Climate variability of tropical cyclones: Past, Present and Future. Storms, 2000
edited by R. A. Pielke, Sr. and R. A Pielke, Jr, Routledge, New York,220-241..
Climate variability of tropical cyclones: Past, Present and Future. Storms, 2000
edited by R. A. Pielke, Sr. and R. A Pielke, Jr, Routledge, New York,220-241..
Worldwide, tropical cyclones are the deadliest and costliest naturaldisasters, as the approximate 300,000 death toll in the infamous BangladeshCyclone of 1970 and the $26.5 billion (U.S.) in damages due to the 1992Hurricane Andrew in the Southeast United States can attest (Holland 1993,Hebert et al. 1997). Pielke and Pielke (1997) show that U.S. hurricanedamages - which exceed those due to earthquakes by a factor of four - accountedfor 40% of all insured property losses for 1984 to 1993. Understandinghow tropical cyclone activity has varied in the past and will vary in thefuture is a topic of great interest to meteorologists, policymakers andthe general public. Some have expressed concern about the possibility thatanthropogenic climate change due to increases in "greenhouse" gases mayalter the frequency, intensity and areal occurrence of tropical cyclones.A review of the interannual variations of tropical cyclones, their causesand seasonal predictability has been covered by Landsea (1999). This chapter,as documented from instrumental records and the emerging field of paleotempestology,will focus instead on what have been the long-term variations in globaltropical cyclone activity, what may be responsible for such variability,and what might occur in future decades through both natural fluctuationsand man-made causes.
"Tropical cyclone" is the generic term for a non-frontal synopticscale "warm-core" low-pressure system that develops over tropical or sub-tropicalwaters with organized convection and a well-defined cyclonic surface windcirculation. It derives its energy primarily by evaporation of water andsensible heat flux from the sea enhanced by high winds and lowered surfacepressure. These energy sources are tapped through condensation in convectiveclouds concentrated near the cyclone's center (Holland 1993). Tropicalcyclones with maximum sustained surface winds of less than 18 ms-1are called "tropical depressions". Once tropical cyclones reach winds ofabout 18 ms-1 they are typically called a "tropical storm" andassigned a name. If winds reach 33 ms-1, they are called: a"hurricane" (the North Atlantic Ocean, the Northeast Pacific Ocean eastof the dateline, or the South Pacific Ocean east of 160°E); a "typhoon" (the Northwest Pacific Ocean west ofthe dateline); a "severe tropical cyclone" (the Southwest Pacific Oceanwest of 160°E or Southeast IndianOcean east of 90°E); a "severecyclonic storm" (the North Indian Ocean); or a "tropical cyclone" (theSouthwest Indian Ocean) (Neumann 1993). Additionally, the category of "intense(or major) hurricane" has been utilized for the Atlantic basin for thosetropical cyclones obtaining winds of at least 50 ms-1, whichcorresponds to a category 3, 4 or 5 on the Saffir-Simpson hurricane intensityscale (Simpson 1974, Hebert et al. 1997).
Before tropical cyclogenesis and further development can occur, severalnecessary environmental conditions must be met (Gray 1968, 1979):
1. Warm ocean waters (of at least 26.5°C) throughout a sufficient depth (unknown how deep, but atleast on the order of 50 m) - are necessary to fuel the tropical cycloneheat engine1.
2. An atmosphere in which temperatures decrease fast enough with heightsuch that it is potentially unstable to moist convection. It is the precipitatingconvection typically in the form of thunderstorm complexes that allowsthe heat stored in the ocean waters to be liberated for the tropical cyclonedevelopment.
3. A relatively moist mid-troposphere. Dry middle levels are not conducivefor allowing the continuing development of widespread thunderstorm activity.
4. A minimum distance of around 500 km from the equator. For tropicalcyclogenesis to occur, there is a requirement for sufficient amounts ofthe Coriolis force to provide for near gradient wind balance to occur.
5. A pre-existing near-surface disturbance with sufficient vorticityand convergence. Tropical cyclones cannot be generated spontaneously. Todevelop, they require a weakly organized system with sizable spin and lowlevel inflow.
6. Low magnitudes (less than about 10 ms-1) of vertical windshear between the ocean's surface and the upper troposphere. Vertical windshear is the horizontal wind change with height. Large values of verticalwind shear disrupt the incipient tropical cyclone and can prevent genesis,or, if a tropical cyclone has already formed, large vertical shear canweaken or destroy the tropical cyclone by interfering with the organizationof deep convection around the cyclone center.
These six conditions are necessary, but not sufficient, as many disturbancesthat appear to have favorable conditions do not develop. Recent work (Velascoand Fritsch 1987, Chen and Frank 1993, Emanuel 1993) has identified thatlarge thunderstorm systems (called mesoscale convective complexes [MCC])often produce an inertially stable, warm core vortex in the trailing altostratusdecks of the MCC. These mesovortices have a horizontal scale of approximately100 to 200 km, are strongest in the mid-troposphere and have no appreciablesignature at the surface. Zehr (1992) hypothesizes that genesis of thetropical cyclones occurs in two stages: stage one occurs when the MCC producesa mesoscale vortex and stage two occurs when a second blow up of convectionat the mesoscale vortex initiates the intensification process of loweringcentral pressure and increasing swirling winds.
