A Quantitative Fuzzy Approach to Assess Mapped Vegetation Classifications for Ecological Applications

Ecologists frequently employ classifications derived from satellite imagery and aerial photography to interpret patterns of the composition and distribution of vegetation. Such maps can be used in conjunction with ecological data to extend locally intensive analyses to the landscape and watershed scale and beyond. In many cases, however, the field data used to produce and validate map classifications are not designed for ecological applications. The objective of this study is to present the fuzzy similarity index assessment method for evaluating the accuracy of maps of natural vegetation based upon methods in plant ecology. Using an example classification and associated data from field plots, a similarity index is employed to compare validation data against the defined composition for each map class, thereby producing a quantitative measure of the similarity of each observation to all classes. Fuzzy membership among classes is quantitatively evaluated using the differences in similarity between validation plots and all classes in the classification. The nature of the gradients in vegetation composition among the map classes can therefore be identified. Consequently, some classes are interpreted as compositionally distinct while others may share traits with many classes. Additional analyses determined the detail of field data collection necessary to produce consistent results using the fuzzy similarity index assessment method. Complete inventories of the presence and abundance of species are not required; rather, a field sampling method biased toward the 12 to 20 most important species in a given class is acceptable.

[1]  R. G. Oderwald,et al.  Assessing Landsat classification accuracy using discrete multivariate analysis statistical techniques. , 1983 .

[2]  M. Hill,et al.  Data analysis in community and landscape ecology , 1987 .

[3]  James D. Barton,et al.  Field Efficiencies of Forest Sampling Methods , 1958 .

[4]  J. R. Jensen,et al.  Subpixel classification of Bald Cypress and Tupelo Gum trees in thematic mapper imagery , 1997 .

[5]  G. M. Foody.  The Continuum of Classification Fuzziness in Thematic Mapping , 1999 .

[6]  Brian Everitt,et al.  Cluster analysis , 1974 .

[7]  T. M. Smith,et al.  A new model for the continuum concept , 1989 .

[8]  H. Gleason.  The individualistic concept of the plant association , 1926 .

[9]  C. Woodcock,et al.  Theory and methods for accuracy assessment of thematic maps using fuzzy sets , 1994 .

[10]  John R. G. Townshend,et al.  Terrain Analysis and Remote Sensing , 1981 .

[11]  Russell G. Congalton,et al.  A review of assessing the accuracy of classifications of remotely sensed data , 1991 .

[12]  J. T. Curtis,et al.  An Ordination of the Upland Forest Communities of Southern Wisconsin , 1957 .

[13]  Fangju Wang,et al.  Fuzzy supervised classification of remote sensing images , 1990 .

[14]  R. Mccoy,et al.  Mapping Desert Shrub Rangeland Using Spectral Unmixing and Modeling Spectral Mixtures with TM Data , 1997 .

[15]  G. Foody.  A fuzzy sets approach to the representation of vegetation continua from remotely sensed data : an example from lowland health , 1992 .

[16]  M. Barbour,et al.  Terrestrial Plant Ecology , 1981 .

[17]  G. H. Rosenfield,et al.  A coefficient of agreement as a measure of thematic classification accuracy. , 1986 .

[18]  J. Settle,et al.  Linear mixing and the estimation of ground cover proportions , 1993 .

[19]  Giles M. Foody,et al.  Fuzzy modelling of vegetation from remotely sensed imagery , 1996 .

[20]  Alan R. Gillespie,et al.  Vegetation in deserts. I - A regional measure of abundance from multispectral images. II - Environmental influences on regional abundance , 1990 .