Mapping aquatic systems with a physically based process chain T. Heege a/b , A. Bogner b , C. Häse b/c , A. Albert b , N. Pinnel a , S. Zimmermann a a Technical University of Munich, Limnological Station, Hofmark 3, D-82393 Iffeldorf, (Germany), thomas.heege@wzw.tum.de b DLR Oberpfaffenhofen, Münchner Str. 20, D-82230 Weßling, Germany c at present: GKSS Geesthacht, Max-Planck-Straße, D-21502 Geesthacht, Germany ABSTRACT Mapping the submersed vegetation can give significant information about the actual state of shoreline vegetation in inland waters. It is of prime importance for the ecological valuation of the entire lake. Remote sensing techniques can accomplish an efficient tool for mapping tasks, if the processing methods are universally valid. The Modular Inversion Program (MIP) follows this concept. It is a processing tool designed for the recovery of hydro-biological parameters from multi- and hyper-spectral remote sensing data. The architecture of the program consists of physical inversion schemes that derive bio-physical parameters from the measured radiance signal at the sensor. Program modules exist for the retrieval of aerosols, sun glitter correction, atmosphere- and water surface corrections, retrieval of water constituents, primary production in optically deep waters and the classification of substrates such as macrophytes and bottom sediments amongst others. For the purpose of mapping the bottom coverage in optically shallow waters, two modules have been added to MIP: The first module calculates the bottom reflectance using the subsurface reflectance, the depth and an approximation of the water constituent concentrations as input. The second module fractionalizes the bottom reflectance to three endmembers of specific reflectance spectra by linear unmixing. The three endmembers are specific reflectance spectra of bottom sediments, small growing macrophytes (Characeae) and tall macrophytes (here: mainly Potamogeton perfoliatus & P. pectinatus). The processing system has been tested with data collected from the multi-spectral airborne scanner Daedalus AADS1268 at Lake Constance, Germany, for multi- temporal analysis. Keywords: Remote Sensing, inland water, shallow water, macrophyte, retrieval, airborne, multispectral 1 INTRODUCTION Monitoring the temporal and spatial dynamics of inland waters is essential for the understanding of freshwater ecosystems. Remote sensing offers a good way to get such information of physical and biological parameters. Current research activity is focused on the development of general, automated and cost-effective methods to create maps of water constituents, phytoplankton primary production and aquatic vegetation at high spatial resolution. These activities are part of cooperation-projects concerning the linkage between remote sensing products and bio- indication systems appropriate for the assessment of the status of aquatic ecosystems. For instance, phytoplankton biomass and phytoplankton productivity are important indicators for the trophic state, and hence for water quality of aquatic systems. The same does apply for species and frequency of submersed littoral vegetation, which is also used for bio-indication tasks [1,2]. Areas fully mapped in high spatial resolution can give significant information about the actual state of shoreline vegetation in inland waters, and are of prime importance for the ecological valuation of the entire lake. An assessment of such information with traditional methods implies significant effort and expenses in terms of man power and time because of extensive data collection in the field. In addition, it is often limited to low areal coverage. Particularly in case of large lakes we are faced with such limitations. At Lake Constance a detailed mapping of submersed macrophytes has been carried out every 10 years by use of aerial photographs and ship based mapping. For inland waters, operational remote sensing methods have only been used to limited extent to assist in human power intensive surveying and mapping campaigns in the field. One reason could be the difficulty in delivering comparable products of satisfying quality. New methods for monitoring indicative parameters promise to be more efficient using up-to-date remote sensing technologies like modern sensors in combination with physical based processing schemes (e.g. MIP). Presented at the 3 rd EARSeL Workshop on Imaging Spectroscopy, Herrsching, 13-16 May 2003 415