Tropical Cyclones - (ARC) Centre of Excellence for Climate System
Transcription
Tropical Cyclones - (ARC) Centre of Excellence for Climate System
Tropical Cyclones Elizabeth Ritchie School of Physical, Environmental, and Mathematical Sciences UNSW – Canberra [email protected] Outline • • • • • • • Climatology Mature Tropical Cyclone Structure Movement Intensification Mechanisms Genesis Extratropical Transition Landfall Impacts Climatology 1851-2006 WPAC CPAC EPAC NATL NIO SIO Australian Region SPAC Climatology • • • • • On average 80 tropical cyclones (> 17m/s surface sustained winds) form globally every year (Gray 1968) as few as 69 and as many as 103 have formed in one year. Formation locations are tied to warmest ocean temperatures Regional seasonality is linked to the regional seasonal variation of ocean temperatures Maximum activity in late summer, early autumn when the ocean heat content is maximum Climatology • NORTHERN HEMISPHERE (Jul-Sep) • North Atlantic & Gulf of Mexico: • peak in activity ~ 10 Sep • ~ 12.1/yr [1,25] • Eastern & Central North Pacific: • Densest TC activity (per km2) • 15-16/yr • Western North Pacific: • Most active basin • TCs in all months • ~26/yr • North Indian Ocean: • Deadliest TCs (Bay of Bengal) • ~4/yr Climatology • SOUTHERN HEMISPHERE (Jan-Mar) • Australian/SEIO: • ~11/yr • Double peak in activity – lull as monsoon moves over Australian continent • Variability from ENSO • SPAC: • 7-8/yr • Mostly west of dateline except in El Niño periods • SWIO: • ~9/yr • Rain records held by TCs making landfall in LaReunion. Mature Tropical Cyclone Structure Hurricane Katrina (2005) • Standard features of a mature tropical cyclone • From satellite: • Eye • Eyewall • Rainbands • Cirrus outflow • From radar: • Eye • Eyewall • Spiral Rainbands • Convective Structure • Moat Regions TC Yasi (2011) Mature Tropical Cyclone Kinematic Structure Hurricane Gilbert (1988) • First-order approximation: • Axisymmetric (k=0) • Asymmetries superposed (k>0) • • • Cylindrical Coordinate system centred on the TC centre. Azimuthally average Decompose the wind field into tangential and radial components. Radial-height cross-sections • Composite Cyclone ࢂ࢘ ࢂࣅ ࢂ Mature Tropical Cyclone Kinematic Structure • Primary Circulation • Tangential component of wind • Strongest component • Near-zero at the centre • Increase linearly to radius of maximum winds (RMW) just outside the eyewall • • • • • Hurricane Gilbert (1988) Composite Cyclone Decrease exponentially outside RMW Secondary maximum in principle rainbands Maximum winds near the surface Fairly constant up to about 500 hPa (barotropic) then drop off rapidly with height Cyclonic to upper levels begin to spiral out and turn anticyclonic at 100-200 km radius in upper-levels. Mature Tropical Cyclone Kinematic Structure • Secondary Circulation - comprises radial and vertical components • Much weaker than primary circulation • Strong inflow in boundary layer (response to friction) • Weaker, deep layer of inflow through middle levels • • • • • Concentrated upper-level, near-tropopause outflow Saturated ascent in the eyewall Weak descent in moat regions and at far radius Inflowing air loses angular momentum through friction, so after it rises in the eyewall and turns outward, it swirls anticyclonically beyond 100-200 km radius Secondary circulation supplies AM and thermal energy, which intensifies the primary circulation and maintains it against frictional dissipation Mature Tropical Cyclone Kinematic Structure • • TCs come in many different sizes and shapes Note that the entire wind field of TC Tracy (cat 4 TC) almost fits inside the RMW of Supertyphoon Tip • Size doesn’t tell