WHOI PO

A ‘Deep’ Perspective on Forecasting the Loop CurrentSystem

In the Gulf of Mexico, the LoopCurrent (LC) system is the dominant feature that drives mesoscale variabilitywith the LC occasionally shedding closed, warm core anticyclones withhorizontal scales of 200-400 kilometers, known as LC Eddies (LCEs). Theseeddies carry significant energy and strong currents, with impacts on the manyindustries operating in the Gulf. Primary understanding of the LC and the LCEseparation process has been driven by altimeter derived surface measurementsand instruments operating in the upper 1000 meters. However, observationalcampaigns – with arrays consisting of Current Pressure Inverted Echo Sounders (CPIES)– and numerical studies have indicated the existence of deep (> 1000-meterdepth), mesoscale eddies which can generate deep energy and interact with theLC. Despite available deep observations, and evidence of the upper-deepcoupling, current forecasting efforts in the Gulf do not assimilate deepobservations, leaving the deep unconstrained.

In this talk, I first review theinfluence of deep, mesoscale eddies on 92-day, 32-member ensemble forecastpredictive skill. I focus on results from two ensemble forecasts covering theLCE Thor separation event. This time-period coincided with deep observationsfrom an array of twenty-four CPIES, providing bottom pressure and velocitymeasurements. Using a generalized ranking system, I determined the best andworst performing ensemble members and compared the deep fields between the twogroups. This revealed differences in the forecasted positioning and magnitudesof deep eddies. The two groups were then contrasted against CPIES data, withthe best member group showing better agreement with deep observations. Moreover,a link between initial deep uncertainty and upper predictive skill is found. Iwill conclude with current efforts to better understand the deep variability inthe Gulf using an extensive 18-year HYCOM nature run.