You are here:
THE USE OF HYPERSPECTRAL REMOTE SENSING FOR THE DEVELOPMENT OF OPTICAL WATER QUALITY INDICATORS IN THE OHIO RIVER BASIN
Citation:
Autrey, B C. THE USE OF HYPERSPECTRAL REMOTE SENSING FOR THE DEVELOPMENT OF OPTICAL WATER QUALITY INDICATORS IN THE OHIO RIVER BASIN. U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-04/170 (NTIS PB2007-106631), 2004.
Impact/Purpose:
The goal of this research is to develop methods and indicators that are useful for evaluating the condition of aquatic communities, for assessing the restoration of aquatic communities in response to mitigation and best management practices, and for determining the exposure of aquatic communities to different classes of stressors (i.e., pesticides, sedimentation, habitat alteration).
Description:
Hyperspectral remote sensing for the assessment of inland water quality can be used in enhancing the capabilities of resource managers to monitor water bodies in a timely and cost-effective manner. The key factor in assessing the accuracy of water quality assessments based on remote sensing is determining the relationships between optical indicators of water quality and remotely sensed data. The usefulness of the optical indicators may depend in large part to their applicability to interpreting data derived from multiple water bodies. The inland water quality parameters have been evaluated using both airborne and satellite-bound sensors. In 1999, a Compact Airborne Spectrographic Imager (CASI) was flown by airplane over the relatively shallow Great Miami River (GMR), Ohio, collecting hyperspectral bands of data. Corresponding water quality samples and field spectrometer data were collected directly from the river during the time of the flight. A similar study was executed in 2001 during which a CASI sensor was flown over a portion of the relatively deep Ohio River while the same types of ground-truth data were collected. Using the remotely sensed, field spectrometer and laboratory analyses data from each of these projects, spectral indices for the analysis of chlorophyll a, turbidity, phosphorus, and nitrogen were developed. The spectral indices developed during the 1999 GMR study were tested against the spectral indices developed for the Ohio River in 2001. While the GMR's spectral index for chlorophyll a was found to be applicable to the Ohio River data, the spectral indices for turbidity, phosphorus and nitrogen were required to be slightly refined in order to be applicable to this new environmental setting.
In 2002 and 2003, hyperspectral data were acquired from the Ohio River using the Hyperion satellite. Near the time that the satellite collected these data, a field crew collected ground-truth data types similar to those collected during the 1999 and 2001 studies. The same spectral index that worked well in predicting chlorophyll a concentrations in the 1999 and 2001 studies were applied to the data collected by the satellite. The results of that application helped validate the chlorophyll-a spectral index.
This study demonstrates the ubiquitous application of the chlorophyll a spectral index while revealing the limited reliability of the turbidity, phosphorous and nitrogen spectral indices. Although differences between the dynamics of the two rivers may have made these spectral indices incompatible, with further refinement they may yet prove to be useful tools that can be modified for use in other rivers to detect potential water quality problems.