This talk is about reexamining old methodology and designing new graphics templates. In many cases the deficiencies of the old methodology are well known. For example Tufte [1983] shreds a 3-D bar plot noting multiple flaws in the particular example. Cleveland and McGill [1984] persuasively demonstrate that dot plots are superior to stacked bars in terms of perceptual accuracy of extraction. While scientific and federal agency authors should be interested in simple graphical encodings with high perceptual accuracy of extraction, habit and the click option ease of spreadsheet graphics tends to carry the day. Have you ever used a dot plot or seen one in a government publication?

This strain of collaborative graphics research began in 1993 in the attempt bring Cleveland’s dot plots to EPA. The research began with designing more aesthetic dot plots and rapidly evolved toward converting visually intimidating tables into apparently simple graphs. Research for BLS included developing a sequel to the old BLS TPL (table production language). A GMU seminar last year covered the resulting JAVA-based Graph Production Class Library (GPL) for distribution of summaries and metadata on the web. The next step in this strain of research focused on linking summaries to maps. This is timely as the federal information services and their GIS shops gear up to communicate a wealth of information to the public. Recent research of others considers how NASA might store and display level 3 satellite products soon to result from the Earth Observing System. This talk presents extensions for global binning and smoothing.

The talk has many graphics examples called row-labeled plots, linked micromap (LM) plots and Icosahedral Snyder Equal Area (ISEA) global grids. Comments about perceptual and cognitive principles motivate the designs. One highlight is a 4 x 8 foot poster relating continental U.S. ecoregions to digital elevation data and an AVHRR-based 8 million pixel data set of vegetation class information. One constituent LM plot represents percents for all 159 vegetation classes for each of the level 3 ecoregions. Another highlight is the resolution 9 (200,000 cells) ISEA plot of elevation data for the globe. Still the simple appear row-labeled plots may draw the most attention because of direct applicability to many scholar publications.