Digital Chart Cartography: Error and Quality Control

This paper puts forward a novel concept of tetrahedron model of chart cartography error analysis and quality control which is necessary for introducing user to the former triangle model of error analysis and quality control, i.e. Reality-Producer-Chart-User. For the human being, high quality does not necessarily represent high accuracy, but represent a concept that what is fit for use. Classical cartography error theory coming from physics and mathematics will no longer meet the requirement of users, so the new concept deriving from communication and cognition sciences will compensate for the research. New model should adopt thought science which is from humanities. In tetrahedron model, producer refers to people who provide data, who manipulate data and who check data. User is considered as a quality controlling factor which has equal status with producer. Data quality control has changed from “Check-Driving” to “User-Driving”. The words for quality assessment change from error or uncertainty to practicability or applicability. Reality here refers to ocean and coastal zone. From a new viewpoint, the chart has been redefined as chart set. And characters of chart differing from terrain map and the specific quality assessment indexes have been discussed. Some innovative ideas of the research work were as follows: 1) Establishing a new tetrahedron model of quality control. The relationship between error and quality assessment was analyzed and practicability was adopted to assess the quality of chart product. 2) Defining the chart set. After changing from paper chart to electronic chart or digital chart, chart product become a set of digitizing collecting, storing and renewing data. 3) Building up the user assessment system of the chart product, which would perform real-time chart quality control. * Corresponding author. Tel.: +86-10-64889132 E-mail address: [email protected] 1. ACTUALITY OF CHART CARTOGRAPHY ERROR THEORY The series of steps of chart cartography, e.g. chart editing and designing, and generalizing (including selection, simplification, combination and replacement) will inevitably influence chart quality. How we estimate the degree of those operations? Is the final chart product fit for use? How is the dependability or reliability of the product? All the questions require scientific assessment of chart quality. The error research of digital chart data has for a long time adopted error and uncertainty theory of spatial data. Because these are part of spatial data, and have many commonness. Now the important norm when evaluate the chart data quality is the error or accuracy of spatial data of chart. The description of error or accuracy is provided by the producer. However, the specific chart cartography process and special user of chart are not deeply discussed. And the reliability and practicability of a chart could not be provided to the user. These lead to reexamine the quality assessment system and more reasonable and efficient quality control model should be built up. Based on the actuality of error theory, a new quality assessment model was established. Then we re-defined the chart set in the digitized society. Finally we built up a user assessment system to perform a real-time quality control. 1.1 Error theory Error, accuracy, uncertainty, quality, the four words are all about the assessment of geographic spatial data, and each has its own emphasis. The essential of error, accuracy, uncertainty, quality is all about the spatial data application system. Study on the error, accuracy, uncertainty, quality theory is carried out at home and abroad. For example, Li Deren, Liu Dajie, Chen Jicheng, Shi Wenzhong and Michael Goodchild, Giles Foody, Gerard Heuvelink, D. Griffith all study on spatial data error and establish the error analysis system, including the error source, error discrimination and measurement method, error propagation model, error administration and some minimizing methods to errors that influencing spatial data quality (Li D., 2002; Liu D., 1999; Cheng J., 2004; Shi W., 2005; Goodchild M., 1989). Traditional mathematical statistics methods are the base to the error analysis theory system. But the classic mathematical statistics theory must be revised and reinforced according to spatial data manipulating characters. A lot of progress has achieved since 1990s, for example, GIS spatial elements uncertainty model establishing methods (Heuvelink), error propagation model of spatial data processing and analysis(Liu,1999), remotely sensed image error and its impact and its error index establishment (Foody, ;Arnoff, 1985). But there are also many areas that need to be researched, for The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol. 38, Part II