Logo 2016 Dragon 3 Final Results Symposium

Conference Agenda

Overview and details of the sessions and sub-session of this conference. Please select a date or session to show only sub-sessions at that day or location. Please select a single sub-session for detailed view (with abstracts and downloads if available).

Session Overview
E2: ID.10549 Forest Change Monitoring
Tuesday, 05/Jul/2016:
2:00pm - 3:00pm

Session Chair: Christiana Cornelia Schmullius
Session Chair: Yong Pang
Workshop: Forest Mapping & Retrievals
Location: Sun Moon Room -1, 5.5 Floor, Junyi Dynasty Hotel

Show help for 'Increase or decrease the abstract text size'
Oral presentation

Deforestation Detection in Subtropical And Tropical Region Using GF-1 Data

Bingxiang Tan1, Mike Wooding2

1Institute of Forest Resource Information Techniques, CAF, Beijing , China; 2Remote Sensing Applications Consultants, UK;

Deforestation Detection in Subtropical And tTopical Area Using GF-1 Data

Tan Bingxiang, Li Shiming, Liu Qingwang

Institute of Forest Resource Information Techniques, CAF, Beijing , China

Mike Wooding, Remote Sensing Applications Consultants, UK

According to the FAO report “Global Forest Resources Assessment 2015”, the total forest area in 2015 was 3 999 million ha, which is around 30.6 percent of the global land area. But in 1990 the world had 4 128 million ha of forest, 31.6 percent of global land area. It means that there was a net loss of some 129 million ha of forest from 1990 to 2015, representing an annual rate of –0.13 percent and a total area about the size of South Africa. The biggest forest area loss occurred in the tropics and subtropics, particularly in South America and Africa, although the rate of loss in those areas has decreased substantially in the past five years. Deforestation, or forest conversion is main reason for the forest area change. So, it is very important to develop techniques for closely monitoring areas of interest (e.g. deforestation) using remote sensing data(SAR or optical satellite data )to quickly detect changes within forested areas, especially using high resolution satellite imagery, and finally produce detailed and timely forest change maps.

For the monitoring of timely forest change, it needs time-series satellite data . At the moment, Sentinel 1 SAR data and GF-1 data are the better choice. The Chinese GF-1 satellite is a new high spatial resolution satellite launched on April 26, 2013. It was equipped with two types of sensors. One is the wide field view sensor (WFV sensor); the other is the panchromatic and multispectral sensor (PMS sensor). The WFV sensor can acquire multispectral image in blue, green, red, and near-infrared bands with 16 meters spatial resolution and 4 days temporal resolution. As GF-1 WFV has much high temporal resolution, it is possible to obtain optical remote sensing images for subtropical and tropical area in which there is more cloud. In our study, the test site is located in Ningming county of Gouangxi Province, China, with forest coverage rate of 27%. In this site, there is much more planted forest area so that the deforestation happens often. Two dates of GF-1 data were selected with the purpose of forest change monitoring (1 year apart), May10, 2014 and April 19,2015, and geometrically corrected using Landsat 8 OLI image.

For change detection, a very common method is called “post-classification comparison ”. It means that the two dates of imagery are separately classified firstly, then two classification results are compared by a pixel-by-pixel to detect changes in cover type, finally, change map is yielded. Using the approach, it will take a lot of time to do the classification of two date images and the it is difficulty to validate the classification accuracy of the first date. Therefore, just to find what the difference between deforestation and non-deforestation with the comparison of two date imageries, finally we selected the optimum bands and to get the combination image for the detection of deforestation. As test results, the selected optimum bands were B3 of second date, B3 of first date, and B2 of second date. In the combination image of this three bands, the deforested area were very clear and had different color. Using the composite image to classify the deforestation area with a supervised classification approach. After post-classification processing, the deforestation map was created.

The results indicate that t is possible to obtain good quality optical images every year in the subtropical and tropical area because GF-1 WFV sensor has high the temporal resolution (4 days). Deforestation areas in GF-1 imagery are very clear and easy to identify. GF-1 imagery is suitable for the detection of forest change over extremely large areas because of the 800km swath coverage. The use of just 3 bands from the 2 date images enabled easy identification and classification of deforested areas. New areas of deforestation are able to be tracked every year across the whole of sub-tropical China using just two cloud free images each year. In the future, the method developed in our study needs to be demonstrated for mapping over large areas. Forest change includes forest loss and forest gain. So far we have focused on deforestation, but future work needs to include mapping of new plantings and regenerated areas.

Keywords: Forest; Deforestation; GF-1, Remote Sensing

Tan-Deforestation Detection in Subtropical And Tropical Region Using GF-1 Data_Cn_version.pdf
Tan-Deforestation Detection in Subtropical And Tropical Region Using GF-1 Data_ppt_present.pdf

Oral presentation

Object-based Forest Cover Monitoring Using GF-1 Satellite Images in Subtropical Area:A Case Study of Shangsi County, China

Shiming Li, Zengyuan Li, Erxue Chen Chen, Bingxiang Tan, Qingwang Liu, Yong Pang, Xin Tian

Institute of Forest Resources Information Techniques, CAF, China, People's Republic of;

Subtropical areas are one of the fast-growing and high-yield plantation forest areas, and forest cover dynamics is changing for deforestation and reforestation. Monitoring forest cover change is of great importance for the scientific decision making of environment protection and forest sustainable management. The acquisition costs of high spatial satellite images are reduced largely because of the development of China High- Resolution Earth Observation System (CHEOS), and make forest cover monitoring in local area become possible. This paper examines an object-based forest cover monitoring approach to detect deforestation and reforestation with the help of the spectral, textural and contextual information from time series GF-1 Satellite Images. The result map is validated with ground truth data.

Li-Object-based Forest Cover Monitoring Using GF-1 Satellite Images_Cn_version.pdf


Method For Forest Vegetation Change Monitoring Using GF-1 Images

Lingyu Yin1, Xianlin Qin1, Guifen Sun1, Xiaofeng Zu1, Guang Deng1, Casanova Joseluis2

1Chinese Academy of Forestry Science , China, People's Republic of; 2Remote Sensing Laboratory,University of Valladolid. Faculty of Sciences;

To the monitoring technology needs for the forest cover change, Yajiang County in Sichuan province of People Republic of China has been selected as the experimental areas. By means of the analysed results on the factors, such as the image spectral characteristics and texture by using the images of the NO.1 high-resolution satellite (GF-1), the change of forest vegetation cover in the experimental region has been analyzed by using the image direct difference method, NDVI difference method, the first principal component difference method and tasseledcap difference method respectively. At the same time, the monitoring results have been compared and verified by using the large area forest fires which had taken place in yajiang County in recent year. The results showed that the overall accuracy and Kappa coefficient difference method are better than the tasseledcap difference method than other methods.

Yin-Method For Forest Vegetation Change Monitoring Using GF-1 Images_Cn_version.pdf

Contact and Legal Notice · Contact Address:
Conference: 2016 Dragon 3 Final Results Symposium
Conference Software - ConfTool Pro 2.6.94
© 2001 - 2016 by H. Weinreich, Hamburg, Germany