Diurnal warming in shallow coastal seas: Observations from the Caribbean and Great Barrier Reef regions

A good understanding of diurnal warming in the upper ocean is important for the validation of satellite-derived sea surface temperature (SST) against in-situ buoy data and for merging satellite SSTs taken at different times of the same day. For shallow coastal regions, better understanding of diurnal heating could also help improve monitoring and prediction of ecosystem health, such as coral reef bleaching. Compared to its open ocean counterpart which has been studied extensively and modeled with good success, coastal diurnal warming has complicating localized characteristics, including coastline geometry, bathymetry, water types, tidal and wave mixing. Our goal is to characterize coastal diurnal warming using two extensive in-situ temperature and weather datasets from the Caribbean and Great Barrier Reef (GBR), Australia. Results showed clear daily warming patterns in most stations from both datasets. For the three Caribbean stations where solar radiation is the main cause of daily warming, the mean diurnal warming amplitudes were about 0.4K at depths of 4–7m and 0.6–0.7K at shallower depths of 1–2m; the largest warming value was 2.1K. For coral top temperatures of the GBR, 20% of days had warming amplitudes >1K, with the largest >4K. The bottom warming at shallower sites has higher daily maximum temperatures and lower daily minimum temperatures than deeper sites nearby. The averaged daily warming amplitudes were shown to be closely related to daily average wind speed and maximum insolation, as found in the open ocean. Diurnal heating also depends on local features including water depth, location on different sections of the reef (reef flat vs. reef slope), the relative distance from the barrier reef chain (coast vs. lagoon stations vs. inner barrier reef sites vs. outer rim sites); and the proximity to the tidal inlets. In addition, the influence of tides on daily temperature changes and its relative importance compared to solar radiation was quantified by calculating the ratio of power spectrum densities at the principal lunar semidiurnal M2 tide versus 24-hour cycle frequency representing mainly solar radiation forcing, i.e., (PSDM2/PSD24). Despite the fact that GBR stations are generally located at regions with large tidal changes, the tidal effects were modest: 80% of stations showed value of (PSDM2/PSD24) of less than 10%.