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

Abstract

ABSTRACT1. 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 (100 m resolution) and the proposed ALOS/PALSAR mosaics (50 m resolution).3. Using the 15 m data it was found that in the Nhecolaˆndia 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 100 m 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.

Using Remote Sensing to Map and Monitor Water Resources in Arid and Semiarid Regions

“…Note that three different marsh vegetation species were identified. Credits: NOAA National Ocean Service, Charleston, SC, USA differently depending on the dominant vegetation type/structure [20] as well as the biomass and condition of vegetation [126,127]. In areas of open water without vegetation, specular reflection occurs and a dark signal (weak or no return) is observed [128].…”

SAR‐based estimates of the size distribution of lakes in Brazil and Canada: a tool for investigating carbon in lakes

“…More details about the problems of spatial resolution are discussed in Costa and Telmer (2007). It is also a strong possibility that the multi-polarimetric capabilities of ALOS will aid in imaging smaller lakes.…”

“…These qualities will allow ALOS to differentiate water from land under the greatest diversity of conditions. Because L-band penetrates vegetation, ALOS should easily identify small lakes covered or fringed with short aquatic vegetation such as those in the Brazilian Pantanal discussed by Costa and Telmer (2007). Its multipolarimetric abilities may improve target recognition in cases where vegetation penetration is needed.…”

China’s wetland databases based on remote sensing technology

“…Other multispectral sensors with finer spatial resolution (such as QuickBird, 2.4 m, and WorldView-2, 0.5 and 1.8 m) can improve wetland mapping accuracy and are increasingly used in wetland mapping (Belluco et al, 2006;Ouyang et al, 2011), but the difficulty and expense of acquiring them (Nagendra et al, 2013) hinder their advance. The limitations of multispectral imagery include the difficulty of obtaining consistent high-quality data due to climate conditions (Costa, 2004;Costa and Telmer, 2007), low accuracy in detecting forest wetlands (Huang et al, 2014), and insufficiency for mapping small wetland patches (Ausseil et al, 2007). Because hyper-spectral data resolve more details than multispectral data, they can better characterize wetlands (Hirano et al, 2003;Schmidt and Skidmore et al, 2003;Klemas, 2011;Zhang and Xie, 2012), including identification of marsh vegetation species and detection of forested wetlands.…”