Gaia3D has developed meteorological data mobile web service using PostgresSQL and GeoServer for weather forecaster in Korea Meteorological Administration(KMA).This system displays weather charts, weather prediction information, weather images, observation data on mobile web for rapid decision of weather forecasters when they are out of office or in remote environment. I will deliver a presentation about experience to develop and launch mobile web service showing weather charts by tuning daily updated big spatial data in terms of database.
Weather charts generated by this system are displayed using OpenLayers Mobile after inserting big size vector data into PostGIS and rendering these data by GeoServer. This system processes 67 million lines of spatial data(approximately 35GB) and generates more than five thousands weather charts everyday.
On previous system, it took five hours to insert data into PostGIS and took tens of seconds to publish single weather charts by GeoServer. Also, there was another problem that file size on PostGIS has unlimitedly increased.
Gaia3D decided to fix the problems and improve this system in terms of data input, data management, and data display. Consquently, the performance of data input has increased about hundred times and the performance of data display has increased about two hundred times. Finally, KMA could successfully and stablely manage the system without increase of file size of three days data.(This system shows data up to previous 72 hours.)