Satellite-based Estimation of Chlorophyll-α Concentration in Turbid Productive Waters

Authors

• Wesley Moses, University of Nebraska-LincolnFollow

First Advisor

Anatoly A. Gitelson

Date of this Version

Fall 12-2009

Document Type

Dissertation

Citation

A dissertation presented to the faculty of the Graduate College at the University of Nebraska in partial fulfillment of requirements for the degree of Doctor of Philosophy

Major: Natural Resource Sciences

Under the supervision of Professor Anatoly A. Gitelson

Lincoln, Nebraska, December 2009

Comments

Copyright 2009, Wesley Moses. Used by permission

Abstract

Inland, coastal, and estuarine waters, which are often turbid and biologically productive, play a crucial role in maintaining global bio-diversity and are of immense value to aquatic life as well as human-beings. Concentration of chlorophyll-a (chl-a) is a key indicator of the trophic status of these waters, which should be regularly monitored to ensure that their ecological balance is not disturbed. Remote sensing is a powerful tool for this.

Due to the optical complexity of turbid productive waters, standard algorithms that use blue and green reflectances are unreliable for estimating chl-a concentration. Algorithms based on red and near-infrared (NIR) reflectances are preferable. Three-band and two-band NIR-red models based on the spectral channels of MODIS and MERIS satellites have been tested for numerous datasets collected with field spectrometers from inland, coastal, and estuarine waters. The NIR-red models, especially the two-band model with MERIS wavebands, gave consistently highly accurate estimates of chl-a concentration in waters from different geographic locations with widely varying biophysical characteristics, without the need to re-parameterize the algorithms for each different water body. The MODIS NIR-red model can be used to estimate moderate-to-high chl-a concentrations.

The NIR-red models were applied to airborne AISA data acquired over several lakes in Nebraska on different days with non-uniform atmospheric conditions. Without atmospheric correction, the NIR-red models showed a close correlation with chl-a concentration for each image. With an effective relative correction for the non-uniform atmospheric effects on the multi-temporal images, the NIR-red models were shown to have a close correlation with chl-a concentration, with uniform slope and offset, for the whole dataset.

The models were also applied to MODIS and MERIS images. Reliable results were obtained from the MERIS NIR-red models. Calibrated MERIS NIR-red algorithms were validated using data from the Taganrog Bay and Azov Sea (Russia) and lakes in Nebraska. The calibrated NIR-red algorithms have the potential for universal application to estimate chl-a concentration from satellite data routinely acquired over turbid and productive waters from around the globe.

Advisor: Anatoly A. Gitelson