I moved to a new blog awhile back. I post to Geospatial Technologies in Education. All the posts from this blog have been ported over there. Follow over there because I don't use this site anymore. I keep this blog alive because I like all the sidebar links, most of which may be broken. Sorry about that part.
Tuesday, April 27, 2010
As of January 1, I am no longer at Vassar College. I started the GIS at Vassar blog for myself, so that I could do my job better, to keep track of links to useful web sites, to remember Google Earth KML network links useful for teaching, to answer questions that faculty and students had around the college, and to remember projects that I found valuable. Over the three and a half years of running this blog (I started it in September 2006), I've had an average of about 50 page views a day. And though I really slowed down on my posting while I've been doing my Fulbright, I still get visitors. So the GIS at Vassar blog and the old posts will stay right where it is, but I will no longer add new posts to it. I will start more posting to a new blog called Geospatial Technologies in Education that I set up on Posterous. I migrated the old posts there but will keep adding new material. Please follow me there.
And if you're in the market for an educational and geospatial technology professional with skills in ArcGIS, Google Earth, field technologies like tablet PCs and GPS receivers, and spatial literacy, I'm available... stewart (dot) meg (at) gmail (dot) com
Posted by Meg at 10:27 AM
Wednesday, April 14, 2010
I have finished up my Google Earth project for the Grenadines. If I didn't have grant proposals, papers, job searches and interviews, and other mundane tasks like that, I would have completed this project in about six weeks. That's the way it goes sometimes.
Here is the link to the Grenadines MarSIS geodata KMZ for your viewing pleasure. The link will launch Google Earth.
I also created (thanks to Valery Hronusov) a KMZ of a nautical chart for the Grenadines. Check it out. It’s big, about 10MB.
And check out the video that describes what’s in the file and explains how to make sense of the many layers of data.
After the first draft of this KMZ, I ran a post that gave a little how-did-I-do-it synopsis. Now I will discuss what the final issues were and how I resolved them.
The original database I was handed was compiled as an ESRI geodatabase. I'll admit it, I have been slow to the geodatabase lifestyle. When I exported shapefiles in my first draft, I had lots of '1's and '2's that looked like this: Island= 1. What I needed to do was export feature classes to a shapefile. I had to switch all the ‘1’s to equal the proper designation that the researcher applied back when the database was built. So now the file shows Island = Mayreau. It was a bear, I must say. But it made it so that the KMZ placemarks contained the information that the original layer contained. It had to be done.
I had to use an HTML editor. I don’t think I mentioned this in my other post, but it is required. That little HTML class I took back in 1996 at the community college still comes in handy. I used HTML-Kit. It is free and works like a charm.
The MarSIS logo used to not stay static. It would stretch and squish when the Google Earth window was resized. I used this helpful site that gave me the code to help me created the screen overlay for the MarSIS logo. This interactive KML code site is superb as it gives you the code to work up in your HTML editor.
I got the snippets, those extra lines of text under the folder or placemark (circled above), to go away. That was done with the HTML editor. Use this snippet code
after the folder or placemark name tag. Find the snippet code discussion here.
I added a legend for the habitats, both shallow water and deep water, using Adobe Illustrator and then creating a screen overlay, see above, with that raster but this time hard-wired to the lower left-hand corner of the screen.
I mentioned in another post that I had fabulous success with SuperOverlay. It bears repeating. I had a large raster file map that I wanted to bring into the project. SuperOverlay was the way to go because it automated the process of slicing up the map to create a smaller KMZ. It was still 10MB so we did not include it with the project KMZ.
Lastly, I had about 200 photographs and 200 videos with geolocations by latitude and longitude to include in the KMZ. Google Spreadsheet Mapper 2.0 made geolocating photos a breeze. Video are a different story. The first draft of the project had a simple screen capture taken from each of the 200 videos that I used as a placeholder until I uploaded each of the videos to a place. That place was YouTube, naturally. The problem is, Spreadsheet Mapper does not take video embed codes, unfortunately. What I did was meticulously copied each video’s embed code as the video finished being uploaded, saved those codes to MS Excel (could have used Google Docs but didn’t), then modified the original previously geolocated placemarks with screen captures that I had created in Spreadsheet Mapper. I erased out the tag for the image and copied in the embed code for the video. It worked very well. All of this was NOT a big deal and I spend more time dreading doing it than I needed to. In retrospect, using the Spreadsheet Mapper to geolocate the video locations was very useful for quickly and accurately finding where the videos belonged in the world.
