“Seeing” large data sets Reanalyses Systems generate huge amounts of data that cannot be absorbed by the human mind without further refinement and interpretation. One tool that helps to make the data more accessible, understandable and usable is data “visualisation”. Users generate a query describing the data elements they want displayed, the time period they want covered and the visual elements they want used (eg. colour, image size, etc.). A graphic is then presented, giving the user a clear visual representation of otherwise complex data. The graphic can be a still image or an animation, for instance:

Tropopause 1000mb Zonal Wind (m/s) composite mean From 20CR Reanalysis system

Air temperature (degK) composite winds October 1899 to Oct 1900

The Weather of October 1918 (20CR system)

NASA has plotted GEOS5 aerosols:

A time series visualisation of weather observation data contained in the International Surface Pressure Databank

Movements of the 250 ships whose logbooks were digitised by the oldWeather project (1914-1923)

A visualisation of the impact that observations contributed by ACRE partners had on the 20CR reconstruction of the weather of 1916. It shows: • earlier contributions made by ACRE partners and those more recently made. • The “fog of ignorance” showing global regions where the reconstructions are still very uncertain because there are too few observations. • The “glow of discovery” showing regions where the new observations have produced a large improvement in the confidence of the reconstructions.

NCAR has plotted an East Asia coastal cyclone