This guest post is written by Jeremy Wang, who worked all summer on a web visualisation system for GISTEMP results, thanks to the excellent Google Summer of Code. This is his second post, here is the first.
It is the end of the summer and I am wrapping up my Google Summer of Code project. My experience with the Google Summer of Code, the Climate Code Foundation, and my mentors here (Nick Barnes, Nick Levine, and David Jones) has been very rewarding. My final project is largely what I proposed at the beginning of the summer. CCF MapView is a map-based visualization of the GISTEMP climate analysis via ccc-gistemp. The tools shows a Google Maps-like interface over which various climate data is overlaid. The map supports zooming and panning. Data sets including topography, night light radiance, cities, and weather stations which can be toggled on and off. The gridded temperature values computed by ccc-gistemp can be overlaid on the map by selecting a source – ocean, land, or mixed. The displayed year can also be changed using arrows or the slider.
Weather stations contributing to the GISTEMP analysis are displayed as square dots with orange indicating stations in urban locations, yellow indicating suburban, and green indicating rural. When the mouse hovers over a station, the station name is shown. Upon clicking a station, detailed information about the station pops up, including geographic information about the station location and two types of chart showing the historical temperature record. Charts show the final temperature difference from the baseline period (1950-1980) along with adjustments made based on partial/missing data and urbanization effect. The basic unit in the GISTEMP analysis is a grid of 8000 sub-boxes designed to contain equal area across the globe (although they appear uneven on a map). Each of these grid cells is shown a different color indicating the temperature delta for the selected year. Red indicates warmer than baseline and blue indicates cooler. When the mouse hovers over a cell, the coordinates for that cell are displayed. Upon clicking a cell, detailed information about the cell pops up, including the list and contribution weights of weather stations contributing to that sub-box average. Also shown is a chart of the temperature record (again, relative to the baseline) for the chosen region.
I hope for this tool to help better explain and illustrate the GISTEMP climate analysis procedure and results, especially for non-scientists and those interested in climate change. The climate data is displayed on top of a modular map framework so that it should be relatively easy to extend the same type of visualization to other climate data sets or other types of GIS data. The source can be downloaded at http://code.google.com/p/ccf-mapview.