The term real-time Web GIS can refer to different kinds of applications. It can be understood as a system visualizing in real-time measured sensor data, like GPS positions or in real-time generated data by human editors, e.g., in collaborative disaster management systems. In the scope of this paper it refers to an almost immediate visualization of GIS analysis results in order to provide a more interactive and responsive user experience, irrespective of the creation time of the input data. Reducing waiting times in web applications is a general aim. Besides that, a real-time linkage between user behavior and analysis results can also be used to enable visual data exploration in data mining processes. Two aspects of the presented approach contribute to the reduction of time between user request and presentation of the results. Spatial analysis execution in the web browser and doing this in an accelerated way using parallel computation on the graphics processing unit (GPU) instead of classic serial computation on the central processing unit (CPU).
In traditional web based GIS applications it is common to use a client-server architecture where the computational analysis part is done on the server side while the web client is just used as a user interface for requesting those computations and to display the processed results. The reason for this is the lack of graphic manipulation capabilities of current web standards. This restriction will change with future web standards and browser releases.
Currently ongoing developments in web technology, like HTML5 (Hickson, 2012; WHATWG, 2012) and WebGL (Khronos Group, 2011a), offer great potential for the future development of more interactive, responsive, efficient and mobile Web GIS applications. This includes the usage of 2D, 3D and even temporal and animated content without the need of any third party plugins.
The following sections provide an overview on related work about parallel computation approaches of GIS analysis. A short introduction on the basics of 3D rendering and a detailed description how to use WebGL for the proposed method follows. Issues concerning the conversion of geospatial raster data to web compatible formats are addressed. An example hillshade algorithm is described and used to demonstrate and measure the performance of the approach. Furthermore some limitations of WebGL based spatial raster processing are discussed.