The procedure for the establishment of a typology, classification and characterization of farms presented in this study is implemented in different phases and is based on automated analysis of the data contained in the Agricultural Census. Such a methodology enables the user to group at the municipal scale all the farms that show the same production pattern within a size class that has been previously established. The spatial distribution of each reference pattern is determined based on these groups of similar elements. The determination of the spatial distribution of farm types is essential in order to choose a reduced number of farms that are representative of each type and provide updated information about the production process (operations, raw materials, labour) and structure (size, machinery, facilities, land, etc.) of the farms. Such information is useful for characterizing the production patterns (type-class combinations) that are representative of each activity. Such production patterns complete the information of the agricultural production planning model developed.
The analysis model applied must not be based on average statistical values of farms that include different production situations in a single value, but on the analysis of each of the production models that are differential and significant. Such an analysis enables a more objective comparison of the results expected for different land units, and of different reference land uses or activities that involve differential production processes and structures.
This methodology can be applied to agricultural production planning and to other decision-making processes (political, administrative, financial, etc.) that require the analysis of alternatives, risks or probable results for farms.
The procedure developed in this study for the establishment of a typology, classification and characterization of farms used the data from the Agricultural Census of 1999 that corresponded to the region of Galicia. This procedure can be adapted to other regions and countries by making slight modifications in the structure of the software application used. Moreover, this procedure can be used with previous or future censuses in order to study and model the evolution of certain elements.