FAO - Africover Spatial Aggregation

The public domain spatially aggregated data set retains the full power of the original data set because it is based on the LCCS coding system: you can exploit all the power of the LCCS coding to perform complex queries and you can re-aggregate the data thematically to extract exactly the information you need.

The thematic content of the spatially aggregated data set is very similar to the original data set: the aggregation is performed at a spatial level setting a threshold under which the polygons (with a surface below the threshold) are dissolved in the polygons containing them. The land cover classes of the spatially aggregated public domain data base are the same of the ones contained in the full resolution one but are characterized by a reduced level of spatial detail.

Under the thematic view point only the classes assigned to units (polygons) characterized by a maximum surface that is minor than the aggregation threshold for the specific group are eliminated. The spatial aggregation and its implications on the thematic contents of the data base depend on the cartographic standards adopted for each country.

The spatial aggregation of the country data bases is performed identifying an area threshold for each of the 8 major LCCS land cover classes:

1) Cultivated and managed terrestrial areas
2) Natural or semi natural terrestrial vegetation
3) Natural or semi natural aquatic vegetation
4) Cultivated aquatic or regularly flooded areas
5) Artificial surfaces and associates area(s)
6) Bare area(s)
7) Natural and artificial water bodies
8) Snow and ice

All the units (polygons) below the threshold are dissolved in the unit (polygon) in which they are contained.

This process is fully documented so that the user can exactly determine the information content of the spatially aggregated data base in relation to the full resolution one. You can analyze the table comparing the characteristics of full resolution and spatially aggregated national databases and further analyze the information content of each data base in the country specific pages.

Starting from this data set you can query it or re-aggregate it thematically or spatially depending on the specific information needs (e.g. rangeland, forest etc.).

To understand the characteristics of the MADE read the inside MADE or visit the technical insight section.

You can refer to the procedure used to aggregate the full resolution data base to understand how to produce your queries and aggregations.

Thematic aggregations: customizing the data to answer your needs

Thematic aggregation is the way that the end user customizes the Africover database to fulfil his/her specific requirements.

The Africover database gives equal level of detail to Agriculture as well as Natural vegetation or Bare Areas etc.

Generally a single user does not need this level of detail for each class type, therefore he will enhance the information of one land cover type and will generalize or erase the information related to other land cover aspects.

The most powerful way to conduct an aggregation exercise is to use the classifiers as basic elements of the exercise. This gives the user the maximum flexibility on the use of data.

The aggregation procedures follows three main conceptual phases:

1. Identification of the classifiers needed for the data customization. In the first example for Tanzania the user wants to know the location of “open to closed tree and scrub (woody) natural vegetation” in the country.

In annex 1 the all classifiers needed are shown. To check and understand the meaning of each of them the user can refers to the “Land Cover classifiers list”, “Legend” and “LCCS glossary” attached to each country data set.

2. Identification of the thematic classes containing the selected classifiers. In annex 1 is shown an automatic procedure using ArcView software and Africover extensions. However the identification can be done manually using the “Legend” file of each country.

3. Creation of the aggregated classes taking into accounts the Africover cartographic standards. In the Africover database due to the MMA (Minimum Mappable Area) chosen, the concept of mixed unit and the inherent characteristic of the study area. Land cover class A can be spatially represented in different ways:

• As single map unit A
• As mixed map unit were A is the dominant feature (more than 50% of polygon area) A/B
• As mixed map unit were A is not the dominant feature (from 20 to 49% of polygon area) B/A
• As mixed map unit were A is not the dominant feature (from 10 to 20% of polygon area) B/A (this is valid only for “Isolated Agricultural Fields”)

Due to the fact in Africover a mixed unit can have up to three classes A/B/C an aggregation class (called 1) can be represented in four (five for agriculture) different ways:

• 1 (were 100% of polygon area represent the aggregation class)
• 1a (60% app.)
• 1b(40% app.)
• 1c(20-30% app.)
• 1d(15% app. Only for agriculture)

The user can further aggregate these classes according to his/her needs.

In annex 1 it is shown a semi-automatic procedure of aggregation using ArcView software and the related Africover extensions.

Annex 2 displays the table of aggregation that depicts which map-codes were assigned to which aggregation class.

See the spatial aggregation thresholds used

The table compares the spatial information content before and after the aggregation. Detailed data for each country are also available.

The mixed units

Due to the cartographic standards applied in the interpretation, if a feature is smaller than the minimum mappable area applied, in the interpretation process, a mixed unit can be developed by the combination of a maximum of three land cover classes; the first class covers always more than 50% of the unit (i.e. the polygon) while the second class must cover at least 20% of the surface.

Different classes in one unit (polygon) are separated by the “/” character. A class can also have a subclass, which normally indicates the crop type. Subclasses are preceded by a hyphen.

Producing your own aggregations

The ADG software is under development to provide an easy way to create you custom aggregation interactively. In the meanwhile you can produce your own aggregations of the spatially aggregated public domain data on the base of procedure provided. Africover developed also an Arcview extension that can be used to simplify the execution of aggregation.

FAO Africover also provides some thematic aggregations designed to answer most user needs; you can access them from the spatially aggregated data bases metadata.

Download the guide to perform thematic aggregations

This guide is derived from the aggregations developed for Tanzania:

Introduction: understanding the procedure (printable PDF version of the paragraph on the left)

Annex 1: the procedure in detail (PDF 236 Kb)

Annex 2: the aggregation tables (XLS 56 Kb)