The techniques described herein have been developed with the IBM Visualization Data Explorer (DX), a general-purpose software package for scientific data visualization and analysis. It employs a data-flow-driven client-server execution model and is currently available on Unix workstations from Sun, Silicon Graphics, Hewlett-Packard, IBM, DEC and Data General . DX provides tools for operating on both scattered and gridded data. The DX Connect module performs the Delauney triangulation used in Figures 5 and 6 while the Regrid module performs the weighted average interpolation used in Figures 8, 9 and 10.
The Regrid module provides independent control of the exponent of the weighting factor, the size of n and the radius of influence from each node of the grid within which to consider data points. In this case a radius of 0.36 degree (of latitude and longitude) was used. It should be noted that for each day of data, not all stations have rainfall measurements. This is NOT the same as a station reporting no rain. DX supports a notion of data invalidity. Hence, for any given day, only those stations having measurement are considered by both the Connect and Regrid modules in creating gridded versions of the data. The choice of modules that support continuous realization is independent of the use of Connect or Regrid even though they result in different mesh structures because these DX operations are polymorphic. These tools were used to create a DX application that supports the study of the rainfall data.
Last modified on February 11, 1999, G.
Scott Owen, email@example.com
Rainfall in Peru - Main Page
HyperVis Table of Contents