其他摘要:We developed and validated a logging detection methodology using ALOS-2/PALSAR-2 data in order to effectively utilize logging area information in the forest cloud, which is becoming widely used by local governments. In Ibaraki Prefecture, we adopted a method of segmenting the images and extracting the areas where the backscatter coefficient decreased between two images. Then, we validated the accuracy and improved the method by conducting a field survey and comparing the observations with optical sensor satellite image. As a result, it was found that screening procedure based on the local microwave incidence angle can reduce false detections, and that using images with different directions of microwave radiation together can reduce undetected loggings, in mountainous areas where PALSAR-2 is not well suited. The detection accuracy was sufficient, with 82% user's accuracy and 76% producer's accuracy, even in mountainous areas. Next, we conducted a demonstration experiment in which we loaded logging area information onto Ibaraki Prefecture's forest cloud system, and four regional governments tested and confirmed it in the field. The results showed that the user's accuracy was 83%, the information was particularly effective in identifying logging areas deep in the mountains, and also showed that 30% of the logging areas were unreported. These results demonstrate the potential of PALSAR-2 data for efficient monitoring of logging areas.