Detecting and Monitoring Harmful Algal Blooms on Florida

Detecting and Monitoring Harmful Algal Blooms on Florida Coast Joseph Tuzzino, Brooklyn Technical High School Jonathan Tien, St. Francis Preparatory ABSTRACT Karenia brevis (K. brevis) blooms occur regularly on the Florida Coast. However, detection still remains a challenge from space due to the uncertainty of atmospheric correction, and interference from high concentrations of organic and inorganic materials in optically complex coastal waters. Our results show that Fluorescence Line Height (FLH) algorithm gives inaccurate results in highly scattering waters. So we used a simple red band difference technique (RBD) and a normalized difference technique, K. brevis bloom index (KBBI), proposed by Amin et al. , 2008, to detect and classify the potential areas of K. brevis blooms from Medium Resolution Imaging Spectrometer (MERIS). We applied these algorithms to satellite images for the blooms documented in the literature and our analysis shows that the RBD and KBBI detect, monitor and classify K. brevis blooms more precisely than FLH. Dr. Alex Gilerson, Dept. of Electrical Engineering, CCNY Ruhul Amin, Dept. of Electrical Engineering, CCNY INTRODUCTION MATERIALS The major natural phenomenon that is investigated in our research is an harmful algal bloom. An algal bloom is a rapid increase in the number of microalgae in water environments. Generally algal blooms are helpful for the environment because they are the source of marine food and they also produce large amounts of oxygen. However, some of the algae are harmful because they produces toxins which are detrimental to plants, marine animals, and even human 1. The species of harmful algae that is of interest is Karenia brevis. This specie is also known for the development of Florida red tide. K. brevis absorbs strongly, scatter weakly, and co-exists with low concentrations of mineral particles. It affects the Southeastern United States and most significantly Florida where blooms cause 20 to 30 million US dollars in damage per episode. §The KBBI technique often gives noise values at the cloud edge and near shore pixels but the RBD technique doesn’t. So both technique should be used together in order to minimize false bloom alarm and improve classification of K. brevis bloom. §Additional errors may be introduced in the classification algorithm due to normalization, particularly when highly absorbing colored dissolved organic matter (CDOM) is very high or due to inappropriate atmospheric correction algorithm. However, such problems are usually eliminated by the RBD technique. PROCEDURE §Database of algal blooms at http: //coastwatch. noaa. gov/hab/bulletins_ns. htm CONCLUSION Used the database from the website, we collected around 100 dates of harmful algal blooms in the Florida region from the year 2004 and recorded them along with latitudes and longitude where they occurred in an Excel spreadsheet § § MERIS satellite images from mercisrv. eo. esa. int/merci/query. Products. do Searched for clear satellite images of the Florida region on the MERIS website for those particular days when a HAB occurred and downloaded these images § §Microsoft Excel Using the BEAM software, we created images applying the RBD and KBBI as parameters to see if the harmful algal blooms could be detected clearly § §Beam software v. 4. 1 for image processing §Our results clearly show that the two band algorithms used, RBD and KBBI, were able to track an annual K. brevis bloom that had been occurring on Florida’s East Coast since 2006. §Our analysis also shows the advantages of applying RBD and KBBI technique over other traditional algorithms such as standard FLH to correctly identify the potential bloom area and to distinguish K. brevis from other blooms, plumes, sediments, and even shallow bottom reflectances. §To minimize false bloom alarm both RBD and KBBI technique should be used together. Our research entails applying remotes sensing measurements of European ocean color sensor MERIS to detect and monitor K. brevis blooms on the Florida Coast. Fig. 1 DISCUSSION Fig. 3 a March 15, 2008 RESULTS Fig. 3 b March 18, 2007 Fig. 3 c FUTURE RESEARCH Fig. 3 d § Apply these bloom detection and classification technique to MODIS § Study different regions in the Gulf of Mexico where K. brevis also blooms RBD= REFERENCES KBBI= §A. Gilerson, et al. “Retrieval of chlorophyll fluorescence from reflectance spectra through polarization discrimination: modeling and experiments. ” Applied Optics 45, 5568 -5581, 2006. Fig. 1: Modeled KBBI values for K. brevis and non-K. brevis blooms. The K. brevis bloom becomes distinguishable from non-K. brevis bloom for very low chlorophyll concentration about 1 mg/m 3 §A. Gilerson, et al. "Fluorescence component in the reflectance spectra from coastal waters. Dependence on water composition. " Optics Express Vol 15, 15702 -15721, 2007 COMPARISON IN DETECTING ALGAL BLOOMS BETWEEN KBBI and FLH §R. Amin, J. Zhou, A. Gilerson, B. Gross, F. Moshary and S. Ahmed. Detection of Karenia brevis Harmful Algal Blooms in the West Florida Shelf using Red Bands of MERIS Imagery. MTS/IEEE Quebec, 2008. Detection (RBD) Identification (KBBI) Fig. 3: a) Bloomed area using the RBD technique b) KBBI image. Detection (RBD) Identification (KBBI) ACKNOWLEDGEMENTS Fig. 3: c) Bloomed area using the RBD technique d) KBBI image. §Dr. Fig. 3 e March 17, 2006 Alex Gilerson, Senior Scientist, Dept. of Electrical Engineering, CCNY Fig. 3 f §Mr. Ruhul Amin, Ph D student, Dept. of Electrical Engineering, CCNY §Dr. Manuel Zevallos, NASA-COSI Summer Research Program Coordinator Fig. 2 a §Ms. Charlene Chan-Lee, NASA-COSI Summer Research Program Instructor Fig. 2 b Fig. 2 c Fig. 2: a) Karenia brevis bloom area detected with KBBI, b) Bloom detection using Fluorescence Line Height technique, c) Comparisons between KBBI and FLH. Detection(RBD) Identification(KBBI) Fig. 3: e) Bloomed area using the RBD technique f) KBBI image. §Ms. Galia Espinal, NASA-COSI Summer Research Program Instructor §Dr. Frank Scalzo, NASA GISS Educational Program Specialist