data using LiDAR, and aerial photos with a resolution in the range of 3″ to 6″ per pixel allows us to visualize and evaluate ground conditions like never before.
Keeping this data as up to date is a monumental task, but very critical as things change almost daily. Think about it . . . new roads are built, boundaries change, land is graded for development, trails are re-routed, sewer lines are installed, and waterways cut new paths across the landscape. A myriad of public and private entities continually track those changes. The features are then organized as layers, categorized by theme, spatially analyzed, and shared in a digital map format using GIS and web mapping services.
The value of geospatial technologies, and the underlying data, is realized when you can stand anywhere on this globe and view a map that is centered directly on your current location. Not only can you see the trail you’re traveling, but you can turn on the aerial imagery and see the stand of trees east of the trail just over the ridge. The ridge can be visualized due to the use of hillshading on the elevation data and its obvious that the last creek crossing is about 3 miles away in the valley below the next 2 switchbacks. Wow . . . you’ve never been there before, but you can see what lies ahead.
A wide array of proprietary and publicly available GIS data makes this possible. It is leveraged to provide the mapping services and downloadable data we so conveniently consume. In most cases, maps for our devices are in-expensive and readily available on the internet in formats that are compatible with our phone and GPS units. It is getting easier to do this everyday and the cost is rapidly diminishing.
Just imagine how handy this has been to Martin while laying out, and fine tuning the Tour of Idaho or for Jeff Stoess and the compilation of the Kentucky Adventure Tour. The aforementioned mapping resources have been key to their success in putting a line on the map we can follow. Their understanding of an area of interest, coupled with the dozens of map layers they studied, made both adventuresome routes possible for us to enjoy.
One important thing that goes along with having great maps at our disposal is using them to stay on the defined trail. Part of being a responsible off-road enthusiast is sticking to the route. Please realize that getting off the trail, or traveling cross country where not permitted, gives us all a bad name. You can’t claim to be lost if you have a GPS or map in-hand!
For many decades, the USGS Topographic Quadrangle Map series was “THE” base for anything mapping related. A wide array of Federal programs required that they be utilized for submittals, surveyors included them on plats as inset maps, and Kentucky-specific statutes relied on them for defining things like “blue line” streams. The “topo” or “quad” maps framed our perception of the nation’s surface and the planimetric features situated upon it.
In the mid 90s, USGS brought the topographic map into the digital age by releasing their Digital Raster Graphics (DRG) product. Finally, the basemap we’d relied on for ages was now in the background on our massive CRT monitors. For better or worse, our relationship to the maps was strengthened as we became mesmerized by the slow rendering raster image on our screens. GIS users quickly realized that the vector data they’d been creating and maintaining for years didn’t always line up perfectly with the DRGs. As a result, agencies quickly scrambled to adjust their points, lines, and polygons to the almighty map that we held as gospel. You know . . . they were made by USGS so they had to be correct!
Flash forward to December of 2011 when the Kentucky Aerial Photography and Elevation Program (KYAPED), or KyFromAbove effort, became a reality. With a focus on acquiring state-of-the-art LiDAR data and full-color, high resolution aerial photography, KyFromAbove promised a new view of the Commonwealth’s surface with a level of detail and clarity that had never been imagined. Since that time, and not surprisingly, one of the primary requests made by stakeholders in the GIS Community has been new LiDAR-derived contour lines.
In early 2017, as the impending reality of a statewide elevation model came into focus, the Kentucky Division of Geographic Information (DGI) began to investigate the creation of a statewide contour dataset and associated cartographic products and web services. Consultation with the Commonwealth’s GIS gurus prompted us to think outside of the box. Sometimes, bending the rules is necessary, even though it may lead to more complex processes and require more resources.
