Alinune Musopole
Climate change is one of the biggest challenges that we are fighting in the 21st century. One of the indicators of climate change is lake surface water temperature (LSWT)-LSWT is expected to be periodic and a move away from periodicity verifies change in climate. With surface temperature of water on a lake obtained at high frequency both spatially and temporally, the volume of data is high. One of the ways used in reducing dimensionality of data is by approaching the data as functional data- functional principal components (fPCs) reduce dimensionality by giving modes of variation that are dominant in the data. In this paper we apply a method called principal periodic components (PPCs) that is capable of separating variability in the data into that which is nearly-periodic and that which is non-periodic, on LSWT data for Lake Malawi. We also carry out a test to check whether there is any exact annual variation in the data or not. The data are remote sensing images. The analysis has shown that there is no any exact annual variation in LSWT data for Lake Malawi- LSWT for Lake Malawi, though with strong periodicity, is not strictly periodic.