Igol version 3

6.Data Delivery

A grand challenge for land observation systems is to make land observation products and science relevant and accessible to a virtually unlimited potential user community. This must be done in ways that respect the policies of data originators and ensures confidentiality and sensitivity where appropriate. Open and unrestricted access, to the greatest extent possible, should be the ambition of data delivery systems for land observations.

6.1.Data and product access

The ideal access strategy for land observations involves unrestricted and no-cost access to all data and products: all data providers are urged to establish open access data policies. For clarification, such a data delivery system uses a comprehensive definition of data and products. These not only include observation, in situ, survey, and other geospatial data sets, but also information products, reports, assessments, metadata, and other forms of documentation.

Because land characteristics and processes are fundamentally place-based and collectively represent the geographic variability of the Earth, all archived land observation data will require some level of geocoding in order to be gathered, combined, and used in environmental models. Standards for geocoding, and implementation of appropriate data integration and analysis capabilities should be an immediate priority so that the overall system capabilities support the necessary functions. The efforts of the Open Geospatial Consortium (OGC) should be considered in the establishment of land observation data system geospatial analytical capabilities.

6.1.1.Data access policies

Data access is limited by data use restrictions and charging for data. Access to land observation data can be restricted by copyright, privacy, or national security or interest issues. For example, the highest quality global digital elevation model data are classified, and access is restricted. Higher resolution data are necessary for activities like hydrological modeling and land use planning. Strong encouragement and support should be given to efforts to lower barriers to access either by wholly enabling new access to existing data or developing data access systems that enable retention of some modicum of spatial uncertainty, anonymity, or recognition yet still maintain copyright, privacy, or security standards.

Socio-economic and in situ inventory data are often restricted due to confidentiality concerns relating to individual privacy concerns. Biodiversity data are often controlled due to either perceived or real threats to the survival of endangered species. In both cases, data aggregation or synthesis capabilities are necessary to respect the confidentiality of all data with real restrictions.

Data pricing policies create significant barriers for many types of observations and applications. Costs certainly restrict access to data. Like lowering other access barriers, lowering costs would increase use of land observation data. In many cases costs are limited to those associated with distribution of a copy of the data to a user. Such minimal costs may be appropriate but still can limit access. In cases where costs are higher, policies that promote access to the widest variety of users are to be preferred. Data users are bound to national data pricing policies, but should negotiate with programs to make costs as low as possible. Implementation of the recommendations offered by GEOSS to make data as inexpensive as possible or at the marginal cost of reproduction is of high importance. Examples of the right directions for appropriate data cost policies are NASA’s, which makes most of its products freely available to all users and ESA’s which, since 1999, provides worldwide access to its data for science and applications development at the cost of reproduction.

The impact of eliminating charging for products can be seen in the cases of Landsat Geocover and CBERS products. For both products the elimination of charging has led to more than a magnitude increase in downloads by users.

It is unreasonable to expect data generators to provide no-cost customization of data sets for special purposes of single customers. Payment for custom product generation is appropriate in any setting. However, costs associated with post-production processing – like extraction, subsampling, re-formatting, etc. – can be reduced through advancements in visualization software and creation of more refined products. Efforts to improve processing software development and to generate easily manipulated land data products should be supported.

6.1.2.Data documentation policies

Several of the challenges to land observation data access arise from the vast number of diverse observations collected at different scales and times. Data collection, authentication, calibration, and quality assurance are necessary, but can be time-consuming and can significantly limit or delay access. For satellite remotely sensed data, efforts to generate validated products have greatly enhanced data access. On the other hand, many attributes cannot be derived from remote sensing and must be collected in situ. In situ data collection is more labor intensive and geographically restricted, and seemingly similar in situ data may be collected for different purposes in different places, may be a-spatial, and are likely to be less uniform in space and time. Production of consistent, comprehensive, validated data products from in situ data is a challenge and likely a strong limitation to data accessibility and utility, potentially precluding some applications. Documentation and metadata are the necessary foundations upon which methods and software to integrate disparate data can be built. Efforts should be supported to assemble data from different sources, develop sound metadata, catalog using these metadata, and collaboratively develop interoperable data products. We strongly support efforts to standardize, complete and clarify metadata documentation of all processing steps and associated estimates of uncertainty, which will enable broad, accurate, efficient use of land observations.

The growing need to address environmental and societal concerns demonstrates the need to minimize barriers to land observation data access and use. It is clear that lowering access barriers, including costs will encourage collaboration and exchange, advancing development of data processing and enhancing data utility; thus we recommend that costs be kept to a minimum and access is open to all users.

6.1.3.Principal recommendations

• 
  Encourage lowering barriers to access either by wholly enabling new access to exiting data or developing data access systems that enable access while maintaining privacy and security standards.
  
• 
  Encourage implementation of GEOSS recommendations to make data available at the marginal cost of reproduction.
  
• 
  Support research to develop methods and software to lower the cost of post-processing data customization.