Assessing positional accuracy and its effects on rice crop area measurement: an application at Coleambally Irrigation Area

Abstract

If management decisions are based on geospatial data that have not been assessed for spatial accuracy, then debate about both the measurements and the decisions themselves can occur. This debate, in part, can be avoided by evaluating the spatial accuracy of geospatial data, leading to heightened confidence in both the data and the decisions made from the data. To increase the effectiveness of environmental compliance monitoring, the spatial accuracies of 2 Geographic Information System datasets were estimated at the Coleambally Irrigation Area, New South Wales. The first, high-resolution digital aerial photography acquired in January 2000, is the Geographic Information System baseline data for Coleambally Irrigation Area. The second, Digital Topographic Data Base roads data, although not a reference dataset at Coleambally Irrigation Area, is often used as a baseline dataset across Australia. Neither dataset met the National Mapping Council of Australia’s standard of map accuracy, so a new version of the digital aerial photography was created that did. The positional accuracy of the improved dataset was over 4 times more accurate than the Digital Topographic Data Base roads dataset and over 2.5 times more accurate than the original digital aerial photography. It was also found that the overall areal error of paddocks measured from the improved dataset decreased as more paddock areas were added together; a finding that has a direct impact on management decisions at Coleambally Irrigation Area. This study both provides a demonstration of how to assess and improve spatial accuracy and shows that this process is not unduly complicated.