Version 12 (modified by mhenning, 14 years ago) (diff)


NDG Progress, as of October 2006 (mhen)


  • The PML MOLES Database has been redesigned. Triggers and postgres-specific extensions (such as arrays) have been removed and relationships have been made explicit. This should make it (slightly) easier to understand, and also more portable. A graphical representation of the new database structure is attached.

  • The database can now be fully populated with data entities, by using a few perl scripts and unix tools.

mhen@pmpc828 ndg_moles # ./ modis meris avhrr seawifs | sort | uniq | ./ | ./

Currently, the database has one MOLES record for each year of data for a particular sensor and area:

mhen=# SELECT production_tool.description AS Sensor, COUNT(data_entity.*) AS RecordCount 
FROM ndg.data_entity, ndg.data_entity_deployment, ndg.production_tool, ndg.deployment 
WHERE deployment.production_tool_id = production_tool.production_tool_id 
AND data_entity_deployment.data_entity_id = data_entity.data_entity_id 
AND data_entity_deployment.deployment_id = deployment.deployment_id 
GROUP BY(production_tool.description);
 sensor  | recordcount 
 AVHRR   |         617
 MERIS   |          32
 MODIS   |          81
 SeaWiFS |         260
(4 rows)


(Total 1054)


  • CSML records can now be generated, by using a combination of perl and python scripts:

mhen@pmpc828 ndg_csml_gen # ./ndg_csml_granule meris200607zx | --format NDG_CSML | ./

The CSML record generated by the above scripts can be seen here:  meris200607zx