Remote sensing data provides high resolution information on a variety of biological and environmental variables that can be utilised for identifying potential links between the environment and higher trophic level organisms (e.g. – drivers of fisheries recruitment). However despite their obvious advantages, data gaps appear quite regularly particularly in the higher latitudes where frequent cloud cover and low irradiance in the winter can limit the satellites spatial coverage significantly. We explored the potential to interpolate data gaps in remote sensing fields by using the DINEOF filling procedure on the chl-a, SST and PAR fields for Icelandic waters. Eight day composites of the variables were used and the accuracy of the interpolation was verified by comparing with in-situ data. The Root Mean Square error was <1 for all variables when at least 40% of the pixels were valid.
The reconstructed datasets were used to explore the biophysical coupling and spatio-temporal variability between chl-a and potential environmental drivers using maximum covariance analysis while generalised additive models were used to explore the spatial variability in the dominant drivers of the spring bloom phenology of chl-a. The findings are discussed in the context of producing indices for fisheries management in higher latitude ecosystems.
Niall McGinty will talk about Filling the gaps: Using novel techniques to interpolate satellite data and explore the environmental and climatic controls of chlorophyll-a variability.
Authors: Niall McGinty; Kai Logemann; Kristin Augustdottir; Kristinn Gudmundsson; Gudrun Marteinsdottir
Other Friday biology lectures – http://luvs.hi.is/is/fyrirlestrar-vorid-2015
Föstudagur, March 6, 2015 – 12:30 to 13:30
Háskóli Íslands, Askja Náttúrufræðihús