Variations of the above broad-scale factors on the order of days, months,years and multi-decades determine how changes in tropical cyclone activityhave occurred in the past and will be manifested in the future.
a. Databases and climatology
Understanding tropical cyclone variability on interannual to interdecadaltimescales is hampered by the relatively short period over which accuraterecords are available.Figure 1 presents thevarious observational platforms available for analyzing tropical cycloneoccurrences in the Atlantic basin. Changes in the tropical cyclone databasesdue to observational platform improvements (and sometime degradations)can often be mistaken as true variations in tropical cyclone activity.For the Atlantic basin (including the North Atlantic Ocean, the Gulf ofMexico and the Caribbean Sea), aircraft reconnaissance has helped to providea nearly complete record back to the mid-1940s. The Northwest Pacific basin(i.e. the Pacific north of the equator and west of the dateline, includingthe South China Sea) also has had extensive aircraft surveillance givingvalid records going back to at least the late 1950s2,though this aircraft reconnaissance program was discontinued in 1987. However,for the remaining basins (the North Indian, the Southwest Indian, the Australian/SoutheastIndian, the Australian/South Pacific and the Northeast Pacific), routineaircraft reconnaissance has not been available and reliable estimates oftropical cyclones only exist for the satellite era beginning in the mid-1960s.Thus, with the instrumental record so limited, it is difficult to makeextensive analyses of trends and of the physical mechanisms responsiblefor the tropical cyclone variability on a global basis. Because of thislimitation, most studies on long-term changes in tropical cyclone activityhave focused upon the Atlantic and Northwest Pacific. However, even withthese limitations, some conclusions can be drawn about past variationsin all of the basins.
The averages and standard deviations over the last few decades for eachtropical cyclone basin are given inTable 1. Forexample, the Atlantic basin averages around 10 tropical cyclones reachingtropical storm strength and, of these, about 6 reach hurricane strength,comprising only about 12% of the global total. By far, the most activeregion is the Northwest Pacific with 27 tropical storms, of which 17 becomingtyphoons - over 30% of the global total. Overall, the global average numberof tropical cyclones reaching 18 m s-1 averages 86 with a rangeof (+ one standard deviation) from 78 to 94. Global hurricane-forcetropical cyclones average 47 yearly with a typical range from 41 to 54.Of particular interest are the tropical cyclones with winds of at least50 m s-1, as these intense tropical cyclones comprise a muchlarger proportion of the tropical cyclone-caused fatalities and destruction.In the Atlantic, for example, intense hurricanes account for only 21% ofall U. S. landfalling tropical cyclones, yet cause over 82% of the tropicalcyclone caused damage (Pielke and Landsea 1998). Intense hurricane-forcetropical cyclones are most common in the Northwest and Northeast Pacificbasins, making up nearly two-thirds of the average of 20 around the globe.
b. Interannual variability of tropical cyclones - a review
Seasonal variations of tropical cyclone activity depend upon changesin one or more of the parameters discussed in section II. Many studieshave focused upon the variations in these values both before and duringthe tropical cyclone season. While the bulk of these studies has been centeredupon the Atlantic basin, the interannual fluctuations in all of the globalbasins have been analyzed to some degree. A detailed survey paper of theinterannual variations of tropical cyclones, their causes and seasonalpredictability has been covered by Landsea (1999). What follows is a briefreview of the topic.
Globally, tropical cyclones are affected dramatically by the El Niño-SouthernOscillation (ENSO). ENSO is a fluctuation on the scale of a few years inthe ocean-atmospheric system involving large changes in the Walker andHadley Cells throughout the tropical Pacific Ocean region (Philander 1989).The state of ENSO can be characterized, among other features, by the SSTanomalies in the eastern/central equatorial Pacific: warmings in this regionare referred to as El Niño events and coolings are La Niñaevents.
In some basins, El Niña events bring increases in tropical cycloneformation (e.g. the South Pacific [Revell and Goulter 1986] and the NorthwestPacific between 160°E and thedateline [Chan 1985]) while others see decreases (e.g. the North Atlantic[Gray 1984a], the Northwest Pacific west of 160°E [Lander 1994], the Australian region [Nicholls 1979]). LasNiñas typically bring opposite conditions. These alterations intropical cyclone activity are due to a variety of ENSO effects: by modulatingthe intensity of the local monsoon trough, by repositioning the locationof the a monsoon trough and by altering the tropospheric vertical shear.
In addition to ENSO, three basins (the Atlantic [Gray 1984a], SouthwestIndian [Jury 1993], and Northwest Pacific [Chan 1995]) show systematicalterations of tropical cyclone frequency by the stratospheric Quasi-BiennialOscillation (QBO), an east-west oscillation of stratospheric winds thatencircle the globe near the equator (Wallace 1973). These relationshipsmay be due to alterations in the static stability and dynamics near thetropopause. Given the robustness of these alterations in tropical cycloneactivity that match the QBO phases, it appears unlikely that the associationsare purely chance correlations. More research is needed, however, to providea thorough explanation of these relationships.
Interannual tropical cyclone variations have also been linked to morelocalized, basin-specific features such as sea level pressures, local SST,monsoon strength and rainfall, sea level pressures and tropospheric verticalshear changes.
Sea level pressures changes in the Atlantic (Shapiro 1982; Gray 1984b)and Australian (Nicholls 1984) basins can force alterations in tropicalcyclogenesis frequency. Lower (higher) pressures are associated with less(more) vertical wind shear, weakened (enhanced) subsidence drying, anda stronger (weaker) intertropical convergence zone [ITCZ]/monsoon promotingincreased (decreased) tropical cyclone activity (Knaff 1997).
Sea surface temperatures in the genesis regions have both a direct thermodynamicand dynamic effect on tropical cyclones. In general, warmer than averagewaters are accompanied by decreased moist static stability, lower thanaverage surface pressures, and reduced shear. Cooler than average watersare usually found in conjunction with a stable troposphere, higher pressure,and increased shear. Somewhat surprisingly, interannual SST variationshave relatively small or negligible contributions toward increasing thetropical cyclone frequency in most basins. Only the Atlantic, SouthwestIndian and Australian regions have significant though small, positive associationsin the months directly before the tropical cyclone seasons begin (Raper1992, Shapiro and Goldenberg 1997). In the Atlantic basin, however, Saundersand Harris (1997) provide substantial evidence that both preceding andduring the hurricane season that SSTs in the "main development region"(i.e. between 10 and 20°N fromNorth Africa to Central America - Goldenberg and Shapiro 1996) contributea large percentage of the variance explained (over 30% during the heightof the season) with the number of hurricanes generated in that area.
