Project Methods

Assessment of Classification Accuracy

Introduction

It is invalid to assume that habitat maps generated from photointerpretation of remotely sensed imagery are 100% correct. It is important that the mapmaker know how reliably a given habitat can be classified. This parameter is called "producers accuracy". The users of a map product want to know what percentage of the polygons of a particular class are correctly attributed. This parameter is called "users accuracy". Furthermore, remotely sensed imagery that may be suitable for mapping coral reef habitats can be acquired from a wide variety of platforms and imaging systems, each having it's own strengths and limitations. It is important to identify the technical merits of each including the thematic accuracy of the map products.

Benthic habitat maps in ArcView GIS format were prepared by visually interpreting the three types of remotely sensed imagery. Conventional methods for assessing the accuracy of living resource maps prepared from visual interpretation of remotely sensed imagery were applied to meet two objectives. The first was to identify and compare thematic accuracy of maps prepared from the three sources of remotely sensed imagery and the second was to generate a scientifically sound statistical estimate of the thematic accuracy of the final map products. To meet the first objective, the accuracy of the benthic habitat maps in the four test areas was completed for each of the imagery types. These were contrasted to compare the accuracy of each. To meet the second objective, the data were then combined and analyzed to establish an accuracy of the final mapped products.

The thematic accuracy of all mapped products was determined at the four most general levels of the classification scheme including unconsolidated sediment, submerged vegetation, coral reef/hard bottom and other. Four coral reef test areas were selected based on the diversity of the habitat types and to assure that all benthic habitats throughout the Hawaiian Islands were represented. The accuracy of all maps is, therefore, considered a conservative representation of the thematic accuracy of the habitat maps prepared using the same methods for imagery collected throughout the remainder of the Main Eight Hawaiian Islands.

Goals of the accuracy assessment:

1. Compare the thematic accuracy of benthic habitat maps prepared from three sources of remotely sensed imagery
2. Establish a statistically sound thematic accuracy for routine production of habitat maps for the Main Eight Hawaiian Islands.