Mapping and monitoring lakes in the Brazilian Pantanal wetland using synthetic aperture radar imagery

被引:12
作者
Costa, Maycira P. F. [1 ]
Telmer, Kevin H.
机构
[1] Univ Victoria, Dept Geog, Victoria, BC V8W 3P6, Canada
[2] Inst Nacl Pesquisas Espaciais, Divisao Sensoriamento Remoto, BR-12201 Sao Jose Dos Campos, Brazil
[3] Univ Victoria, Sch Earth & Ocean Sci, Victoria, BC, Canada
[4] Univ Estadual Campinas, Inst Geosci, Campinas, Brazil
关键词
wetlands; radar; lake size; Pantanal;
D O I
10.1002/aqc.849
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
1. The Pantanal wetland in South America is being threatened by human development, such as the construction of hydropower dams in the upper Paraguay River and the Hydrovia. The consequences of these developments are difficult to predict, but will potentially cause irreversible loss of habitat and biodiversity. 2. Radar satellite imagery can be used to establish baseline information that is crucial for monitoring changes in this region. In the case of the Pantanal, the pixel spacing of the imagery is a major concern if the goal is to count, map, and monitor lakes. To address this, Radarsat and JERS-1 data were acquired and used to produce a series of coarser-resolution images in order to understand the capabilities of the various resolutions to map Pantanal lake ecosystems. The simulations included resolutions of the existing JERS-1 mosaics (100m resolution) and the proposed ALOS/PALSAR mosaics (50m resolution). 3. Using the 15 in data it was found that in the Nhecolandia region of the Pantanal, only about 7% of lakes have an area larger than 0.2 km(2), 47% have an area larger than 0.05 km(2), and 50% are smaller than 0.05 km(2). Therefore, it is important to consider the substantial number of lakes smaller than 0.05 km(2). 4. The size of the errors in determining lake area depends on lake size, lake shape, and the image pixel spacing - long and small lakes have greater errors, and 100m resolution images show unacceptable errors. The error in estimating lake area associated with the pixel spacing of the Radarsat ScanSAR image is 12.5% and 25%, and for the proposed mosaics of ALOS/PALSAR it is 30% and 47% for lakes larger and smaller than 0.05 km 2 respectively. Therefore, based on simulated results, the spatial resolution of the observing SAR was an important determining factor in the capacity to discriminate small lakes in the Pantanal. Copyright (c) 2007 John Wiley & Sons, Ltd.
引用
收藏
页码:277 / 288
页数:12
相关论文
共 29 条
[1]  
Abdon MD, 1998, PESQUI AGROPECU BRAS, V33, P1675
[2]   The JERS Amazon Multi-season Mapping Study (JAMMS): observation strategies and data characteristics [J].
Chapman, B ;
Siqueira, P ;
Freeman, A .
INTERNATIONAL JOURNAL OF REMOTE SENSING, 2002, 23 (07) :1427-1446
[3]   Estimate of net primary productivity of aquatic vegetation of the Amazon floodplain using Radarsat and JERS-1 [J].
Costa, M .
INTERNATIONAL JOURNAL OF REMOTE SENSING, 2005, 26 (20) :4527-4536
[4]   Utilizing SAR imagery and aquatic vegetation to map fresh and brackish lakes in the Brazilian Pantanal wetland [J].
Costa, Maycira P. F. ;
Telmer, Kevin H. .
REMOTE SENSING OF ENVIRONMENT, 2006, 105 (03) :204-213
[5]   Use of SAR satellites for mapping zonation of vegetation communities in the Amazon floodplain [J].
Costa, MPF .
INTERNATIONAL JOURNAL OF REMOTE SENSING, 2004, 25 (10) :1817-1835
[6]   Biophysical properties and mapping of aquatic vegetation during the hydrological cycle of the Amazon floodplain using JERS-1 and Radarsat [J].
Costa, MPF ;
Niemann, O ;
Novo, E ;
Ahern, F .
INTERNATIONAL JOURNAL OF REMOTE SENSING, 2002, 23 (07) :1401-1426
[7]  
COSTA MPF, 1998, CANADIAN J REMOTE SE, V24, P339, DOI DOI 10.1080/07038992.1998.10874698
[8]  
DASILVA CJ, 2004, P WETL C 2004 25 30
[9]   New perspectives on global ecosystems from wide-area radar mosaics:: flooded forest mapping in the tropics [J].
De Grandi, GF ;
Mayaux, P ;
Malingreau, JP ;
Rosenqvist, Å ;
Saatchi, S ;
Simard, M .
INTERNATIONAL JOURNAL OF REMOTE SENSING, 2000, 21 (6-7) :1235-1249
[10]   Spectral reflectance characterization of shallow lakes from the Brazilian Pantanal wetlands with field and airborne hyperspectral data [J].
Galvao, LS ;
Pereira, W ;
Abdon, MM ;
Novo, EMML ;
Silva, JSV ;
Ponzoni, FJ .
INTERNATIONAL JOURNAL OF REMOTE SENSING, 2003, 24 (21) :4093-4112