the whole story as Tracy, in spite of its midget size, wiped out the heart of Darwin with a direct hit Extreme sizes: Midget Cyclone Tracy: 100 km Super Typhoon Tip: 2220 km Average TC Kerry: 550 km Mature Tropical Cyclone Thermal Structure • In cylindrical coordinates: • The Thermal wind equation can be written:࢜ࣅ ࢌ ࢜ࣅ ൌ ࢘ ࢘ • ࣔ Which can be re-written by talking ࣔࢠ of both sides and substituting ࣔ ࣔࢠ ൌ ࡾࢀ ࡴ as ࢜ࣅ ࢜ࣅ ࡾ ࢀ ࢌ ൌ ࡴ ࢘ ࢘ ࢠ • For ࢜ࣅ with z , T with r . That is, the TC is warmest at the centre warmcore system. Mature Tropical Cyclone Thermal Structure Furthermore … ࢜ࣅ ࢜ࣅ ࡾ ࢀ ࢌ ൌ ࢘ ࢠ ࡴ ࢘ ฺ ࣔࢀ̱ ࢜ࣅ ࣔ࢜ࣅ ࣔ࢘ ࡴ ࢌ ࢘ ࣔࢠ ࡾ Scale Analysis: L ~ 100 km, f ~ 5 x 10-5 s-1, U ~ 50 m s-1. ࢾࢀ ̱ ࢁ ࢌ ࡸ ࢁࡴࡸ ࡴࡾ ̱ ൈ ൈ ି ൈ ̱ ૠԨ ૡૠ Mature Tropical Cyclone Thermal Structure Hurricane Hilda (1964) • Perturbation Ԩ • interpolated from flight-level data above and below (dashed lines) • Low surface pressure at center hydrostatically balanced by warm core aloft Hawkins and Rubsam 1968 Mature Tropical Cyclone Thermal Structure • In General: • Highest temperature perturbation located near 200-300 hPa • Cold anomaly above • Thermal wind balance tells us that the strongest winds must be near the surface and decrease with height – as we have already observed Mature Tropical Cyclone Kinematic Structure • Recall: The Rossby number can be used to assess the relative importance of the Coriolis force: Centrifugal Acceleration ݒఒ ଶ Τݎ ݒఒ ܴ ൌ ൌ ݂ݎ ݂ ݒఒ Coriolis Acceleration The flow is in: gradient balance if R0 ~ 1 - TC outer region ݒఒ ଶ ͳ ߲ ݂ ݒఒ ൌ ߩ ߲ݎ ݎ cyclostrophic balance if R0 >> 1 - TC core region (inside RMW) geostrophic balance if R0 << 1 - Environment. ݂ ݒఒ ൌ ݒఒ ଶ ͳ ߲ ൌ ߩ ߲ݎ ݎ ͳ ߲ ߩ ߲ݎ Movement • Generally initially steered by deep tropospheric easterly winds – so move to the west • If in the deep tropics then westward movement is typical • If slightly further poleward, then can get steered around the subtropical high and recurve into the midlatitude westerlies – head back to the east. • In NIO tracks tend to be directly northward Movement • And sometimes they just do seem to do really weird things. Movement Some basic controls on motion include:• Environmental Steering: • to first order this is a deeplayer mean of the environmental winds within about 1000 km • Interactions with other weather systems: • can consider weather systems as vortices and approximate vortex interactive behaviour Movement • The Earth vorticity gradient: • Puts a northwest (southwest) motion on the TC in the Northern (Southern) hemisphere • Convective asymmetries in the ૠԨ TCs own circulation: • low-level convergence associated with strong convection induces positive vorticity tendencies. TC moves toward positive vorticity tendencies – small track deflection Tropical Cyclone Intensification • It is fairly well accepted that tropical cyclones rely on ocean fluxes as their energy source – this is why activity is maximised in the late summer early autumn • The primary circulation is maintained/intensified against the effects of frictional dissipation because of the high ࣂࢋ air brought in via the secondary circulation, lifted in convection and then expelled in the outflow layer. • As this air flows in the inflow branch, angular momentum is lost to friction, and heat is lost to adiabatic expansion. If this energy is not replaced somehow, then the air that is lifted in deep convection does not add much energy to the mid-to-upper troposphere. Tropical Cyclone Intensification • As the air parcels adiabatically expand and cool in the inflow layer, they pick up energy from sensible and latent heat fluxes from the warm ocean • When they arrive at the eyewall they have very high