One aspect recently uncovered is the association of a tropical cyclonebasin with its generating (or nearby) monsoon trough. Evans and Allen (1992)identified that variations in the Australian monsoonal flow can be associatedwith changes in tropical cyclone activity such that a strong (weak) monsooncirculation during La Niña (El Niño) events is accompaniedby many (few) tropical cyclones. Over the Atlantic basin, June throughSeptember monsoonal rainfall in Africa's Western Sahel has shown a veryclose association with intense hurricane activity (Gray 1990). Wet yearsin the Western Sahel (e.g. 1988 and 1989) are accompanied by dramatic increasesin the incidence of intense hurricanes, while drought years (e.g. 1990through 1993) are accompanied by a decrease in intense hurricane activity.Variations in tropospheric vertical shear and African easterly wave intensityhave been hypothesized as the physical mechanisms that link the two phenomena(Landsea and Gray 1992), although Goldenberg and Shapiro (1996) have demonstratedthat changes in the vertical shear probably dominate.
Some of this work has led to real-time seasonal forecasting efforts.The Atlantic basin has generated the most interest with predictions methodsdescribed in Gray (1984a, 1984b), Gray (1992, 1993, 1994), Elsner and Schmertmann(1993), Hess et al. (1995) and Lehmiller et al. (1997). Nicholls (1979,1984, 1992) has developed forecasts for the Australian basin as well. Currently,no other group has issued real-time forecasts for basinwide tropical cyclonesbased upon peer reviewed research.
c. Interdecadal variability of tropical cyclones
Among the basins with relatively short reliable records, Nicholls (1992)identified a downward trend in the numbers of tropical cyclones occurringin the Australian region from 105-160°E,primarily from the mid-1980s onward. However, a portion of this trend islikely artificial as the forecasters in the region no longer classify weak(greater than 990 hPa central pressure) systems as "cyclones" if the systemsdo not possess the traditional tropical cyclone inner-core structure, buthave the band of maximum winds well-removed from the center (Fig.2a - Nicholls et al. 1998). These changes in methodology around themid-1980s have been prompted by improved access to and interpretation ofdigital satellite data, the installation of coastal and off-shore radar,and an increased understanding of the differentiation of tropical cyclonesfrom monsoonal depressions (McBride 1987) and subtropical storms (Neumannet al. 1993). By considering only the moderate and intense (less than orequal to 990 hPa) tropical cyclones, this artificial bias in the cyclonerecord can be overcome.Figure2b shows that even with the removal of this bias in the weak tropicalcyclones that the frequency of the remaining moderate and strong tropicalcyclones has been reduced substantially over the years 1969/70-1995/96.(The intense tropical cyclones with minimum central pressure dropping below970 hPa has a very slight upward trend - not shown.) Nicholls et al. (1998)attribute the decrease in moderate cyclones to the occurrence of more frequentEl Niño occurrences during the 1980s and 1990s. However, the relativelysmall trend in the intense tropical cyclones implies that while ENSO modulatesthe total frequency of cyclones in the region, ENSO does not exert a controlon the intensity of the systems after formation.
For the remaining short record basins based upon data from the late1960s onwards, the Northeast Pacific has experienced a significant upwardtrend in tropical cyclone frequency, the North Indian a significant downwardtrend, and no appreciable long-term variation was observed in the SouthwestIndian and Southwest Pacific (east of 160°E) for the total number of tropical storm strength cyclones(from Neumann 1993). However, whether these represent longer term ( > 30years) or shorter term (on the scale of tens of years) variability is completelyunknown because of the lack of a long, reliable record.
For the Northwest Pacific basin, Chan and Shi (1996) found that boththe frequency of typhoons and the total number of tropical storms and typhoonshave been increasing since about 1980 (Fig. 3).This recent trend holds true whether the curvilinear fit is utilized forthe years 1972-1994 or on the whole 1959-1994 time series. However, theincrease was preceded by a nearly identical magnitude of decrease fromabout 1960 to 1980. It is unknown currently what has caused these decadal-scalechanges. Additionally, no analysis has been done as of yet on the numbersof intense typhoons (winds at least 50 m s-1) because of anunremoved overestimation bias in the intensity of such storms in the 1950sand 1960s (Bouchard 1990, Black 1993).
There has been an extensive analysis of the North Atlantic basin duein part to the reliable record for both the entire basin (back to 1944)and U. S. landfallings (back to 1899)3.Similar to the problems with the Northwest Pacific data, the all-basindata also has had a bias in the measurement of strong hurricanes: duringthe mid-1940s through the late 1960s, the intensity of strong hurricaneswas likely overestimated by 2.5-5 m s-1 (Landsea 1993). Thisbias has been crudely removed to provide estimates of the true occurrenceof intense (or major) hurricanes. No estimate of the true occurrence ofall-basin intense hurricanes is attempted for the era before the mid-1940sbecause of the lack of reliable data on the strong inner core of the hurricanesexcept for very infrequent measurements conducted by unlucky ships' crews.The U.S. landfalling hurricane records back to the turn of the centuryare very reliable as opposed to open-water storms because of the use ofactual central pressure measurements at landfall (Jarrell et al. 1992).
Examination of the record for the Atlantic numbers of tropical storms(including those designated as subtropical storms41968 onward) shows substantial yearly variability, but no significant trend(Fig. 4). In contrast, the numbers of intensehurricanes have gone through pronounced multidecadal changes: active duringthe late 1940s through the mid-1960s, quiet from the 1970s through theearly 1990s, and then a shift again to busy conditions again during theextraordinarily active years 1995 and 1996(Fig.5). Concurrent with these frequency changes, there have been periodsof strong mean intensity of the Atlantic tropical cyclones (mid-1940s-1960sand 1995-1996) and weak mean intensity (1970s-early 1990s), though therehas been no significant change in the peak intensity reached by the strongesthurricane each year (Landsea et al. 1996a).
