This study evaluates the efficacy of remote sensing technology to monitor species composition, areal extent and density of aquatic plants (macrophytes and filamentous algae) in impoundments where their presence may violate water-quality standards. Multispectral satellite (IKONOS) images and more than 500 in situ hyperspectral samples were acquired to map aquatic plant distributions. By analyzing field measurements, we created a library of hyperspectral signatures for a variety of aquatic plant species, associations and densities. We also used three vegetation indices. Normalized Difference Vegetation Index (NDVI), near-infrared (NIR)-Green Angle Index (NGAI) and normalized water absorption depth (DH), at wavelengths 554, 680, 820 and 977 nm to differentiate among aquatic plant species composition, areal density and thickness in cases where hyperspectral analysis yielded potentially ambiguous interpretations. We compared the NDVI derived from IKONOS imagery with the in situ, hyperspectral-derived NDVI. The IKONOS-based images were also compared to data obtained through routine visual observations. Our results confirmed that aquatic species composition alters spectral signatures and affects the accuracy of remote sensing of aquatic plant density. The results also demonstrated that the NGAI has apparent advantages in estimating density over the NDVI and the DH. In the feature space of the three indices, 3D scatter plot analysis revealed that hyperspectral data can differentiate several aquatic plant associations. High-resolution multispectral imagery provided useful information to distinguish among biophysical aquatic plant characteristics. Classification analysis indicated that using satellite imagery to assess Lemna coverage yielded an overall agreement of 79% with visual observations and >90% agreement for the densest aquatic plant coverages. Interpretation of biophysical parameters derived from high-resolution satellite or airborne imagery should prove to be a valuable approach for assessing the effectiveness of management practices for controlling aquatic plant growth in inland waters, as well as for routine monitoring of aquatic plants in lakes and suitable lentic environments. ?? 2010 Blackwell Publishing Ltd.
|Title||Differentiating aquatic plant communities in a eutrophic river using hyperspectral and multispectral remote sensing|
|Authors||Y.Q. Tian, Q. Yu, M.J. Zimmerman, S. Flint, M.C. Waldron|
|Publication Subtype||Journal Article|
|Series Title||Freshwater Biology|
|Record Source||USGS Publications Warehouse|