And now I am done and popping some sparkling wine. On to the next project...organizing the CERMES department's GIS data library. Cheers!
Mashup image above uses a Creative Commons photo from titanium22’s photostream.
Posted by Meg at 9:03 PM
Friday, April 09, 2010
Posted by Meg at 3:07 PM
Thursday, March 18, 2010
The Grenadines MarSIS project web page now links out the the KMZ file that I've been working on. Find it here. You can also get the nautical chart. We made a video to help folks navigate the many layers of geo-data. I think it might help since there are a lot of layers to keep track of.
Posted by Meg at 12:51 AM
Wednesday, March 03, 2010
You know those times that you have a large scanned map of your field area or project site and you want to use Image Overlay to add the map to a Google Earth project? If you want the file size to remain small and portable, you need to re-sample the map and lose all your useful information, or you can try to keep the map as the original file size and then you have an enormous KML to share. What to do? Use Superoverlay!
Superoverlay, developed by Valery Haronusov, has been around as long as I've been dabbling in Google Earth, I just never gave it a try. Until now.
I've been working on a Google Earth project based in the Grenadines. One of the files that I needed to add to the project folder is a nautical chart (shown above) that in digital form is a 110 MB file. It would be too unwieldy - both geographically and in digital file size - to try and do an Image Overlay for this file so I needed to make it smaller and so used Superoverlay. Superoverlay cuts up the raster file into rectangular chunks and creates smaller raster tiles. This makes it so that when you zoom in, the tiles that draw are more clear and have the nice resolution that you want maintained. You can do this by hand. But why?
I followed the Superoverlay instructions, which are great. Because the chart was in UTM, I needed to re-project to the WGS84 coordinate system. Superoverlay uses the world file (i.e., *.jpx, or *.tfw) for spatial reference. The spatial reference is read directly from the world file. The output I got for the KMZ file of the nautical map was 10.5 MB.
There was one slight problem. The nautical chart looked great but I have other files in my project that need to draw over the chart, and the chart was not drawing on the bottom. This had something to do with the DrawOrder tag. I contacted Valery to see how this could be fixed and he found this ordering issue to be a bug and he fixed it!
If you want your Superoverlay map to draw under your other files, use a negative number in the Start draw order box (circled in green above). I found that -20 gave me the order that I was looking for. By the way, it may seem like you can just reorder your KML files in the folder and place the Superoverlay map at the bottom and then this problem is fixed (like in ArcGIS). It doesn't work that way in Google Earth, unfortunately.
Shown above is the mapping of the shallow marine habitat near Mayreau and Union islands with the nautical chart layered underneath. Just the way I wanted it.
If you're interested in GTOPO30 maps for the world (shown below), one has been created by Valery using Superoverlay. Click here to find it.
This is how to make the file shown above. So, if you have super huge raster datasets, what do you do? Use online storage. Here is some useful information from Valery on the process of these larger Superoverlay projects:
"I use Dropbox and Amazon S3 storage for tiles. For huge rasters I use Global Mapper + Superoverlay or KML2KML. Global Mapper makes large tiles (2048 x 2048) from huge and Superoverlay makes small tiles (256 x 256) from large. And with the last one, we can copy to cloud web storage (S3) an absolute "authoring" independent low cost solution. I use Cloudberry as an Amazon S3 explorer."Thanks, Valery! This is all great stuff and so helpful.
Posted by Meg at 6:37 AM
Friday, February 05, 2010
The big news this past week was Steve Jobs' throwing the coverlet off of Apple's long-awaited iPad. Though it has long been rumored that Apple would make a tablet PC, none had materialized. The wait continues. Touted as a 'tablet PC,' the iPad is not that. The fever surrounding the launch of the yet-unnamed-Apple-tablet was incredible. Then once it hit the streets, everyone and her auntie had gushy words for it. On Twitter, #iPad was number one with a bullet all week, even in the face of people still buried under rubble in Haiti and the President giving his first State of the Union address. It's still a Trending Topic today. It's weird. Even my favorite political podcast at Slate, last week, couldn't resist weighing in on the iPad. Why? I really don't get all the fuss as yet. Another favorite podcast of mine, Digital Campus, seemed to think that the iPad would be awesome for archaeology or geology students because those students would have all this internet information at the their fingertips when they're out on their digs. What?! Field researchers need data collection tools (like a tablet PC) when they're out there in the muddy and the dusty and the boggy. Maybe I'm missing the point.
