GEOSS Banner

Bottom up aggregation of land cover data

Table of contents
No headers
Subject area/Theme

Data and Architecture/Governance

Best Practice:

Land cover data should be collected on subsidiary political levels (local, regional, for small countries national) and aggregated upwards to higher political levels. This upwards aggregation involves a generalisation and reduction of details that are circumstantial for the higher political levels.

Explain why is there a need for this Best Practice?

Users operating on the highest political levels (EU-wide, US-national, continental, global...) typically require generalised data products that provide an overview, whereas users on subsidiary political levels require more detail. Data products that are produced for higher political levels can hardly be adapted to the needs of subsidiary political levels.

Provide an example application(s):

Several European countries have initiated their own land cover systems. Some of them apply the European product CORINE** land cover as a standard to which their national data are then compatible. The advantage is two-fold: Their mapping matches that of neighbouring countries, so that environmental issues can be compared and tracked cross-boarder; and their participating in European level mapping is facilitated because they can fall back on efforts already done, rather then duplicating work by applying firstly national and then, in a second mapping campaign, European standards.

)** Coordination of information on the environment, http://reports.eea.europa.eu/search_...chTitle=CORINE

How widely deployed is this practice(if applicable)

One example is described above.

Owner (Originator) Contact Information:

GNU.jpg

Submitter Contact Information:

GNU - http://www.gmes-network-of-users.eu

Detailed Description of Best Practice

The smaller the area is to which particular environmental data refer, the higher the degree of detail required. GMES* currently applies a down-stream model. Gor land related applications, this implies that data products conceived for the European level will be produced by a core service. Subsequently, downstream services will  adapt or further develop these data products so they meet the needs of entities that operate on subsidiary political levels.

The core service data products provide the overview, European level users need. However, although useful on the European level, these data products typically do not meet the accuracy (general and thematic) needed by subsidiary level users. In the down-stream model, such rather general products would have to be refined, to make their level of details and their accuracy increase, which is rarely possible. A bottom up model appears to be more practical: Data are firstly collected on subsidiary levels – with the degree of details the related user communities need – and then aggregated upwards, thereby being generalised so they meet the needs of users of higher political levels. The main task for co-ordinating this effort would be the standardisation that enables both the compatibility of data collected in various areas and the upwards aggregation process. This coordination and aggregation could then be among the major functions of the related GMES* core service.

The details of a land cover related bottom-up approach still must be worked out in a participatory process. Firstly, there is still a considerable amount of research needed to solve compatibility issues. The most important open question is, where the "bottom" should be. Eg, it may work well to upwards aggregate European national data, but then local users may again consider this a top-down approach. Moreover, how costs can be shared between the various political levels must yet be resolved.

)* Global Monitoring for Environment and Security, the European contribution to GEO

Files (0)

 
You must login to post a comment.