These trends for the entire Atlantic basin are mirrored by those intensehurricanes striking the U. S. East Coast, from the Florida peninsula throughNew England(Fig. 6). The quiet period of the1970s to the early 1990s is similar to a quiescent regime in the firsttwo decades of this century. A more active regime began in the mid-1920sand continued into the 1960s, with a peak in landfalling intense hurricanesfrom the 1940s through the mid-1960s. During two particularly busy periods,the Florida peninsula and the Carolinas to New England each experiencedseven intense hurricane landfalls in seven years (1944-1950 and 1954-1960,respectively). Other regions within the Atlantic basin - such as the CaribbeanSea and surrounding land masses - also have experienced these multidecadalchanges with even greater amplitude(Fig. 7).In contrast, a subset of the Atlantic basin consisting of the U. S. GulfCoast from Texas to the Florida panhandle(Fig.8) has observed much weaker multidecadal variability in intense hurricanestrikes. Going back even farther into the historical records, Fernández-Partagásand Diaz (1996) estimate that the overall Atlantic tropical storm and hurricaneactivity for the years 1851-1890 was 12% lower than the corresponding fortyyear period of 1951-1990, though little can be said regarding the intensehurricanes.
Finally, hurricane-caused damage in the United States - when properlynormalized - can also provide an independent indication of multiyear changesin tropical cyclone activity. Pielke and Landsea (1998) standardized theamount of U. S. destruction from tropical cyclones by taking into accountinflation, coastal county population changes and trends in personal propertyamounts.Figure 9 shows the time series of normalizeddamage amounts when these three factors are taken into account. Note theextreme destruction in 1926 (due to the near worst case scenario of a largeCategory 4 hurricane striking first the Miami-Ft. Lauderdale region inFlorida, then the Florida panhandle and Alabama as a Category 3 hurricane),lowered values of damage in the early and mid-1930s followed by $3-7 billiondamage per year for nearly every five year period from the late 1930s untilthe late 1960s. During the 1970s and 1980s, the normalized damage in theUnited States was substantially smaller ($1-3 billion per year) than inearlier decades. During the first five years of the 1990s, damage againreturned to higher levels due primarily to the destructiveness of HurricaneAndrew in 1992.
Gray (1990) and Gray et al. (1997) have attributed these multidecadalvariations in intense hurricane activity to changes in the Atlantic SSTstructure. Warmer (cooler) than average conditions in the Atlantic northof the equator coupled with cooler (warmer) than average SSTs in the SouthAtlantic favor increased (decreased) intense hurricane activity. Such adipole structure of the Atlantic SSTs also forces drought and wet periodsin the North Africa's Western Sahel (Fig. 10,Folland et al. 1986), which at least partially explains why there is astrong concurrent link between the year-to-year Sahel rainfall variationsand intense Atlantic hurricanes (Reed 1988, Gray 1990, Landsea and Gray1992, Landsea et al. 1992). The SST dipole pattern appears to alter theoverlaying tropospheric circulation such that warm North/cold South Atlanticconditions correspond to reduced vertical wind shear in the main developmentregion favoring the formation and intensification of tropical cyclones.In contrast, a cool North/warm South Atlantic acts in concert with enhancedtradewind easterlies and upper tropospheric westerlies and thus increasedtropospheric vertical wind shear (Gray et al. 1997). Additionally, theseSST variations likely play a direct role in providing changes of the heatinput available to the incipient tropical cyclone by changing the boundarylayer moist enthalpy values (Saunders and Harris 1997, Landsea et al. 1998).
The strong sensitivity of Atlantic intense hurricanes to these changeswhile the frequency of named storms remains relatively constant is likelydue the formation differences between the two. The vast majority of Atlanticintense hurricanes develop from easterly waves exiting the North Africancoast and moving across the tropical North Atlantic (Landsea 1993). Conditionsthroughout the main development region are usually unfavorable for anytropical cyclone to form and intensify, so typically the most that is realizedis a tropical storm or a weak hurricane (Gray et al. 1993). In active intensehurricane years such as 1995 and 1996, the vertical shear is lowered andthe SSTs are warmer along the main development region allowing a few easterlywaves to develop up to intense hurricanes (Goldenberg and Shapiro 1996,Saunders and Harris 1997). In contrast during quiet seasons for intensehurricanes such as 1991 through 1994, tropical storms can occur in relativeabundance in the subtropical latitudes (20-40°N) forming from upper level lows, stationary frontal boundariesand easterly waves that survive the hostile tropical latitudes.
The lack of a distinct multidecadal variation of intense hurricanesin the Gulf of Mexico is likely due to local conditions that dominate overthese basinwide SST changes (Landsea et al. 1992). Since 1967 (when satellitemonitoring made it possible), only intense hurricanes that were spawnedfrom easterly waves have made landfall along the U.S. East Coast, whilemid-latitude systems (e.g. stationary frontal boundaries or upper-troposphericcutoff lows) can occasionally form an intense hurricane that makes landfallalong the U.S. Gulf Coast. Hurricane Alicia, which struck the Texas coastin 1983, is a notable example of this latter phenomena. Additionally, verticalshear changes in the Gulf are not correlated highly with variations ofENSO or West Sahel rainfall, unlike the main development region (Goldenbergand Shapiro 1996).
