Linked Micromap Plots for Evaluating Trends in Multi-Pollutant Deposition
Robert R. Gillies1, Jürgen Symanzik2, Esmaiel Malek1 and Alan Moller1
The National Atmospheric Deposition Program (NADP) traditionally provides multi-pollutant data summaries as color choropleth maps. Linked Micromap (LM) plots, however, provide an alternate arrangement to displaying such statistical information that avoids some of the restrictions of choropleth maps. LM plots were first introduced at the Joint Statistical Meetings 1996 in Chicago. Since then, they have been used for various applications. Interactive, Web-based versions of micromaps are also available on-line. Similar to the micromaps accessible on Web sites, micromaps compiled for NADP data interpretation consist of three to four columns: The first (leftmost) column shows a series of small maps. Different sub-regions (here the states of the US) are highlighted on each map. The next column shows the identifiers of the sub-regions (here the state names). Typically, one or two columns follow with statistical data (here different air pollutants such as NO3 and NH4). All columns are sorted according to one of these data columns, say NO3. The columns are linked by color so that the same color is used to highlight a state, the state name, and the color used in the statistical columns representing this state. Then, the top map highlights the five states with the highest NO3 values; the second to top map highlights the five states with the next highest NO3 values, and so on. By highlighting a few states in a particular map, strong geographical relationships were detected. Moreover, by emphasizing a median state, it becomes possible to easily identify a subset of states with similar NO3 concentrations and to determine whether these states form a geographic pattern. For NO3 (in 2004), such patterns were spotted: There is a band of states with high NO3 values stretching from the Southwest (Arizona) to the Northeast (New Hampshire). On the other hand, the Northwest and the Gulf Coast form additional strong geographic patterns. Additional data columns (here NH4) show the associations between statistical variables and immediately highlight geographic outliers. New Jersey, although having the maximum NO3 value, only had a very low NH4 value in 2004.
1 Utah Climate Center, Utah State University, Department of Plants, Soils and Climate, 4280 Old Main Hill, Logan, UT 84322-4280, USA. (435)797-2190 , ,
2 Utah State University, Department of Mathematics and Statistics, 3900 Old Main Hill, Logan, UT 84322-3900, USA. (435)797-0696