Visualizing gridded datasets with large number of missing values

Much of the research in scientific visualization has focused on complete sets of gridded data. The paper presents our experience dealing with gridded data sets with a large number of missing or invalid data, and some of our experiments in addressing the shortcomings of standard off-the-shelf visualization algorithms. In particular, we discuss the options in modifying known algorithms to adjust to the specifics of sparse datasets, and provide a new technique to smooth out the side-effects of the operations. We apply our findings to data acquired from NEXRAD (NEXt generation RADars) weather radars, which usually have no more than 3 to 4 percent of all possible cell points filled.