IV. Paleotempestology - the prehistoric record of tropical cyclones
The study of pre-historic tropical cyclones, or "paleotempestology"as it could be called, may be a way to extend back these records to providemeasures of longer-term tropical cyclone climate variability. Recent effortsto address this issue with a variety of creative methodologies includeexamining: shallow coastal lake bed cores to locate storm surge sand layers(Liu and Fearn 1993), cyclone-produced sediment deposits in shallow offshorewaters (Keen and Slingerland 1993), pollen changes recorded in coastalforest floors due to canopy blow down (Bravo et al. 1997) and oxygen isotopevariations found in coastal cavern stalactites5(Malmquist 1997). Liu and Fearn's (1993) work suggests that at one particularlocation in coastal Alabama, U.S., there were strikes by Saffir-SimpsonCategory 4 or 5 hurricanes at around 3400, 2800, 2200, 1300 and 700 yearsago (Fig. 11) implying an annual probabilityof occurrence of about 0.17%. Before about 3400 years ago, there is noevidence in the sediment record for Category 4 or 5 hurricanes, meaningthat either the climate did not allow for such strong hurricanes to occur,that the tracks of such hurricanes were altered away from this Gulf ofMexico location, and/or the geomorphology of the region changed so thatthe technique could not provide an accurate measure of such strong hurricanesbefore this time. These methods provide promise in extending records oftropical cyclones well beyond the current few decades of reliable standardhistorical data, provided that they are able to be calibrated accuratelyagainst hurricanes that occurred in the instrumental record.
V. Future Climate - how tropical cyclones may change in coming years
a. Extrapolation of past variations
As a first approximation of tropical cyclone activity in the next decadeor two, one can simply extrapolate past variations in the data - that isassuming that such trends are not artificially-induced and that a quasi-periodicityactually exists in the cyclone activity. Three basins - the Australian,the Northwest Pacific and the Atlantic - have been examined in enough detailpossibly to allow some suggestions for what the late 1990s and first decadeof the 21st century may bring.
In the Australian basin as detailed earlier, Nicholls et al. (1997)identified an substantial downward trend in the numbers of tropical cyclonesover the period of 1969/70 through 1995/96, the non-artificial portionof which is linked to having more frequent El Niño events duringthe late 1970s through the early 1990s than earlier. If more frequent ElNiño events were to continue in the coming decades, then the Australianregion would likely continue to receive fewer than normal tropical cyclones.The continuation in the trend in ENSO is dependent upon its cause. Onepossibility is that the increased El Niño activity is due to naturalvariability of the ocean-atmosphere system (e.g. Gray et al. 1997). However,Trenberth and Hoar (1996) suggest that the extremely long-running El Niñoevent of late 1990 through early 1995 was not due to natural fluctuations,but instead may be due to climate changes associated with increases ingreenhouse gases. Such a statement is not supported by general circulationmodel (GCM) simulations because GCMs characterize ENSO variability poorly.Thus until the cause of the trend in ENSO is known, suggestions that therewill be a continuation of frequent El Niño events resulting in fewerAustralian tropical cyclones for the next decade or two is probably notvery prudent.
Chan and Shi (1996) uncovered a decrease in Northwest Pacific typhoonsand total number of tropical storms from the late 1950s through the late1970s, followed by a nearly comparable increase from the early 1980s untilthe mid-1990s. However, since the mechanism for these variations is unknown,a further extrapolation of the increase in the 1980s and 1990s into thefuture would also be unfounded.
The one region where it may be possible to make a reasonable assessmentof future climate trends is the Atlantic, because of the multidecadal variationsin hurricane activity that have been described and to some extent understood.As described above, while the total number of tropical storms and hurricanesdo not vary greatly on a multidecadal time scale, the intense hurricanesshow a strong variation. More numerous intense hurricanes occur, such asin the decades of the 1940s through the 1960s, while the North Atlanticis warmer than average and the South Atlantic is cooler. Converse conditionsof few intense hurricanes were observed in the 1970s through the early1990s while the North Atlantic was cool and the South Atlantic was warm.It has been hypothesized (Gray et al. 1997) that these multidecadal oceanictemperature and hurricane changes are regulated by the strength of thethermohaline circulation and North Atlantic deep water formation - portionsof the global "Great Ocean Conveyor" (Broecker 1991). Given that the Saheldrought and wet regimes also occur in conjunction with the Atlantic intensehurricane quiet and active periods, respectively (Gray 1990, Landsea etal. 1992), and that the Sahel has experienced several multidecadal periodsof wet and dry conditions over at least the last few hundreds of years(Nicholson 1989), it stands to reason that these fluctuations are a naturalmanifestation of the ocean-atmospheric system and that an end to the Saheldrought and Atlantic hurricane quiet period of the 1970s-early 1990s wouldsoon come to an end. In fact, Gray (1990) predicted as much:
"If these past variations are a reasonable indication of the future,then we should expect an eventual recurrence of somewhat heavier WesternSahel precipitation, possibly during the 1990s and the early years of the21st century. With such a rainfall increase, we should also expect a returnof more frequent intense hurricane activity in the Caribbean Basin andalong the U. S. coastline."
The hyperactive Atlantic hurricane seasons of 1995 and 1996 with a totalof 32 named storms, 20 hurricanes and especially 11 intense hurricanesmay indicate the start of such a return to active conditions (Gray et al.1996, Goldenberg et al. 1997). The 11 intense hurricanes over two yearsrepresents a 450% increase over the frequency of intense hurricanes during1991-1994 and a 139% increase over the long-term (1950-1990) average of2.2 intense hurricanes per year. Along with the increase in hurricane activity,the West Sahel rainfall has returned to near average conditions for 1994-1996,the first three year stretch of near to above normal rainfall since 1965-1967(Landsea et al. 1996b). Corresponding to, and most likely leading, thesechanges in the Atlantic intense hurricane and West Sahel rainfall, arerather dramatic increases in the North Atlantic SSTs from 5°N to 60°N anda cooling of the South Atlantic SSTs from 5°N to 50°S (Grayet al. 1996). It is also possible that such changes were beginning to occurin 1988-1989 - which were two years of high West Sahel rainfall and activeAtlantic hurricane seasons - but that the highly anomalous long-runningEl Niño event of late 1990 through early 1995 acted to mask theenhancing effects of the Atlantic SSTs (Goldenberg et al. 1997). More indepth research is needed to better define if indeed this change in theAtlantic SSTs with the attendant effects on Atlantic hurricanes and Sahelrainfall has occurred; if it has switched, when the change took place;and how long would an active intense hurricane regime stay in place.
