Web Map Spatial Clustering Strategy

The web map displays a large number of data points, one for each tree in our system, and thus must compute and render aggregate region based totals to ensure good performance at low zoom levels. We have implemented a strategy that assigns clustering regions to each zoom level and uses these to pre-calculate region totals. Within a given zoom level, clustering regions can be assigned a priority, enabling a given region to pre-empt other regions. Each region also allows for the location of its cluster annotation to be specified.

Examples

• A region, for example, could be a continent, a country, a state, or a city. These regions can be assigned to appropriate zoom levels that make sense for the size of the region - thus appropriate aggregate totals of trees are displayed at appropriate scales.

• A good example of pre-emption by priority would be when sub-districts are available for a certain governmental administrative area, but not for another. A zoom level could implement district level totals by default, but when sub-districts are available these can be given a higher priority and yield cluster totals by sub-district in those areas.

Technical Details

Database Schema for Spatial Clustering

Current Cluster Regions