Vishful thinking…

A Wind power prototype – using Google maps, Google charts and ExtJS

Posted in Uncategorized by viswaug on June 1, 2008


I just finished up the above prototype application that displays the installed wind power capacity by state in the US over the years. The application is built on top of the ExtJS JavaScript framework to provide layout functionality with collapsible left and bottom panels. I was going for a close-to-ArcMAP kind of look for the application. So, the left-side panels contains the legend information and some related links on another sub-tab. The attribute table data appears in a grid that is docked to the bottom. The grid allows the user to sort and hide columns as required. Both the left and the bottom panels can be completely collapsed to allow for maximum screen real-estate for the map. The possibilities with the ExtJS framework can be much more powerful and even more user-friendly.


Google maps provides the mapping for the prototype. The state polygons are overlayed on the maps and are symboloized based on the year selected by the user and the installed MW capacity for the selected state. The user can also click on a state on the map or on the grid to see a bar chart (generated by Google charts) that displays the year over year growth in the installed MW capacity for the selected state. The current year’s value is highlighted in red on the bar chart.


The state polygon and the attribute data is delivered by the server in a GeoJSON format making it easy to interpret. Please take a look at the GeoJSON.NET library if you are interested in generating GeoJSON yourselves. I will also soon be posting about a GIS data framework that I have been working on and that will make generating GeoJSON even easier for various other data formats like FeatureClasses and SQL Server 2008 spatial.

Over all the development of the prototype was pretty smooth. The task that took some time was the reducing the state boundaries shapefile from National Atlas into shapes with a reasonable number of points so that it can be displayed in Google maps without too much of a performance hit (to download and to draw) and still looks good in a cartographer’s eyes. I tried ESRI’s “Generalize” tool on the state shapes, only for it to pop-up a message box saying that the geometry is empty. SQL Server 2008’s “Reduce” function also didn’t get the job done on the shapes. I tried‘s map reduction functionality and it got the job done on the shapefile. But not everyone was happy with the cartographic quality of the generated shapes. Finally, one of my colleague took the pains to edit the polygons manually to get a better cartographic prototype (kudos for Chris French for doing that).

I also hope to posting about “Google Maps vs VirtualEarth” from purely a developer’s perspective pretty soon.