Remote sensing of California estuaries: Monitoring climate change and invasive species

The spread of invasive species and climate change are among the most serious global environmental threats. The goal of this dissertation was to link inter-annual climate change and biological invasions at a landscape scale using novel remote sensing techniques applied to the San Francisco Bay/Sacramento- San Joaquin Delta Estuary. I evaluated the use of hyperspectral imagery for detecting invasive aquatic species in the Delta using 3 m HyMap hyperspectral imagery. The target invasive aquatics weeds were the emergent water hyacinth (Eichhornia crassipes) and the submerged Brazilian waterweed (Egeria densa). Data were analyzed using linear spectral mixture analysis (SMA). The results show the weeds were mapped with a classification accuracy of 90.6% compared to 2003 sample sites and 82.6% accuracy compared to 2004 sample sites. Brazilian waterweed locations were successfully mapped but the abundances were overestimated because we did not separate it from other submerged aquatic vegatation (SAV). I evaluated 3 m HyMap imagery, from 2004, for SAV species in the Delta, including: Brazilian waterweed ( Egeria densa), Eurasian watermilfoil (Myriophyllum spicatum ), curlyleaf pondweed (Potamogeton crispus), coontail (Ceratophyllum demersum), American pondweed (Potamogeton nodosus), fanwort (Cabomba caroliniana), and common elodea (Elodea canadensis). Data were analyzed using SMA with a classification accuracy of 84.4%. Spectral simulations of Brazilian waterweed and American pondweed show how spectral properties can change at different water depths and varying water quality. Finally I address the effect of inter-annual climate change on the estuary ecology in the San Francisco Bay by analyzing current (2002) and historical (1994-1996) Airborne Visible Infrared Imaging Spectrometer (AVIRIS) datasets to map salt marsh species distribution. The species in the estuary, Salicornia virginica, Spartinia foliosa, Scirpus robustus, and Distichlis spicata undergo dramatic changes in above ground distribution in response to El Nino and La Nina cycles. We used SMA to analyze 2002 AVIRIS imagery. Overall accuracy was 80.94%. 1994-1996 AVIRIS imagery was analyzed to determine the distribution of species and to relate changes to annual rainfall data. Based on the known salinity tolerances, the spatial distribution and zonation of these species respond to salinity stresses resulting from the El Nino and La Nina cycles.