Although it was painful at times, going through this process was fruitful in that the KyTopo Map Series was conceived. The idea was to create a cartographic product that could be printed and shared as a cached web mapping service. Then, the primary derivative datasets (i.e. contours, hillshade, spot elevations, . . .) would be made available for download and published to the KyGeoNet. Two statewide contour datasets are slated for creation. One will be somewhat aligned with the USGS contour intervals in Kentucky (10’, 20, & 40’) and another statewide set of 5’ contours will leverage scale threshold settings and group layer functionality to adjust dynamically to the viewer’s map scale.
The KyTopo map series will be Kentucky-specific in several ways. First, an entirely new set of landscape-oriented quadrangle tiles have been developed. These new tiles align with our 5k tiling scheme and are in Kentucky’s Single Zone coordinate system rather than the traditional USGS UTM-based maps. The map area is exactly 30” wide (60,000’) by 20” tall (40,000’) and fits nicely on a standard Arch D (24” x 36”) printed page. The typical 1:24,000 (1” – 2000’) scale has been maintained and there are only 549 tiles as opposed to the 779 tiles offered by USGS. Importantly, the new tiles have square corners unlike the UTM version!
Next on the agenda was the task of coming up with names for the new map tiles. In many instances, the names from the old USGS quads were directly adopted. However, there were many cases where it just didn’t make sense. DGI staff studied the USGS methodology used to name its maps many years ago and employed that approach. Basically, the most prominent feature within a given tile was used for the name. For example, the largest city or the most centrally located place name (GNIS) was selected. In undeveloped areas, State Parks, State Forests, and natural features (i.e. streams, ridges, lakes, . . .) were used for naming the tile they dominated.
After pondering contour interval and index values, the standard USGS 10’, 20’ and 40’ intervals were embraced along with their associated 40’, 100’ and 200’ indices. Having these different interval levels is necessary due to great variance in elevation change as you move from east to west across the Commonwealth. Generally, the KyTopo intervals by tile, align fairly well with their USGS counterpart.
Weeks were spent gathering authoritative data for the project and fine tuning the map layout and symbology. Lots of thought was put into which layers should be included and which should be cast aside. As previously noted, all of the data is available in the public domain and much of it has been sourced directly from Kentucky State Government and a variety of Federal agencies. In fact, over 75% of the layers were already published to the KyGeoNet. It is anticipated that the maps will be updated on an annual basis so changes in the transportation network, forest cover, boundaries and other features can be updated accordingly.
The production hardware environment for the KyTopo project consists of one VMware-based 64-bit virtual server with 4 cores, 64GB of RAM, and 2TB of fiber-attached storage running Windows Server 2012 R2. Esri’s ArcGIS Desktop 10.3.1 and its Data Driven Pages functionality are the primary tools utilized for cartographic production. Additionally, Esri’s Production Mapping extension has proven useful in terms of handling layout constraints and customized map elements.
The map output process is done through a series of Python scripts that take advantage of the 64-bit background processing option within ArcGIS. Using this approach, the time taken to output each map was reduced by 10-15 minutes per map. Initially, draft maps were output to 300dpi PNG files for proofing and review. Feedback during this proofing process was used to further refine various aspects of the map series. Two types of georeferenced maps will be produced for distribution: collared and un-collared. The collared version will include all layout elements (i.e. legend, inset maps, title, . . . ) whereas the un-collared ones will be limited to the area within the data frame.
Significant effort was put into establishing an appropriate symbol set for this map series, and DGI staff consulted the USGS Topographic Map Symbols publication for guidance. Some of its symbols were adopted while others were adjusted to enhance readability. Once a symbol was selected, it was added to the KyTopo.style library using the ArcGIS Style Manager. This tool has proven to be an effective and efficient way to manage symbology for the production effort.
Typography, or cartographic labeling, is another aspect of this process that has received a tremendous amount of attention. Feature labels, titles, legends and notes are just some of the text-based components of a cartographic product. Using the correct typeface for each element is critical when compiling a quality map. Discounting the importance of typography during cartographic production can lead to a map that is difficult to interpret or one that draws the user’s attention away from the subject. Per the principles of thematic map design, serif typefaces were used for the map title, layout elements such as graticule labels, and hydrographic features. Road shields, road names, elevation features, and some selected layout elements employ sans serif typefaces. Feature label placement was handled by enabling Maplex functionality in the map document. Hours were devoted to fine tuning the Maplex labeling rules so more prominent features were placed prior to those of less importance.