b. The effects of anthropogenic global warming
Two impacts of anthropogenic climate change due to increasing amountsof "greenhouse" gases that may occur (Houghton et al., 1990, 1992, 1996)are increased tropical sea surface temperatures(Fig.12) and increased tropical rainfall associated with a slightly strongerITCZ(Fig. 13). Note in these figures the 0.5to 1.5°C warming of the tropicaland subtropical SSTs and an overall increase in the ITCZ precipitationnear the equator, though the precipitation changes show a "noisier" signal.Because of these possible changes, there have been many suggestions basedupon global circulation and theoretical modeling studies that increasesmay occur in the frequency (AMS Council and UCAR Board of Trustees 1988;Houghton et al. 1990; Broccoli and Manabe 1990; Ryan et al. 1992; Haarsmaet al. 1993), area of occurrence (Houghton et al. 1990; Ryan et al. 1992),mean intensity (AMS Council and UCAR Board of Trustees 1988; Haarsma etal. 1993), and maximum intensity (Emanuel 1987; AMS Council and UCAR Boardof Trustees 1988; Houghton et al. 1990; Haarsma et al. 1993; Bengtssonet al. 1996) of tropical cyclones. In contrast, there have been some conclusionsthat decreases in frequency may result (Broccoli and Manabe 1990; Bengtssonet al. 1996). Finally, one report concluded that any changes in frequencyor intensity due to increased greenhouse gases would be "swamped" by thelarge natural variability (Lighthill et al. 1994). As discussed earlier,there is currently no evidence at the present time that there have alreadybeen systematic changes in the observed tropical cyclones around the globe.
Any changes in tropical cyclone activity are intrinsically tied in withlarge-scale changes in the tropical atmosphere. One key feature that hasbeen focused upon has been possible changes in SSTs. However, SSTs by themselvescannot be considered without corresponding information regarding the moistureand stability in the tropical troposphere. What has been identified inthe current climate as being necessary for genesis and maintenance fortropical cyclones (e.g. SSTs of at least 26.5°C) would likely changein a 2 x CO2 world because of possible changes in the moistureor stability. It is quite reasonable that an increase in tropical and subtropicalSSTs would be also accompanied by an increase in the SST threshold valueneeded for cyclogenesis because of compensating changes in the troposphericmost static stability (Emanuel 1995). Such difficulties then make it problematicto address the issue as Ryan et al. (1992) did in using Gray's (1979) "GenesisParameter" to diagnose changes in large-scale fields from GCM output fortropical cyclone frequency and area of occurrence issues. Indeed, Wattersonet al. (1995) found that Gray's Genesis Parameter, while quite useful fordiagnosis of the mean climatology of tropical cyclone frequency and areaof occurrence, was not able to correctly anticipate interannual fluctuationsin tropical cyclone activity and thus, probably would not be useful foranalysis of future climate states.
Additionally, besides the thermodynamic variables, changes in the tropicaldynamics will also play a big role in determining changes in tropical cycloneactivity. For example, if the vertical wind shear over the tropical NorthAtlantic moderately increased (decreased) during the hurricane season ina 2 x CO2 world, then we would see a significant decrease (increase)in activity because this particular basin is marginal for tropical cycloneactivity. Another large unknown is how the monsoonal circulations may change.If the monsoons become more active, then it is likely that more tropicalcyclones in the oceanic monsoon regions would result. In contrast to otherGCM results (e.g. the "variable cloud" run in Broccoli and Manabe 1990,Haarsma et al. 1993, etc.), Bengtsson et al. (1996) show that a GCM climatewith doubled carbon dioxide amounts compared with pre-industrialized valuesproduces substantially fewer tropical cyclones around the world becauseof a weakened ITCZ and monsoonal circulations. However, the downscalingtechnique utilized (i.e. a high resolution atmospheric GCM run for fiveyears run from the SST boundary conditions from the 60th year of a lowresolution GCM run) appears to be flawed because the ITCZ response to increasedcarbon dioxide was actually opposite in the low resolution model (Landsea1997b), thus calling into question the validity of Bengtsson et al.'s results.
One last final wild card in all of this is how ENSO may change in a2 x CO2 world, as ENSO is the largest single factor controllingyear-to-year variability of tropical cyclones globally (Landsea 1997a).If El Niño events occur more often or with more intensity, thenthe inhabitants along the Atlantic basin and in Australia would likelyhave fewer tropical cyclones to worry about, whereas people living in theSouth Central Pacific would have more storms to prepare for. The reversewould be true if La Niña events became more prevalent. As describedearlier, El Niño events indeed have become more frequent in occurrenceduring the most recent two decades, actually resulting in some of the changesnoted above. It is currently unknown whether this trend toward more ElNiño events is simply natural variability or is due to anthropogenicforcing.
Overall, it is difficult to assess globally how changes of tropicalcyclone intensities (both the mean and the maximum), frequencies, and areaof occurrence may change in a 2 x CO2 world. It is because ofthis uncertainty that the 1995 Intergovernmental Panel on Climate Changeassessment (Houghton et al. 1996) came out with this straightforward admission:
"The formation of tropical cyclones depends not only on seasurface temperature (SST), but also on a number of atmospheric factors.Although some models now represent tropical storms with some realism forpresent day climate, the state of the science does not allow assessmentof future changes."
Clearly, much more investigation is needed to narrow down the uncertaintiesthat are currently in this field of tropical cyclone climate change. Currently,there is no convincing evidence that there will be a systematic increasein the tropical cyclone mean intensity, maximum intensity or frequencydue to increases in "greenhouse gases". Nor, for that matter, is therestrong evidence for decreases in hurricane, typhoons and tropical cyclones.It may turn out that changes around the globe will not be consistent; someregions may experience more activity, others less.