This line from Steve Jobs' announcement demo doesn't help give me any more clarity: "It’s phenomenal to hold the Internet in your hands." Umm, I'm using a netbook /Blackberry/iPhone /tablet PC right now and I'm actually holding the Internet in my hands. Weird.
Look, I want more choice in hand-held devices just as much as the next guy. But what the 1st Gen iPad seems to be is a large iTouch. Maybe before the 2nd Gen iPad comes out, Apple will listen to all the howling cries of ...So What!...*yawn*...Big Deal!...and actually add some functionality to the pretty little thing. Or it could just remain a Kindle alternative. That's fine with me, except for the fact that I don't really read books anymore. I listen to them on my iPOD!
Dear Apple, If you really want to make a killer tablet PC, make it so that I can do Adobe Photoshop and Illustrator using the tablet's pen, I don't want to use a Wacom tablet because I think your new iPad would deem them obsolete. While you're at it, give me some GPS location-based ability. Didn't you hear that everything these days is really GEO-everything? In an ideal world, I could actually make maps on an iPad using a GIS software and make drawings using something like AutoCAD, but I digress. But for goodness sake, I want to make phone calls, make videos, shoot photographs, record a podcast. Awesome price on these little 1.5 lb babies, by the way. Don't change that, alright? Alright.And all you fans and rushers-out-to-get-the-new-
So, for the Apple iPad, take my advice: wait. And as for the unfortunate name...look at 'Google'?
Posted by Meg at 1:56 PM
Tuesday, January 19, 2010
While traveling to St. Vincent, Grenada and Union Island this past November, I took some video of our workshops and the intervening travel between workshops. Kim Baldwin, PhD student at CERMES, and I were there to talk about her MarSIS marine mapping project and show how to view the geodata in Google Earth. I only just now was able to get to finishing up editing the video from our trip. I blogged about the workshops on my other blog, if you're more of a reader than a viewer of video.
Posted by Meg at 10:15 AM
Sunday, November 29, 2009
I don't usually promote clothing in this space but this site, Geo-Tee, crossed my desk and I thought they have some cool GISy offerings just in time for the holidays. They've got T-shirts and stickers for the GIS-inclined. Here's what they've got to say about themselves:
We supply apparel for the seasoned old-school GIS’r that was doing GIS with punch cards, to the newbie who just figured out the difference between a datum and a projection.
And frankly, the shirt above, Will Map for Food, may be one I'll need to order in the near future!
Posted by Meg at 7:25 PM
Tuesday, November 24, 2009
I haven't written in awhile nor kept this blog updated on my project work. I've been busily working away at CERMES on a decent first draft of of the Grenadines MarSIS Google Earth file that I discussed previously on my other blog. Though I have a post that says I got the geodata back on September, I really got the final, final data about two weeks before a workshop trip, launching the file that I'm going to explain in this post.
I went with MarSIS project leader and GIS data collector, Kim Baldwin, on a presentation tour to three public workshops the week of Nov 9 to 13, 2009, and you can read about that trip here. It was amazing. She presented the GIS work she's done and I showed the participants how to use Google Earth and how to use the MarSIS data in Google Earth. The MarSIS project is a marine-based resources project for the Grenadine Islands, between (but not including) the islands of St. Vincent and Grenada in the Caribbean. That's all about the workshops for now.
As a result of the workshops, we got terrific feedback on the usability and functionality of the project KML. There were some things that were not so clear, too. We are not ready to show or give out the final version of the KML. Once the file is ready, it will be launched on the MarSIS web site as a Google Earth API plug-in and I'll blog about it. The plan for this user-ready final product is February 2010.
Above, is a list of the subsets of data layers found in the MarSIS project. I basically took Kim's Geodatabase feature layers and exported them using the free ArcScript Export to KML or the not free Xtools Pro. There was some iteration involved but I used one for one thing and one for another. It was pretty painless.
