Introduction

The Great VRML Model - The 3D Earth is a three dimensional VRML globe model of planet earth. The input for the model is taken from the GTOPO30 data set published by the USGS. GTOPO30 is a global digital elevation model (DEM) with a horizontal grid spacing of 30 arc seconds (approximately 1 kilometer).

The Great VRML Model - The 3D Earth also displays populated place information of about 90000 places derived from the Digital Chart of the World data set published by the Pennsylvania State University.

The Great VRML Model - The 3D Earth consists of about 800 million polygons that get seamlessly downloaded while you browse the model.

Prerequisites

The Great VRML Model - The 3D Earth can be viewed with blaxxun Contact a free VRML viewer that can be downloaded from blaxxun interactive. This programs runs only on WINDOWS computers that have a graphics card with at least 2 MB of video memory. A setup with WIN 98 or WIN 2000 and with DirectX 8 and a 32 MB AGP video card is prefered.

Enter The Great VRML Model - The 3D Earth

Enter the name of a place and SEARCH for it:


Name:

Or select a place where you want to start: A random place on earth

NEW YORK	ATLANTA	DENVER	SAN FRANCISCO
MEXICO CITY	LIMA	RIO DE JANEIRO	SANTIAGO
LONDON	MUNICH	ROMA	ISTANBUL
KABUL	NEW DEHLI	BEJING	TOKYO
TANGER	ALEXANDRIA	KINSHASA	CAPE TOWN
MANILA	CANBERRA	HILO	DJAKARTA

Or enter the latitude and longitude of a place in degrees and minutes and hit SUBMIT:


Latitude: 0000 - 8959 North  South  Longitude: 00000 - 17959 East  West 

Screenshots from The Great VRML Model - The 3D Earth

San Francisco	Mount Fuji	Bavarian Alps	Jerusalem

Frequently Asked Questions

Why is it called The Great VRML Model - The 3D Earth?

It is a 3D VRML model of the Earth. "Great" is used here in the meaning of "big", as in the "Great Lakes". The model allows you to browse planet earth seamlessly, it uses a vertex for each 30 arc seconds. Thus the model consists of 360 * 60 * 2 * 150 * 60 * 2 == 777600000 vertexes. Let me know if you know of a VRML model with more vertexes.

Is it really the entire planet earth?

No it is not. I omitted Antartica, because GTOPO30 uses a different orientation for those tiles and I was too lazy to write the special code necessary to handle it.

The model covers the planet earth from 179 Deg West to 179 Deg East and from 60 Deg South to 90 Deg North.

You cannot walk over the 180 Deg E/W line and I do not know what happens if you go further north than 90 Deg and further south than 60 Deg.

Why does it solely work with blaxxun Contact?

I wanted to make the model multi-user in order to allow joint browsing of the planet. The easiest solution that I know for making it multi-user is the blaxxun Contact Frameset Generator.

I have a similar digital elevation model which I want to visualize, would that work?

Well, if you can access your dataset from a cgi script and can write a cgi script that outputs the elevation data of a certain tile of your data if asked for a specific latitude and longitude, we can talk about it.

E.g. the cgi request url gtopo30?TPL=pwrl/tilegtop&LAT=480000N&LON=0120000E fetches the data of one tile located at 48 North, 12 degrees East.

Seriously, I did this project for fun and I enjoyed it very much, so if you have a data set that promises to show something new I am more than happy to collaborate.

What perspective and projection are you using?

The elevation data is taken from GTOPO30. This data is only cached for processing, no further calculations are done.

How does the seamless browsing work?

When you enter the model the main world with the HUD, the globe, etc. gets loaded. This world inlines 9 sub tiles. The sub tiles are arranged in a square so that the user position is on the middle one of the 9 tiles. If you move off of the middle tile, the 3 tiles that are now far behind you are moved in front of you and new geometry gets loaded into those 3 tiles.

If you move too fast or your internet connection is too slow you will notice that the tiles are not loaded underneath or in front of you. Stop moving and wait a little while in this case.