Every great map series includes both index and inset maps so users can see the location of each map tile in context. The KyTopo index map is situated in the upper right corner of the layout and highlights the given map tile on a statewide view. The inset map has been positioned below the legend and shows the eight tiles that surround each map. Including both of these components on the map has proven to be quite helpful. Using separate data frames for each, in conjunction with Data Driven Pages functionality, made this an automatable task.
One of the most frustrating layout elements to deal with was the legend. ArcGIS has some nice legend generation tools but DGI was unable to fully utilize its output. The legend created with the automated tools just didn’t produce the most desirable results. For example, the symbol for spot elevations would appear in the legend but the actual elevation value would not be placed beside it. Reluctantly, a partially manual process that uses Adobe Photoshop and Illustrator was put into place. It is a bit cumbersome, but the resulting legend is much nicer and better reflects the features within the map series.
At the time of the writing of this article, draft maps for KyTopo have only been generated for 394 of the 549 tiles. This is being driven by the current availability of LiDAR data for the Commonwealth. Statewide LiDAR coverage should be a reality by the end of 2017. Once that has been completed, contours will be generated for the remaining tiles and final map production will commence. It is planned that the map products and ancillary data elements will be made available on the KyGeoNet by Spring of 2018.
Stay tuned for final release dates and updated product specifications!
A little over ten years ago an article entitled “GIS as a Utility: Kentucky’s Enterprise Implementation” appeared in Esri’s 2005 summer edition of ArcNews. The article provided an overview of Kentucky’s new Enterprise GIS implementation, discussed the challenges associated with making spatial data accessible in a networked environment, and highlighted how various agencies were beginning to leverage this new resource. It has been over a decade since that article’s publication, and the Commonwealth’s enterprise GIS has matured and become a critical component of business processes both within and outside State Government. This article outlines the efforts managed by the Kentucky Division of Geographic Information (DGI) that have made this implementation successful and sustainable throughout evolving technologies, tight budget cycles and changing policy directions.
One of the least desirable tasks in information technology is documentation. No one likes to document code, routine processes, or data for that matter. In the GIS world, creating “metadata” is the task that no one likes to see on their “to do” list but successful implementations require that it is completed on a layer-by-layer basis without exception. Like it or not, metadata is the keystone of Kentucky’s spatial data services. It has been tough at times to get agencies on this bandwagon, but once they’ve bought in, the value becomes more apparent. The policy states, “All spatial data resources shall have a complete metadata record in order to be included in the Commonwealth’s spatial data repository.” The existence of and strict adherence to this policy has had a positive impact on geospatial data integrity that carries on today.
Back in 2005, the Kentucky Geography Network (KyGeoNet) was based upon ArcIMS Metadata services, and there were less than 50 metadata records created by a half a dozen publishers. The whole concept of creating metadata was new and many just saw it as something else they had to do on top of everything else on their plate. For many GIS practitioners, it was a low priority until they realized that the Commonwealth’s spatial data repository would not grow if metadata was not created for the valuable resources scattered across the GIS community.
The value in making geospatial data accessible certainly outweighs the amount of effort required to complete a minimal metadata record and publish it to a repository. Essentially, this is a form of “cost avoidance” in that if users can find a resource, read the metadata to learn about its characteristics, and then download it for their use, they are less likely to call or e-mail with questions or requests. This becomes quite apparent once you’ve fielded 20+ calls about a given dataset and then realize that a published metadata record could have answered questions about the resource. Being asked the same question over and over again makes for grumpy GIS staff.
