To fully appreciate the power of OS OpenData, it is best used within a geographical information system (GIS). You may already be using a commercial GIS such as ESRI, MapInfo or Cadcorp within your business, but if you do not have such systems in place, you may want to consider using an open-source solution.
The use of open-source software and OS OpenData offers the choice between cutting costs and improving delivery and both of these options offer significant bonuses for end users.
There are implications when choosing free software which must be taken into consideration. In the case of a commercial application, technical support can be paid for as part of the software package. However, with OpenSource you are dependent on the growing web community and developers. As a lot of OpenSource software is relatively new, long established commercial software has a well-developed user community which in turn is supported and championed by the provider. This means that with commercial software someone has probably ‘done it before’. In the case of OpenSource you may well be trying these applications for the first time.
Before implementing an open-source GIS, there are a few questions to consider:
- How is the data to be stored, either as flat-file on a local drive or server or in a spatial database?
- Who is the intended audience for the final output?
- What data you hold that is already spatial?
- What data do you need to link to a spatial data set?
- What questions do you need to answer?
- What level of mapping detail you require; for example, consider what scale the data is viewed at and the granularity of the data required, for example, postcode unit vs. full postal address?
- Is there a need to capture and store spatial data?
- Do other users need to access the data, so is there a need to have a shared server or a local installation?
- How are the results going to be published, as static images or as live data on the web?
A typical GIS comprises a number of elements which can be broken down into the following components:
Open-source GIS software infrastructure model
Data input – as data may not be available in a format that can be read by a particular GI application, a translator may be required to read data and convert it into a compatible format.
- Edonica provide a tool for viewing OS OpenData.
Data capture and maintenance – tools are required to enable users to capture and maintain their own data sets.
- Geospatial Data Abstraction Library (GDAL) is a library of translation utilities for use with raster data.
- FWtools is a set of open source GI tools created by Frank Wamerdam for both Windows and Linux.
Data storage – this may be either as simple files in a directory or in a spatially enabled database.
- PostgreSQL is an OpenSource SQL database and has powerful spatial functionality when the PostGIS spatial extension is loaded. Connecting PostgreSQL is supported by a number of GIS including ArcGIS, MapInfo, CadCorp and QGIS.
- MySQL is an Oracle OpenSource data which has some spatial functions but is not yet fully supported by GIS software.
Data analysis – either using tools within a GIS or a spatial database.
- QGIS is probably the most popular open-source desktop GI application, with a range of tools for use in analysing spatial data.
- MapWindow is a GIS created by Idaho University.
- Geographic Resources Analysis Support System referred to as GRASS was originally created by the US Army. It has a wider range functions for spatial analytics.
Publishing mapping – either for viewing in a GIS or via a web browser.
As well as analysing your data with the GIS software mentioned above, these can also be used to publish the results of your analysis in a graphical format.
- Geoserver is an OpenSource server for sharing geospatial data across the web.
- GeoExt uses OpenLayers for the creation of powerful GIS apps.
- Mapserver for publishing spatial data and interactive mapping applications to the web.
It is also worth noting that commercial software will have a well-developed Graphical User Interface (GUI), in the case of some open-source applications, it may well be that unless a developer has seen the need to develop a GUI then none will exist. This could mean that using some OpenSource applications may require some detailed technical knowledge for users to be able to perform tasks which may be considered straightforward.