Navigation path

Presenting geographic data in Overview maps

In order to present maps containing continuous phenomena such as vegetation absorption or rainfall statistics we need to take care of the cartographic knowledge on presenting 'just' enough data to the user. This implies that we do not present so-called 'high resolution' data at a continental map of the size of a common computer screen. We choose to group data together at cells of 5 sizes, often referred to as pyramids. This allows the user to encounter visually attractive maps at continental or country scale and high resolution when being zoomed in. The values and legends at small scales are therefore to be seen as indicative, it is helping to locate an area of interest. Currently we present 'averages' from the cells grouped in the cell being visualized. We might choose for specific maps other mathematical rules, especially if that would help the user to locate quickly an area of interest. Currently we implemented 'pyramids' for the themes called 'Soil Moisture', Fapar and Fapar Anomaly' and SPI Blended and Interpolated. The latter start only at a resolution of 1/4 by 1/4 decimal degree.

PYRAMID NAME: general
GRIDCELL DIMENSION [DD]MIN SCALEMAX SCALEFRACTION OF A CELL
OF FIRST GRID
NUMBER OF CELLS
IN A CELL OF FIRST GRID
FRACTION OF A CELL
OF PREVIOUS GRID
NUMBER OF CELLS
IN A CELL OF PREVIOUS GRID
GRID_1DD11:1000000001:70000001 / 1 = 11--
GRID_05DD0.51:70000001:30000001 / 2 = 0.541 / 2 = 0.54
GRID_025DD0.251:30000001:15000001 / 4 = 0.25161 / 2 = 0.54
GRID_083DD0.08331:15000001:7500001 / 12 = 0.08331441 / 3 = 0.33339
GRID_0416DD0.04171:7500001:11 / 24 = 0.04175761 / 2 = 0.54

Performance issues
Displaying maps using the EDO database is not an impulsive task. The EDO database contains archives using high resolution data for whole Europe containing time-series up to 50 years. Without noticing you might be consulting database tables containing 60 million records.
We do our best to give a user optimal performance but our service is not comparable with commercially run services. You will for example notice that a map when displayed the first time can be slow, up to 20 seconds to generate. If you subsequently change layer but not change the zoom area your next map will be faster thanks to a caching system that we use.
Do not hesitate to give up constructive comments regarding the service and performance.
For more information or to leave feedback
please contact Alfred de Jager