ࣂࢋ values – above the ambient tropics This air is then lifted in deep convection. The latent heat release that occurs just offsets adiabatic expansion and adds little heat to the upper atmosphere • • The high ࣂࢋ air is lifted nearly moist adiabatically and for values of ~365K can produce a central pressure of ~960 hPa. Tropical Cyclone Intensification • A second mechanism kicks in at about 33 m/s – eye formation • Back of the envelope calculations suggest that a subsidence of w ~ -2 cm/s can produce a net warming of ૡǤ ιȀࢊࢇ࢟ • • This subsidence also acts to lower the maximum warm core from the tropopause (100 hPa) to 200-300 hPa as we saw in TC Hilda. • The additional warming results in much lower surface central pressures via hydrostatic balance than would otherwise be possible. Tropical Cyclogenesis Stages of Intensity Tropical Disturbance ??? 1. 2. Tropical Depression < 17 m/s Genesis: A mixture of large-scale (environmental), mesoscale, and convective-scale processes interact 3. Named Tropical Cyclone 17 - 33 m/s 4. Intense/Severe Tropical Cyclone > 33 m/s Intensification/Steady-state maintenance: WISHE Tropical Cyclogenesis 30 N 20 N 10 N 90 W 75 W 60 W 45 W Tropical Cyclogenesis Large-scale conditions necessary for genesis (Gray 1968): Thermodynamic Conditions: x Sufficient ocean thermal energy x Enhanced mid-tropospheric relative humidity Thermodynamic conditions for deep convection x Conditional instability Dynamic Conditions x weak vertical shear at the genesis location x at least 5° latitude away from the equator x Enhanced lower tropospheric relative vorticity Deep Convection Non-zero Coriolis Pre-existing disturbance x Transient large-scale circulations can enhance the environment for genesis – MJO, Equatorial waves (Kelvin, Rossby, inertia-gavity) Tropical Cyclogenesis Types of “parent” disturbances: x Monsoon trough – provides horizontal shear vorticity environment x Monsoon gyre – provides both horizontal shear vorticity and rotational vorticity x ITCZ – shear vorticity – can become unstable and spontaneously break down Tropical Cyclogenesis Types of “parent” disturbances: x Easterly waves – weak rotational vorticity strongest at 600-700 hPa x Common in the North Atlantic x African easterly waves are triggered by an instability on the low-level jet that forms as a result of the reversal of the largescale PV gradient over the Sahara x Wavelength: ~1500-2000 km Tropical Cyclogenesis n=1 Equatorial Rossby Wave: x Triggered by large-scale persistent heating in the tropics x westward phase speed of ~ 5 m/s x Dispersive x Wavelength: 2000-4000 km x associated with twin TC developments straddling the equator Tropical Cyclogenesis Two-stage process to genesis: • Stage 1: Preconditioning of the environment. • Stage 2: Mesoscale development. Tropical Cyclogenesis • Key questions: 1. Reduction of the Rossby radius of deformation, LR the kinematic response to a cumulus cloud is to excite gravity waves that disperse the energy to LR. The rotational wind response to the heating in the cloud is at the scale of LR. Mean tropical Rossby Radius is ~2000 km whereas the TC core is < 100 km. Scale problem Tropical Cyclogenesis • Key questions: 2. Disorganised convection to organised convection (Ooyama 1960s) for an individual air mass cumulus cloud, the downdrafts on the edge of the cloud, and finally within the cloud itself offset, and may even dominate, the updraft effect that produced boundary layer convergence and a spin up of low-level vorticity. Tropical Cyclogenesis There is both a kinematic and thermodynamic cooperation that is the key to this question. - Convection acts to moisten the lower and middle levels of