Fast forward to 2016 and there are now approximately 700 metadata records being actively managed by over two-dozen responsible publishers on what is known as the KyGeoPortal. The service is currently based upon a recent release of Esri’s open source product known as Geoportal Server. This solution includes the ability to search for and create metadata records associated with web mapping and image services, as well as downloadable datasets, static map images, standards documents and web mapping applications. It is a wonderful resource that is leveraged frequently by thousands of users on a monthly basis.
Average Monthly FTP Server Requests for 2016 (through July):
5.9 million requests/several thousand unique visitors
As noted in the 2005 article, KyVector and KyRaster are the primary enterprise-based services that are accessible to the hundreds of ArcGIS Desktop and CAD users within Kentucky State Government. Each user is on the Wide Area Network (WAN) and enjoys fast, network-based access to these geospatial services hosted at the Commonwealth Computing Center in Frankfort, Kentucky. This level of service would not be possible without Commonwealth Office of Technology’s (COT) robust WAN and server infrastructure as the delivery mechanism.
Centralizing the data repository has reduced direct agency storage costs and eliminated the issue of outdated data residing on agency-based file servers. Agencies no longer have to expend resources on updating data that originates from external sources. All data is current, easily accessible, and includes associated metadata. This level of data availability and integrity promotes confidence in results when GIS technologies are utilized for inventory, planning, response, or business process decision support.
KyVector, an ArcSDE-based server solution that runs in conjunction with Microsoft SQL Server is hosted on three separate physical boxes. One server (kysdewww) is dedicated to providing vector-based data to State Government agencies that host their own web mapping services and applications. Another server (kysdewan) is reserved for the hundreds of ArcGIS Desktop users on the WAN, while the third is used for staging and failover/outage scenarios. This proven configuration reduces contention for server resources and ensures that data delivery to users and applications is carried out in the most efficient and effective manner.
All data resources housed in KyVector have met the prerequisites of having a full metadata record and having been published to the KyGeoNet. In 2005, it was noted that 125 layers were in KyVector. Today, the number has more than doubled and stands at just over 300 layers. New layers are added periodically as they are created or if there is a need to share the resource across the enterprise.
Access to KyVector has been streamlined by using layerfiles that include connectivity properties, appropriate display scale thresholds, and suggested symbology settings. The organization, maintenance and distribution of the layerfiles is carried out by DGI. Users can start a new, or add to an existing, map document with ease by browsing a thematically organized folder of layerfiles.
Weekly data updates to KyVector, are fed from the agencies to the staging server then subsequently moved to production during the update procedure. Users access this resource with confidence knowing that it is the most current version available at any given time.
A decade ago, KyRaster was ArcSDE-based but newer technologies have prevailed. ArcGIS Image Server is now employed to provide users with access to raster-based data such as aerial photography, topographic maps, digital elevation models, and land cover. The primary benefit of ArcGIS Image Server over raster datasets stored in ArcSDE is the significant reduction in the time needed to process raster data so that it is accessible to end users. In the past, days or even weeks would be consumed just loading imagery into a raster dataset in ArcSDE. But now, with mosaic datasets and Image Server, we are able to get data out to the users in hours, or days at the very most.
Additionally, DGI’s raster storage footprint has been reduced by employing function chains within mosaic datasets. This functionality allows us to store the data once but serve it up in multiple projections, with different band combinations, or with special rendering options. For example, a function can be inserted into the chain of a digital elevation model (DEM) that allows it to be rendered as hillshade, shaded relief, or percent slope. The DEM is only stored once on disk, however the functions can be applied to each request on the fly.
As with KyVector, ArcGIS Desktop users connect to KyRaster using layerfiles that include connection properties, display scale thresholds, compression settings and custom rendering options. The performance of Image Server is on par with ArcSDE on the WAN, and those that aren’t on the State’s network have learned to change compression settings on the client-side to achieve better performance on slower networks. This approach has worked well in state field offices, at local governments, and by private sector users such as engineers, surveyors, and consultants.
