I just got an email from Google Earth Outreach about the release of Google Earth 5. As speculated over and over, and discussed here just over a year ago, it includes the new bathymetry layer, and it now has ocean depths. You can grab the update from the Google Earth website.
A coworker of mine, Jason Jones, created a projected change in population file for the state of California from the year 2000 to 2030. This was a calculation based on data downloaded from the Natural Resource Ecology Lab. In the map below, gray patches are varying degrees of current urbanized areas. White patches are private, rural areas. These locations will probably not change in the next 20 years. Shades of red show increasing amounts of population growth. Overall, lightest pink shows some urban edge infill and far suburbs. The darkest red is mostly suburban growth, as you can see these colors basically ring around metropolitan areas. I tossed in public/protected lands to show the ownership landscape and to highlight threatened areas in need of conservation. An interesting thing to note: if you look at northern central Marin (just north of San Francisco) (sorry about the lack of labels for those unfamiliar with the area), there is a swath of white, rural, that is surrounded by protected areas to the west and projected population growth to the east. My guess is that this private rural land, a prime location for development in the Bay Area, actually has a ton of conservation easements in place on it. This diminishes the chances of population growth by 2030.
Okay, so I can’t get over these data quite yet. One more post. Oh, and yes, Happy Holidays! I’m heading to New York this evening — nothing like a Christmas Eve red eye, but I’m excited to see family. Back to LIDAR for one second. I wanted to show what the calculation of the DEM with trees minus the DEM without trees (which equals trees, basically) looks like without symbolizing by land cover type. I originally got this idea from an article discussing LIDAR analysis of gigantic redwoods along the Northern California coast, a project funded by Save the Redwoods League. So, just below you can see the height of vegetation simply shown with a color ramp. The darkest red represent the tallest trees, the light blue shows the water or barren land. This image says so much, it kind of blows my mind. In the northwest, you can see where the tallest trees follow the natural ravine, along the banks of a stream. But that square in the center, and the line of trees in the southeast, that could only be created by one thing: humans and their crazy “property rights.”
So, the next image shows parcel boundaries, and yes, they line right up with the drastic difference in tree height. This goes to show that just because an area may be forested with identical tree species, there can still be extreme differences in habitat type caused by forest management.
Is there really a difference between looking at these images and just inspecting some aerial imagery? There is, check it out below. Zooming to the property line in the southeast, you see the original calculation, just as above, then an aerial image with parcel lines, followed by just an aerial image. If you look close enough at the aerial, you can make out a property line, but it is definitely not as glaring as the original calculation. Cheers! Happy Holidays!
A search for California LIDAR brought up a NASA Planetary Geodynamics Laboratory link with some very localized LIDAR geotiffs. I recognized the name Fort Ross, so downloaded these images, and they’re great. There are four files for each area, a digital elevation model (DEM) with tree heights, a DEM without tree heights, and hillshades of each DEM. I thought about how I could call out forested areas, and came to the conclusion that I could subtract the DEM without trees from the DEM with trees to just get heights of objects above the ground. The biggest differences, the highest trees, are shown as dark green, the smaller differences are shown as yellowish grasses and shrubs, and no difference is shown as blue water. Set transparencies between this layer and both hillshades, and we get a pretty great land cover map. Click on the image below to see a high resolution version, and further below you will see the DEMs and calculation described above.
I found some really great historical photos for about half of Illinois today. The sample here shows historic downtown Geneva in 1939. For those not in the know, Geneva is the middle city (and arguably more superior) of the Tri-Cities on the banks of the Fox River in the suburbs of Chicago along with Saint Charles to the north and Batavia to the south. If you compare the two images below, you see an incredible change to the landscape as farms disappear and are replaced by subdivisions. The second image was taken in 2007. Forests and tree-lined streets have become much more dense. It’s interesting, in the soutwest of town, there is a quarter-circular suburban development that is being constructed in the 1939 photo, suggesting new growth to the town center that was to become the defining land use trend of the next 75 years throughout the United States. New golf courses have been added, but the smaller, rectangular course just to the west of downtown was already established. I’ve georeferenced the image in Google Earth if you want to take a look around for yourself (~3.5mb) or if you don’t want to go for the download, here’s an animated GIF to check out.
Tim the Cartographer for Obama
The points below represent houses in Chittenden County, Vermont, around the Burlington area. Yellow dots are houses in the year 2000, and red dots show houses by the year 2004. When overlayed with each other, we see where developments have been built over the course of just four years. One of the largest developments is pulled out in the image below. As expected, most of these new homes are outside of town centers, adding to the continuous sprawl of the region around Burlington. These data come from the Vermont Center for Geographic Information (VCGI).
The organization that I work for, GreenInfo Network, has been composing some great web sites lately, and I wanted to take a post to let people know about them. A while back we started creating our ParkInfo website and it has recently come to fruition. At this site you can search throughout California for park names directly or search by addresses, cities/locations or zip codes to find parks in your area. Once you click GO, the map zooms to that location and a list of parks shows up under search results. Single click on a protected area on the map, it will highlight, and a bubble pops up with ownership and acreage information.
This site showcases one of our primary data sets here at GreenInfo, the California Protected Areas Database (CPAD). CPAD is a collection of federal, state, county, city, special district and non-governmental agency protected lands. You can learn more about this initiative and download the spatial data at our CPAD site. In an effort to continually improve CPAD, you can help review the data and provide feedback on particular properties that we will use to edit the database before our next release. And I’ll report back in the future about an exciting online property boundary editing tool that will help streamline the review process.
Both of these sites were developed using the Google Maps API. TileCache and MapServer are being used to generate the custom basemap in ParkInfo and the parks layer.
Rumor has it that Pennsylvania’s Department of Conservation and Natural Resources is flying all about the great Keystone State collecting and then processing LIDAR data. They are creating 3.2-meter DEMs and 2-foot contours from the data, which kind of blows my mind, and is really impressive. So far, it looks as though Luzerne County data are posted on the Penn State PASDA clearinghouse site, and rumor also has it that at some point, raw LIDAR data will be released. So what that means is that you will be able to analyze canopy, understory, and ground elevations for some random forest in middle-of-nowhere Pennsylvania. And this, I think, is brilliant. Well done, Pennsylvania.
This is admittedly a bit off topic, but, there are moments where GIS is used in the video, so I thought I’d give it a little post. A bit in the same boat as the Radiohead video I composed, ArcScene was also a major player in this endeavor. I began writing the song two years ago and had moments of inspiration in the past couple of weeks that pushed me to finish it. The youtube video is just below, but for a higher resolution version and to download the song, follow this link…
