Estimation of suspended sediment concentration based on bio-optical mechanism and HJ-1 image in Chaohu Lake

Abstract

Synchronization experiment was conducted in June, 2009 to get Inherent Optical Properties (IOP) of water component in Chaohu Lake. Water bio-optical mechanism was studied combined with multispectral data of Environmental Satellite 1 (CCD), and then inversion models of total suspended matter (TSM), inorganic suspended matter (ISM) and organic suspended matter (OSM) concentration were built. The data indicated that: the absorption ratio of suspended particulate matter and CDOM to total were almost no change from band 1 to band 2 with about 85% and 9%, respectively. The ratio of pure water to total increased from 0.4% to 5.6%. Water reflectance in these two bands were influenced by absorption of three kinds of components: algae particles absorption surpassed non-algal particles in band 3, and so played an important role in total absorption with about 35.7%; the proportion of pure water absorption and particles matter backscattering both were 99% in band 4, so these two components decided the main inherent optical properties in band 4. The models of TSM and ISM concentration inversion based on band combination (band 3 + band 4)/(band 1 + band 2) were built, while OSM concentration was estimated by band 4/ (band 1 + band 2) index. Inversed by image data, RE of TSM concentration between modeled and measured was 33.4%, and RMSE was 18.68 mg/L. RE of ISM and OSM concentration were 39.9% and 35.2% respectively. The inversion was more accurate when satellite-ground data were just in the same day. At this situation, RE of ISM concentration dropped to 25.4%, and that of TSM and OSM reduced to 26.5% and 26.8% as well. HJ-1 Satellite, bio-optical property, Chaohu Lake, suspended matter Citation:Wang Q, Jin X, Li Y M, et al. Estimation of suspended sediment concentration based on bio-optical mechanism and HJ-1 image in As it is well known, water columns show weak remotely sensed signals owing to its strong absorption. Therefore, accurate retrievals of chlorophyll concentrations, suspended matter concentrations and others are always difficult tasks in the field of water color remote sensing. The increase of suspended matter content in waters can induce corresponding increase of spectral reflectance, which supports a chance for extracting suspended matter information. However, the collected spectral reflectance reflect mixtures of multiple material information, such as chlorophyll, colored dissolved organic matter (CDOM), which increase difficulty of extracting suspended matter contents in waters. Thus to accurately derive suspended sediment concentration, a key point is to explore optical parameter which can enhance suspended matter information to a large extent, and can also effectively decrease interference from other water components [1]. Many scholars proposed inversion models with different band combinations and image data from different sensors.

Absorption and backscattering coefficients and their relations to water constituents of Poyang Lake, China

“…Currently, many of these lakes are facing a number of problems, such as a decrease in lake area and water quality decline caused by natural or anthropogenic factors, which might hamper the sustainable development of lake functions [25]. Absorption and backscattering coefficients have been measured and analyzed in several Chinese lakes, such as Taihu Lake [29][30][31][32][33] and Chaohu Lake [34]; however, the IOPs of most Chinese lakes are still unknown.…”

What is the minimum river width for the estimation of water clarity using medium-resolution remote sensing images?

“…The HJ‐1A CCD data have similar spectral and spatial resolutions as the Landsat TM/ETM+ resolutions; however, the temporal resolution of the HJ‐1A data, i.e., a 2 day retour period, is much higher than for the Landsat TM/ETM+ data (http://www.cresda.com/). The HJ‐1A/B images have been successfully used to estimate various water quality parameters, such as total suspended sediment and total nitrogen concentration [ Wang et al ., ; Yan et al ., ; Yu et al ., ]. The multispectral WorldView‐2 image has a 1.8 m spatial resolution with five bands (450–510, 510–580, 630–690, 770–890, and 860–1040 nm), and the multispectral Pléiades image has a 2.0 m spatial resolution with four bands (430–550, 500–620, 590–710, and 740–940 nm).…”

Development of environmental monitoring satellite systems in China

“…The special issue "Key Technologies in Application of HJ-1A/B Satellite Data" has exhibited the new productions and key technologies in the applications of HJ-1A/B satellites, which covered studies in various fields of environmental remote sensing. The research contents include basal technologies of analyses of data features, image registration [18] and analyses of scaling effects [19], environmental monitors of land cover [20], surface water [21,22], urban heat island effect [23] and plague natural epidemic foci [24], and quantitative retrieval of aerosol optical thickness [25], photosynthetically active radiation [26], shrub LAI [27], forest aboveground biomass [28]. These papers reflect the recognitions on the data applications of environmental satellites to various fields, and also demonstrate the directions for the future development of environmental satellites in China.…”