The release last year of NASA's World Wind global product in many ways marked the end of an era in the world of remote sensing and global visualization. The release of Google's Google Earth product puts an emphatic exclamation point on the earlier NASA statement

When I started this site in 2000, the availability of global DEM data and particularly remotely sensed satellite imaging was virtually non-existent. The site was started in order to make the limited USGS ASCII DEM and SDTS DEM data more accessible to the non-traditional, non-academic user. This required a lot of custom programming and the use of a variety of software applications, each of which did part of the job that in the end produced a usable if less than perfect result.

The free Google Earth software application produces the highest yet quality DEM overlay visualizations of the earth's surface at virtually any point o the earth's landmass. Moreover, it does so using an exceptionally well designed and intuitive user interface. In fact you don't even have to consult the 'Help' menu to construct terrain visualizations that would have required weeks of work to produce just a couple of years ago.

Google Earth is very similar to World Wind in its presentation to the user. A globe rendering appears on the screen upon startup. The user manipulates the globe using the mouse, and then zooms into the area of interest.

Three features make Google Earth superior to World Wind. The Keyhole-developed transitioning from low to high-resolution image presentation is even better than the sophisticated MRSID technology developed by Lizardtech (which is excellent). Unlike World Wind, there is virtually no refreshing as higher resolution satellite image overlays are applied to the rendering. Transitions are almost as fast as you can zoom your mouse.

The maximum resolution for the remotely sensed images far surpasses that of World Wind for many areas. Google Earth benefits from the inclusion of much Digital Globe Quickbird imagery, which is of considerably higher resolution than the Landsat imagery used by World Wind. Although the underlying DEM data is the same (SRTM) the overlaying imagery is worlds better (no pun intended). As pointed out in previous articles, the resolution of the overlay can improve the apparent resolution of the overlay considerably.

Finally, Google Earth does not suffer from the huge bug in World Wind that causes intolerable screen jitter in many cases when maximum image resolution is invoked. In addition unlike World Wind, the file save utility of Google Earth actually works so that images can be saved in the conventional manner without resorting to screen shots.

The attached images show some Google Earth images of one of the most famous (or infamous) locations in the world, the southern Manhattan former site of the Word Trade Centers. The location is shown at various levels of zoom. The higher levels are of significant interest as the resolution of these Digital Globe images (probably 0.8m) is kind of scary. At the highest level, it is actually possible to see individual people walking on the site (although even at this resolution not every individual on the site may be resolved.) As pointed out in a previous article, the position of the sun whereby it throws long shadows helps in the apparent resolution of small objects, as often their shadows are resolved even if the objects themselves are not.

Such high resolution is not available over most of the Google earth. Much of the earth is apparently rendered at Landsat resolution of 30m. Only global population centers are rendered at high resolution, the even the somewhat limited high resolution coverage is astounding in its quality.

The final shot shows one of the "trick" features of some of the Google urban images, whereby buildings are rendered in 3D. I am not sure where the source data comes from. It may be from high resolution Digital Globe stereo pairs or perhaps from separate aerial Lidar scans.

Google Earth marks the end of an era and makes much of the technology described on this page obsolete and irrelevant. However, new challenges remain. The most important of which include access to DEM data with greater than 90m resolution. Future work on this site will focus on this problem, among others of interest to the author.