VI. Summary
Tropical cyclones - including hurricanes and typhoons - have been andcontinue to be extremely disruptive events for inhabitants in the globaltropics and subtropics. Knowledge of how and why the characteristics ofthese coupled ocean-atmospheric systems have changed in the past is a topicof much interest. With a more complete understanding, we will be betterprepared to answer the question: "How will tropical cyclone activity changein future years?".
Progress is being made in analyzing both the interannual fluctuationsof tropical cyclone activity (see review by Landsea 1997a) as well as themultidecadal facet. In this chapter, three basins - the Australian, theNorthwest Pacific and the Atlantic - have been examined in detail. In theremaining basins, reliable records are too short to demonstrate reliabletrends. For long-term trends in total frequency of events, the Australianbasin has shown a decline (since the late 1960s), the Northwest Pacificis now showing an increase after experiencing a decrease in frequency fromthe late 1950s through 1980, while the Atlantic has been fairly constantsince the mid-1940s. For mean intensity, the Australian basin has verylittle trend, the Northwest Pacific has shown a downward trend during the1960s and 1970s and an upward trend in intensity of events since, and theAtlantic has been observed to have a quasi-cyclic multidecadal regime thatalternates between active and quiet phases of mean intensity on the scaleof 25-40 years each. Such variations in Atlantic hurricanes are mirroredby normalized destruction amounts that occurred in the United States. Formaximum intensity, only the Atlantic has been examined revealing no substantialtrend or consistent variation in the strongest hurricane each year.
While the multidecadal variations for the Northwest Pacific currentlyhave no explanation, there exist plausible reasons for the changes in regimesin the Australian and Atlantic basins. The decline in the Australian tropicalcyclones are due to increasing El Niño events during the late 1970sthrough the early 1990s. The quiet decades of the 1970s to the early 1990sfor intense Atlantic hurricanes are likely due to changes in the AtlanticOcean SST structure with cooler than usual waters in the North Atlanticand warmer in the South Atlantic. The reverse situation of a warm NorthAtlantic and a cool South Atlantic was present during the active 1940sthrough the 1960s. A natural fluctuation of the Great Ocean Conveyer andthe associated North Atlantic deep water formation has been hypothesizedas being responsible for such SST and Atlantic hurricane changes.
Extrapolation of these multidecadal variations into the future is uncertain.Until an explanation is found for the Northwest Pacific upward trend infrequency and intensity (as well as the downward fall in both during thelate 1950s through the late 1970s), a decadal forecast is doubtful. Forthe Australian basin, if one could be assured that the more frequent ElNiño events would continue, then a continued low number of tropicalcyclones would be expected for this area. However, the mechanism for suchEl Niño changes over the past couple of decades is not understoodand thus, a decadal scale forecast of Australian cyclones would not beprudent. The only basin that might be reliably forecast is the Atlanticbecause of the large, apparently natural fluctuations of the Atlantic SSTson a multidecadal timescale. It is possible that the hyperactive seasonsof 1995 and 1996 signal a return of the active regime (of an unknown duration)to the Atlantic, though more research is needed to confirm or deny sucha hypothesis.
Over even longer timescales, the question has been raised as to thepossible impact of anthropogenic global warming on tropical cyclones aroundthe world. Unfortunately, due to our inability to simulate tropical cycloneson the scale needed within the context of a GCM, because of conflictingmodel results, and due to our lack of knowledge about the processes oftropical cyclogenesis and intensification, there is no convincing evidencefor systematic changes to occur in the frequency, mean intensity, maximumintensity, and area of occurrence of tropical cyclones. Indeed, lookingfor a systematic global signal common to all tropical cyclone basins isnot the most reasonable approach. Because of strong links with global phenomenasuch as the El Niño-Southern Oscillation, tropical cyclone activityin various basins is not independent of one another. An increase in activityin one region may be instead be accompanied by a decrease in tropical cycloneactivity in another basin. It will take continued efforts toward increasingour understanding before more definitive answers are available for theglobal warming question.
VII. Acknowledgments I would like to acknowledge the helpfulsupport and encouragement of Hugh Willoughby and Stan Goldenberg here atthe NOAA/AOML/Hurricane Research Division. Bill Gray of Colorado StateUniversity has also sparked many useful and enlightening discussions onthe topic. Roger Pielke, Sr., Roger Pielke, Jr. and three anonymous reviewersprovided quite helpful comments that clarified and enhanced this reviewchapter. The author thanks the Bermuda Biological Research Station's RiskPrediction Initiative for providing financial support through a grant onthe topic of interannual tropical cyclone variability.
VIII. References
American Meteorological Society (AMS) Council and University Corporationfor Atmospheric Research (UCAR) Board of Trustees, 1988: The changing atmosphere- challenges and opportunities.Bull. Amer. Meteor. Soc.,69,1434-1440.
Bengtsson, L., M. Botzet and M. Esch, 1996: Will greenhouse gas-inducedwarming over the next 50 years lead to higher frequency and greater intensityof hurricanes?".Tellus48A, 57-73.
Black, P. G., (1993): Evolution of maximum wind estimates in typhoons.Tropical Cyclone Disasters, J. Lighthill, Z. Zhemin, G. Holland,and K. Emanuel, Eds. Peking University Press, Beijing, 104-115.
Bouchard, R. H. 1990. A climatology of very intense typhoons: Or wherehave all the supertyphoons gone?.1990 Annual Tropical Cyclone Report,U.S. Naval Oceanography Command Center, Joint Typhoon Warning Center, COMNAVMARIANAS,PSC 489, Box 12, FPO San Francisco CA, 96630-2926, U.S., 266-269.
Bravo, J., J. P. Donnelly, J. Dowling, T. Webb, III, 1997: Sedimentaryevidence for the 1938 hurricane in southern New England.Preprints ofthe 22nd Conference on Hurricanes and Tropical Meteorology, Amer. Meteor.Soc., Fort Collins, CO, 395-396.