the atmosphere - As the convection becomes more long-lived in the form of MCSs then a dynamic response occurs within the saturated midlevels of the MCS – a midlevel vortex (100-200 km) (Chen and Frank 1993) - Acts as an organizing feature for more convection- continued moistening reduces the downdrafts so that surface convergence is maintained and gradually enhanced (number of refs) - Process can happen very quickly (within 24 hours) or as a pulsing process over several days depending on the supporting background environment Tropical Cyclogenesis Example: TS Ofelia (1993) - Flight centered on 0000 UTC 23 July 1993 Potential Vorticity along 16 㼻N Divergence along 16 㼻N Tropical Cyclogenesis TS Ofelia (1993) 500 mb 700 mb 850 mb Tropical Cyclogenesis TS Ofelia (1993) - Flight centered on 0800 UTC 24 July 1993 Potential Vorticity along 17 㼻N Relative Vorticity along 17 㼻N Tropical Cyclogenesis Re-initiation of Convection … TS Ofelia (1993) Subsequent to 2nd flight… Voila … TS Ofelia! Extratropical Transition STY Chataan 2002 0709/00Z 0709/12Z 0710/12Z SLP Rising 973 mb 0711/12Z TC Dissipating 982 mb STY Phanfone 2002 0818/12Z 969 mb 0819/12Z SLP Rising 0820/12Z 985 mb ETC Intensifying 00821/00Z ~ 975 mb Extratropical Transition These systems generate heavy seas and very large waves (speed of translation resonates with maximum winds to generate large swell) → impacts shipping, for example… • The QEII encountered a rogue 100’ wave in an ET (Hurricane Luis) system in 1995. • The “1991 Halloween Storm” (Hurricane Gladys and strong midlatitude trough off U.S. east coast in 1991) – loss of a fishing trawler and all on board. • The 1998 Sydney-Hobart yacht race Sub-tropical system and rapidly developing trough between Victoria and Tasmania. Loss of 7 yachts, 57 sailors rescued, 6 died. 44 out of 115 completed the course. 40 m/s (90 mph) winds and over 80’ waves. Landfall and Impacts Tropical Cyclones are socially and economically the most destructive natural phenomena on the globe Considering all natural hazards, they account for*: • 14% of natural disasters • 26% of natural disaster economic damage Considering all meteorological hazards, they account for: • 48% of meteorological disasters • 69% of human impact of meteorological disasters Finally, 77% of the disasters associated with TCs in the last decade (EM-DAT): • were associated with heavy rain, flooding, and associated impacts * Data for 1990-2013 (Guha-Sapir et al. 2013; World Bank 2014, International Disaster Database, 2014) Landfall and Impacts Definition of landfall: when the center of the storm moves across the coast. - in strong TCs this is when the eye moves over land. > 63 m/s > 50 m/s - The severest wind impacts of these systems are concentrated near the eyewall. Landfall and Impacts BUT … damaging winds can extend more than a hundred kilometres from the eye Storm surge can extend several hundred kilometres Tornadoes develop typically in outer rainbands as they come on shore. Large amounts of rain from the core and rainbands can continue to fall far inland of landfall. If combined with steep terrain, significant flooding and mudslides can develop > 63 m/s > 50 m/s > 33 m/s > 25 m/s > 17 m/s Landfall and Impacts Wind Storm surge Rain Landfall and Impacts Flooding TC Oswald (2013) – Australia TY Saola (2012) – Philippines Cyclone Nargis (2008) – Myanmar Land/Mudslides H. Mitch (1998) – Nicaragua H. Mitch (1998) – Nicaragua TS Manuel (2013) – Mexico Landfall and Impacts Landfall and Impacts Global Landfalling TC Impacts *1991 and 2008 excluded Thank You!!! Elizabeth Ritchie School of Physical, Environmental, and Mathematical Sciences UNSW – Canberra [email protected]