14
BAMOS Sep 2020
Figure 5 . High pass filtered (< 36 hour ) time series of ( a ) east‐west ; and , ( b ) north‐south components of currents at mooring TR50 . Vertical bands reflect diurnal changes ( positive / negative ) in currents reflecting circular motion through the water column . Units are ms -1
2 . Diurnal‐inertial resonance
Inertial currents are circular motions that are generated by wind stress in combination with the Coriolis force ( Ekman 1905 ; Simpson et al . 2002 ). The time taken to complete the circle is defined as the inertial period (= 2π / ƒ where ƒ is the Coriolis parameter ). When the period of wind forcing is close to the local inertial period , a resonance condition occurs . At the ' critical latitude ' ( 30 ° N – S ± 10 °) the inertial period is close to 24 hours which is also the period of local wind forcing resulting from the land‐sea‐breeze ( LSB ) cycle . Thus , diurnal‐inertial resonance is most likely to occur in regions close to the ‘ critical latitude ’. These resonant wind‐current responses have been addressed in a variety of field and theoretical studies and have been shown to enhance the upper ocean velocity field and vertical mixing ( e . g . Simpson et al . 2002 ; Kim and Crawford 2014 ). South‐west Australia experiences one of the most persistent and strongest LSB systems globally ( the Fremantle ‘ doctor ’) and the local inertial frequency is 22.6 hours . Low‐frequency , background currents also have an influence by changing the effective Coriolis frequency ( Kunze 1985 ; Lerczak et al . 2001 ). The local relative vorticity ( ζ ) is able to shift the planetary Coriolis frequency to the effective Coriolis frequency : ƒ eff
= ƒ + ζ / 2 and Mihanović et al . ( 2016 ) indicated that diurnal‐inertial resonance occurs when ƒ eff
= 24 hours . Diurnal‐inertial resonance was identified offshore Fremantle using field measurements ( HF Radar , ADCP and thermistors ) during the austral summer , when LSB system dominated the wind regime ( Mihanović et al ., 2016 ). Here , strong anti‐clockwise diurnal motions associated with winds and currents ( amplitudes > 0.3 ms – 1 ) penetrated to depths > 300 m with vertical isotherm diurnal fluctuations up to 60 m .
Time series of currents collected from a mooring offshore Fremantle ( TR50 , Figure 2 ) located in 500 m water depth indicated a period resonance in January / February 2020 ( Figure 5 ). Resonance ( ƒ eff
≈ 24 hours ) begins on 16 January with currents penetrating to water depths ~ 200 m . Ekman depth in the region is ~ 70 m and thus the currents penetrate much deeper than the Ekman depth . Over the period 27 January to 5 February , there was strong resonance with currents through the water column , forced by the LSB system penetrating to ~ 400 m ( Figure 5 ).
3 . Tropical storms / continental shelf waves
Fluctuations of sea level and currents with periods of 5 – 15 days (‘ weather ’ band ) are a common feature of many continental shelf regions . Most of this variability can be attributed to the generation and propagation of forced and free coastallytrapped waves in the coastal ocean generated by atmospheric forcing . A coastally‐trapped wave is a Kelvin wave that travels parallel to the coast , with maximum amplitude at the coast and decreasing offshore ( Huyer , 1990 ). Here we consider continental shelf waves ( CSWs ) that are generated by fluctuations in the alongshore winds and controlled by the cross‐shelf depth profile . They travel with the coast on their left in the southern hemisphere and along the Australian coast , shelf waves propagate anti‐clockwise relative to the landmass . Storm surges caused by tropical cyclones are well known to pose a risk for coastal regions . However , a lesser‐known effect of tropical cyclones is the generation of CSWs that can propagate along the coast and influence water levels thousands of kilometres away . Eliot and Pattiaratchi ( 2010 ) analysed many CSWs generated by tropical cyclones in the north‐west shelf propagating along the west and south coasts of Australia . These waves propagated up to 3500 km from the source region with speeds of 450 – 500 km day −1 ( 5.2 – 5.8 ms −1 ). Maximum trough to crest wave height was 0.63 m , comparable with the mean daily tidal range in the south‐west . The shelf wave is identified in the coastal sea level records , initially as a decrease in water level , 1 – 2 days after the passage of the cyclone and has a period of influence up to 10 days . The amplitude of the CSW was strongly affected by the intensity and tropical cyclone track , with those travelling parallel to the coast typically producing higher signals . This was due to the resonance effect that occurs when the speed of a tropical cyclone is equal to the speed of the CSW ( Fandry et al ., 1984 ).
As an example , tropical low 14U formed in the Gulf of Carpentaria on 23 January 2017 , moved westward and , on 26 January , was over open water north of Broome ( Figure 2 ). A maximum 10‐minute mean wind speed of 23 ms -1 was recorded at 0200 28 January ( AWST ) as 14U moved west , parallel to the coast at a mean speed of 6.6 ms -1 . The low did not develop into a tropical cyclone . The low‐pass filtered sea level records at coastal stations between Broome and Fremantle ( distance ~ 2400 km ) indicated the formation of a CSW that propagated over a distance ~ 2000 km over period of ~ 5 days with a mean speed of 5.5 ms -1 ( Figure 6 ). The maximum wave height ( 0.55 m )