The two screen captures below show some of the results of the export process. There's no legend which is something I will work on in the coming months.
Above, shows the shallow water habitat polygons, locations of whelks, and sea turtle nesting beaches around and near Mayreau Island.
And this screen capture above shows Space-Use Patterns (pink font) and Marine Resource Users (blue font) layers turned on for the area surrounding Union Island.
Kim has a lot of underwater photographs that she's used to identify habitat type and map the sea floor. Above is an example of what one of the photos looks like in the KML. I also blogged about the preliminary results on another blog I have. She also took a lot of underwater video for the deeper areas for the same reason, to map the sea floor. The videos will be included, but for the recent presentations, we used a representative single frame as a jpeg. There are nearly 400 locations and images that are in the KML. Here is how to put lot's of photos into KML.
Use Spreadsheet Mapper 2.0 You are given a choice of six placemarks templates that you can customize if you wish. You use a template built in a Google Doc spreadsheet that you then create a network link so that you can "automagically" build the placemarks for your georeferenced photos. Here is the Google Doc template. When you start to load you're information, you should prepare a table with latitude and longitude of each shot, URL to each photo (or video), some metadata perhaps (i.e., "The habitat shown in the photograph is classified as Sea Grass, the fisher classification is Sea Grass and the research description is: Rubble, w/macroalgae & syrigodium. The depth at this location is 29 feet.") which is semi-easily obtained from the shapefile table with a little concatenation. Link to the project page or blog. Then Publish your Google Doc and you're live!
Finally, I made a KML Screen Overlay described at the Google Earth blog with the MarSIS project logo. You can see it in the screen captures.
The plan for taking this MarSIS KML to a final version in the next couple of months is: 1) create legends for the polygonal habitat maps, 2) figure out how to use SuperOverlay to slice up and load a nautical map of the Grenadines (shown as an empty folder above called "Imagery/Maps"), 3) fix some of the metadata that didn't export properly from the geodatabase (i.e., island = 6 should be island = Palm Island), 4) figure out how to get rid of snippets, and 5) embed the map into the MarSIS web page using Google Earth API.
Note: Here is the final installment on steps I took to create this KMZ file.
Posted by Meg at 9:30 AM
Wednesday, October 21, 2009
We're always looking for geospatial data to teach with. Finding data for the U.S. is relatively easy, but finding data for other places has often been more challenging. AfricaMap is a web mapping interface created by Harvard University's Center for Geographic Analysis that can be used stand-alone in a classroom because there is a rich set of data available for viewing. The layers that I looked at, some shown below, drew quickly and came with a legend(!). Very nice accomplishment for CGA after being in operation just under three years. Here is a description of the AfricaMap project: This project attempts to address a basic problem for all scholarship on Africa that treats where things happen as necessary to understanding how and why they happen: finding places on a map. Despite the existence of excellent public maps for Africa, to date there is no common source that allows students, researchers, and the general public to:
"AfricaMap is based on the Harvard University Geospatial Infrastructure (HUG) platform, and was developed by the Center for Geographic Analysis to make spatial data on Africa easier for researchers to discover and explore. Sounds great! Here are some maps available through the AfricaMap site and some comments on how to use the site.
The above map is the 1722 Delisle Carte d'Afrique found under Map Layers -> Historic Maps 1600 to 1800. I lead with this map because I love old maps.
This map shows a map of ethnolinguistic families (mapped in 2001). Remember to add the Legend found over to the right in Map Layers.
Under the Environmental layer, you will find just three layers, rivers, soils and, shown above, surficial geology. Note that this legend is not all that useful.
AfricaMap allows for downloading of their data, but only some of the layers are currently available. That may be, in part, why they this web map is in beta. How to get to the download function: Click on Map Layers, scroll all the way over to the right. There is a Download tab, don't click there, follow the arrow down to your layer and click "download" there. If the data are available, you will be taken to the web page and source of the geospatial data, if not available, you will be given this message "Mapping data will be made available for download." That is quite hopeful.
Thanks go to Diana Sinton for sharing this web map site.
This project attempts to address a basic problem for all scholarship on Africa that treats where things happen as necessary to understanding how and why they happen: finding places on a map. Despite the existence of excellent public maps for Africa, to date there is no common source that allows students, researchers, and the general public to:
Posted by Meg at 9:33 AM