Modeling the diurnal variability of sea surface temperatures

In this study a one-dimensional mixed layer ocean model is customized for the purpose of estimating the diurnal signal of temperature in the near-surface ocean layer, generically referred to as sea surface temperature (SST). The model is initially run with data from three mooring locations. It is then demonstrated how operational forecast data sets can be utilized to estimate diurnal signals over a wide area. Daily diurnal variability maps are produced for a weeklong period over the Atlantic Ocean. These maps highlight the transient nature of diurnal SST signals with day to day changes in their magnitude and spatial distribution. The resulting diurnal variability maps are evaluated using a combination of infrared and microwave satellite-derived SST observations taken over the day. These matchups result in a mean error of 0.09 degrees C and a standard deviation of 0.54 degrees C. Advantages of modeling the diurnal cycle as opposed to using a persistence assumption are discussed.