Building an Automated Integrated Observing System to Detect Sea Surface Temperature Anomaly Events in the Florida Keys$^{\ast}$

Abstract

<para xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> Satellite-derived sea surface temperature (SST) images have had limited applications in near-shore and coastal environments due to inadequate spatial resolution, incorrect geocorrection, or cloud contamination. We have developed a practical approach to remove these errors using Advanced Very High Resolution Radiometer (AVHRR) and MODerate-resolution Imaging Spectroradiometer (MODIS) 1-km resolution data. The objective was to improve the accuracy of SST anomaly estimates in the Florida Keys and to provide the best quality (in particular, high temporal and spatial resolutions) SST data products for this region. After manual navigation of over 47 000 AVHRR images (1993–2005), we implemented a cloud-filtering technique that differs from previously published image processing methods. The filter used a 12-year climatology and <formula formulatype="inline"><tex Notation="TeX">$\pm$</tex> </formula>3-day running SST statistics to flag cloud-contaminated pixels. Comparison with concurrent ( <formula formulatype="inline"><tex Notation="TeX">$\pm$</tex></formula>0.5 h) data from the SEAKEYS <emphasis emphasistype="italic"> in situ</emphasis> stations in the Florida Keys showed near-zero bias errors <formula formulatype="inline"><tex Notation="TeX"> $(≪ 0.05\ ^{\circ}\hbox{C})$</tex></formula> in the weekly anomaly for SST anomalies between <formula formulatype="inline"><tex Notation="TeX">$-$</tex></formula>3 <formula formulatype="inline"><tex Notation="TeX">$^{\circ}\hbox{C}$</tex> </formula> and 3 <formula formulatype="inline"><tex Notation="TeX">$^{\circ}\hbox{C}$</tex></formula>, with standard deviations <formula formulatype="inline"><tex Notation="TeX">$≪ 0.5\ ^{\circ}\hbox{C}$</tex></formula>. The cloud filter was implemented using Interactive Data Language for near-real-time processing of AVHRR and MODIS data. The improved SST products were used to detect SST anomalies and to estimate degree-heating weeks (DHWs) to assess the potential</para>