Aggregation issues of spatial information in environmental research
Political decisions concerning the environment are frequently based on aggregated versions of scientific results, obtained by post-processing . Decision makers generally use such results, without the luxury of questioning them. GIS experts often act as intermediary between decision makers and scientists. To support that role, guidelines are needed to gain insight into the aggregation issues involved in presenting spatial information. This paper provides such guidelines. The procedure focuses on aggregation prior to modelling (approach A) or afterwards (approach B) and considers two major aggregation techniques (mean and worst case). Some simple indicators are used to quantify the differences between maps. The procedure is tested with an actual case of agricultural pesticide risk, which demonstrates its use. It proved very helpful to improve on the insight needed. It showed the complexity of the issue and the sensitivity to level and method of aggregation, as well as to classification and model type.