STAO: a component architecture for raster and time series modeling

Implementation of spatially distributed models using Geographic Information System (GIS) involves accessing and processing of spatial and time series data. Most of the GIS software do not provide an explicit functionality to support handling of time series data in modeling and also requires exclusive client storage of one-dimensional array for raster data access. This paper describes an attempt to overcome such short comings by developing Spatial and Time series data Access Objects (STAO) using Microsoft's Component Object Model (COM (R)). STAO has advanced enough modeling objects which are developed for the raster and time series data access and analysis while they are simple enough for the user to implement them in spatial models in a variety of native proprietary formats. The STAO architecture defined in this study has multiple advantages in terms of interoperability, reusability, and extendibility. The usefulness of these objects has been demonstrated by implementing: (1) the USDA's Soil Conservation Service (SCS) method of rainfall runoff estimation and (2) an algorithm for finding time area diagram at sub-watershed level. The results of implementing SCS using STAO in Kolli Hills watershed is discussed and the computed runoff values are in good agreement with field observed values. (c) 2004 Elsevier Ltd. All rights reserved.