Web Mapping Services
Both KyVector and KyRaster are the backbone of the Commonwealth’s publically accessible web mapping and image services. Dynamic and cached map services are hosted on kygisserver.ky.gov and provided in Web Mercator projection using ArcGIS Server. These services power hundreds of web mapping applications internal and external to State Government. The Commonwealth Base Map and the Commonwealth Street Base Map are the two primary cached services, but other statewide imagery and land cover layers have been cached as well.
The balance of the services are considered to be dynamic, as they pull data directly from KyVector to fulfill every single server request. Each time a user zooms in, zooms out, pans one direction or another, a request is made to the map server which in turn reaches out to KyVector for the data.
Average Monthly Map Server Requests for 2016 (through July): 9 million requests/several thousand unique visitors
Kentucky’s image and elevation services are hosted on kyraster.ky.gov in Kentucky Single Zone Projection using ArcGIS Server and the Image Server extension. Employees with ArcGIS Desktop in State Government are the primary users of this server, however there are a growing number of users outside the WAN. When ArcSDE was the server solution for imagery, only users on the WAN could gain access. Image Server now allows the Commonwealth to meet the needs of internal and external users with the same computing infrastructure yet still maintain a high level of performance and availability.
Average Monthly Image Server Requests for 2016 (through July): 7.5 million requests/several thousand unique visitors
Like many other users of geospatial technologies, DGI has incorporated ArcGIS Online (AGOL) into its solutions and workflows. AGOL is Esri’s scalable and secure software-as-a-service cloud-based mapping platform. It includes a rich collection of web mapping application templates, visualization and analytical tools, and ready-to-use data sources that make geospatial resources more accessible and user friendly. DGI’s organizational site on AGOL, KyGovMaps, showcases a wide variety of Kentucky-specific web maps, applications, and services. Data and maps that were once only usable by those with expensive GIS software are now available to most everyone. GIS isn’t just for propeller heads anymore!
A customized gallery page on the KyGeoNet makes it easy to locate featured web maps, mapping applications, and thematic galleries that interest a wide array of users. Lately, some of the most popular applications have been those in our collection of Story Maps. Recreational resources, imagery, historic landmarks, demographics, elevation data, and physiographic regions are a sampling of the topics covered using story map templates that we’ve customized.
The Next Decade
If the pattern continues, combined total server requests for the 2016 calendar year will easily exceed 250 million. This alone is a true testatment to the success of the Commonwealth’s Enterprise GIS implementation. The growth in adoption and usage of geospatial technologies during the past decade has been astounding, and Kentucky’s capabilities in this realm have kept up with the pace of advancement. It will be interesting to look back at this, and the preceding 2005 article, during the middle of the next decade to see where we stand. Live data feeds, UAV acquired imagery, and 3D visualization are just some of the new trends that will shape the makeup of the Kentucky’s Enterprise GIS down the line. Stay tuned!
One nice part about my job doing GIS for State Government is the access to the most current aerial photography and other data layers. The Federal Government through the National Aerial Inventory Program (NAIP) has acquired leaf-on aerial photography of the entire state at a 1 meter resolution every two years since 2004. In 2006 2 foot resolution (much better) was obtained. We just got the 2010 photography a couple weeks ago and have been working to mosaic it all together and make it available for everyone to use and/or view.
Checkout the photo above of Horseshoe Bend taken this summer as a part of the NAIP effort. You can see every twist and turn of the MX track and if you know where to look you’ll see where the KORHS course crossed some fields and where it tied into the MX track. If you look closely you can see where scoring was just inside the little kids track. Also, every jump and corner is very clear on the Supercross track on the left hand side of the image.
I cruised around the state looking at the new imagery and in most of the KORHS race locations there were few indications of our presence that were visible in the imagery. Mainly, the scoring areas and some corners on the grass track sections. These types of things are visible at the Sawmill in the area next to the road. Now in the case of Salt River Run, the imagery was acquired before the event. Thus, there is no sign there was a race there this year, but you can count the roll bales in one of his fields with great precision!