Where do the colors come from?

The coloring is entirely mine. First I find out whether a given point is on a hill, on a slope or in a valley and use different colors for these three cases. Furthermore the coloring depends on the location on the planet. E.g. Europe is greener than North Africa. But this algorithm needs more work.

What do the yellow and white flags mean?

The data of the place names is derived from Digital Chart of the World. The yellow flags are names of populated places and the white ones are names of airports.

The data of the DCW is at least 10 years old, therefore a lot of names are not correct anymore. If you know of any better data that is available on the net please let me know.

I entered a couple of names of lakes ( blue flags ) and mountains ( white flags ) by hand because I could not find any other source.

If anybody knows about a source that maps names of mountains and lakes to latitude and longitude in degrees and with a resolution of at least 30 arc seconds please let me know.

Why are there no names of places in Russia?

Digital Chart of the World uses a strange longitude/latitude scheme for Russia which I could not figure out. If anybody figures it out and can retrieve the data in the form

Lat    Lon      P Name    Country
480827N0113234E P MUNCHEN DE

I will include it.

What happened to the map shown in the screen shots?

The map was retrieved from the Xerox Map Viewer. This link is no longer active, I took it out.

Are the altitude readings on the HUD correct?

Yes, while you move around the HUD continously displays your current position in degrees and your altitude above sea level in meters.

What is the little red spot on the globe?

The marker shows your current position on the globe, it moves over the globe while you move around.

Why is the search label displayed?

All yellow flags of place names are anchors that start a search on Google with the place name as query string.

Where do the country codes come from?

I have added the country codes to the place names, they are taken from the codes given to the web domains of the countries, e.g. DE for Germany. For the US states I used the 2 letter postal codes, e.g. CA for California.

Can you say more about the inner workings of the program?

The cgi program written in C gets called with a request to print one sub tile: /cgi-bin/gtopo30?TPL=pwrl/tilegtop&LAT=480000N&LON=0120000E

The program reads a 41 by 41 elevation point square with the given location at the upper left corner from the GTOPO30 data. The number 41 is a value that I defined, it leads to a sub tile with 1600 vertexes. 9 of those subtiles are displayed at any given time, my computer can handle the 15000 vertex VRML model resulting but not much more :-).

With 30 arc second grid spacing of GTOPO30 the nine subtiles together exactly display one square degree.

Next the program assigns the color values to the points. Hilltops, slopes and valleys get different color values, the color depends also on the location of the data on earth and on absolute altitude. Lakes and the sea are blue. The algorithm is rather complex, heuristic, unscientific and from me. ;-)

The program retrieves the place names of all places that are located on the tile printed.

Last the program prints a VRML template and fills the heights and color values into an ElevationGrid with colorPerVertex color. The flags for the places located on the tile get also printed to the VRML. The resulting VRML file gets gzip'ed on the fly and then sent on its way to your browser. The gzip'ed sub tile VRMLs are between 5 and 10 KB.

In order to display the entire scene the program gets called with a request to print the main world first.
/cgi-bin/gtopo30?TPL=pwrl/gtopo30&LAT=480000N&LON=0120000E. This uses a different VRML template for the main world. The main world VRML in turn calles the program 9 times to print the 9 subtiles.

The main world VRML also contains the code that loads the tiles needed while you walk around. This code is written in VRML script.

Acknowledgements

Thank you to the USGS - NASA Distributed Active Archive Center for making the GTOPO30 data set available on the net.

Thank you to the Pennsylvania State University for making the Digital Chart of the World data set available on the net.

Thank you to Google for their wonderful search engine.

Thanks to the web 3D consortium for providing the mailing list and the GeoVRML working group.

Thanks to Teresa Reuter for consulting on the color schemes.

Special thanks to Holger Grahn for writing the best VRML browser and for support on VRML questions.

And last not least thanks to Robert G Stein for supplying the music for the project.

copyright c 2001 - 2007 by Peter Graf