![]() ![]() ![]() |
![]() |
|
|
![]() ![]() ![]() ![]() ![]() |
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Return to Advanced Models and Languages/Architectures for Data Analysis (Session D3) Earth observation (EO) and simulation data share some core characteristics: they resemble raster data of some spatio-temporal dimensionality; the complete objects are extremely large, well into Tera- and Petabyte volumes; data generation and retrieval follow very different access patterns. EO time series additionally share that acquisi- tion/generation happens in time slices. The central standardization body for geo service interfaces is the Open GIS Consortium (OGC). Earlier OGC has issued the Web Map Service (WMS) Interface Specification which addresses 2-D (raster and vector) maps. This year, the Web Coverage Service (WCS) Specification has been added with specific focus on 2-D and 3-D rasters ("coverages"). In this paper we present operational applications offering WMS/WCS services: a 2-D ortho photo maintained by the Bavarian Mapping Agency and a 3-D satellite time series deployed by the German Aerospace Association. All are based on the rasdaman array middleware which extends relational DBMSs with storage and retrieval capabilities for extremely large multidimensional arrays. ![]() ©2004 Association for Computing Machinery |