Assessment of water quality in Sheyang Estuary (China) using hyperspectral data

ABSTRACT Estuaries are biologically productive and diverse coastal areas that are also vital to commerce, transportation, and recreation activities. In this paper, we demonstrated the potential of hyperspectral data for monitoring water quality in estuary. Many historical surveys showed that the water quality in Sheyang estuary was mostly deteriorated by the rich of dissolved oxygen (DO), chemical oxygen demand (COD), nitrate nitrogen (NTN), nitrous nitrogen (NSN), ammoniacal nitrogen (AMN) and total phosphorus (TP). Based on the regr ession analysis between spect ra radiometer measurements and ground reference data acquired synchronously, the sensitive bands for the estimation of the above six parameters were decided. An IDL-based atmospheric correction code, complemented with an air/water interface correction, was used to convert Hyperion at-sensor radiances into subsurfa ce irradiance reflectance values. These reflectance values were comparable to in situ reflectance spectra measured during Hyperion’s overpass, except at longer wavelengths (beyond 910 nm), where the reflectance values we re contaminated by severe atmospheric adjacency effects. The Hyperion spectra curves exhibit band to band spikes or dips and a selection of single bands could match some spikes. The binning of bands was used instead of single channel to develop the hyperspectral models. The model validation have relative RMSE values basically less than 30% but for TP validation, which indicates that Hyperion imagery could act as a landmark for moving forward the operational use of RS-related technologies. Integrated with traditional survey procedures, hyperspectral data could provide useful information for the dynamic monitoring of water quality in estuary. Keywords : hyperspectral data, water quality , estuary, Hyperion, models