Broccoli, A. J., and S. Manabe, 1990: Can existing climate models beused to study anthropogenic changes in tropical cyclone climate?Geophys.Res. Letters,17, 1917-1920.
Broecker, W. S., 1991: The great ocean conveyor.Oceanography,4,79-89.
Chan, J. C. L., 1985: Tropical cyclone activity in the Northwest Pacificin relation to the El Nino / Southern Oscillation phenomenon.Mon. Wea.Rev.,113, 599-606.
Chan, J. C. L., 1995: Tropical cyclone activity in the western NorthPacific in relation to the stratospheric quasi-biennial oscillation.Mon.Wea. Rev.,123, 2567-2571.
Chan, J. C. L. and J. Shi, 1996: Long-term trends and interannual variabilityin tropical cyclone activity over the western North Pacific.Geo. Res.Letters,23, 2765-2767.
Chen, S. A., and W. M. Frank, 1993: A numerical study of the genesisof extratropical convective mesovortices. Part I: Evolution and dynamics.J. Atmos. Sci.,50, 2401-2426.
Elsner, J. B., and C. P. Schmertmann, 1993: Improving extended-rangeseasonal predictions of intense Atlantic hurricane activity.Wea. Forecasting,8, 345-351.
Emanuel, K. A., 1987: The dependence of hurricane intensity on climate.Nature,326, 483-485.
Emanuel, K. A., 1993: The physics of tropical cyclogenesis over theEastern Pacific.Tropical Cyclone Disasters. J. Lighthill, Z. Zhemin,G. J. Holland, K. Emanuel, (Eds.), Peking University Press, Beijing, 136-142.
Emanuel, K. A., 1995: Comments on "Global climate change and tropicalcyclones": Part I.Bull. Amer. Meteor. Soc.,76, 2241-2243.
Evans, J. L., and R. J. Allen, 1992: El Niño/Southern Oscillationmodification to the structure of the monsoon and tropical activity in theAustralian region.Int. J. Climatol.,12, 611-623.
Fernández-Partagás, J., and H. F. Diaz, 1996: Atlantichurricanes in the second half of the 19th Century.Bull. Amer. Meteor.Soc.,77, 2899-2906.
Folland, C. K., T. N. Palmer, and D. E. Parker, 1986: Sahel rainfalland worldwide sea temperatures, 1901-1985.Nature,320, 602-607
Goldenberg, S. B. and L. J. Shapiro, 1996: Physical mechanisms for theassociation of El Niño and West African rainfall with Atlantic majorhurricane activity.J. Climate,9, 1169-1187.
Goldenberg, S. B., L. J. Shapiro, and C. W. Landsea, 1997: Are we seeinga long-term upturn in Atlantic basin major hurricane activity related todecadal-scale SST fluctuations?Preprints of the Seventh Conferenceon Climate Variations, Amer. Meteor. Soc, Long Beach, CA, 305-310.
Gray, W. M., 1968: Global view of the origins of tropical disturbancesand storms.Mon. Wea. Rev.,96, 669-700.
Gray, W. M., 1979: Hurricanes: Their formation, structure and likelyrole in the tropical circulation.Meteorology Over the Tropical Oceans.D. B. Shaw (Ed.), Roy. Meteor. Soc., James Glaisher House, Grenville Place,Bracknell, Berkshire, RG12 1BX, 155-218.
Gray, W. M., 1984a: Atlantic seasonal hurricane frequency: Part I: ElNiño and 30 mb quasi-biennial oscillation influences.Mon. Wea.Rev.,112, 1649-1668.
Gray, W. M., 1984b: Atlantic seasonal hurricane frequency: Part II:Forecasting its variability.Mon. Wea. Rev.,112, 1669-1683.
Gray, W. M., 1990: Strong association between West African rainfalland US landfall of intense hurricanes.Science,249, 1251-1256.
Gray, W. M., C. W. Landsea, P. W. Mielke, Jr., and K. J. Berry, 1992:Predicting Atlantic seasonal hurricane activity 6-11 months in advance.Wea. Forecasting,7, 440-455
Gray, W. M., C. W. Landsea, P. W. Mielke, Jr., and K. J. Berry, 1993:Predicting Atlantic basin seasonal tropical cyclone activity by 1 August.Wea. Forecasting,8, 73-86.
Gray, W. M., C. W. Landsea, P. W. Mielke, Jr., and K. J. Berry, 1994:Predicting Atlantic basin seasonal tropical cyclone activity by 1 June.Wea.Forecasting,9, 103-115.
Gray, W. M., J. D. Sheaffer, and C. W. Landsea, 1997: Climate trendsassociated with multidecadal variability of Atlantic hurricane activity.Hurricanes, Climate and Socioeconomic Impacts, Edited by H. F. Diazand R. S. Pulwarty, Springer, Berlin, 15-53.
Guard, C. P., L. E. Carr, F. H. Wells, R. A. Jeffries, N. D. Gural,and D. K. Edson, 1992: Joint Typhoon Warning Center and the challengesof multibasin tropical cyclone forecasting.Wea. Forecasting,7,328-352.
Haarsma, R. J.,Mitchel, J. F. B. and C. A. Senior, 1993: Tropical disturbancesin a GCM.Clim. Dyn.,8, 247-257.
Hebert, P. J., J. D. Jarrell, and M. Mayfield, 1997: The deadliest,costliest, and most intense United States hurricanes of this century (andother frequently requested hurricane facts).NOAA Tech. Memo., NWSTPC-1, Miami, Florida, 30pp.
Hess, J. C., J. B. Elsner, and N. E. LaSeur, 1995: Improving seasonalpredictions for the Atlantic basin.Wea. Forecasting,10,425-432.
Holland, G. J., 1993: Ready Reckoner - Chapter 9,Global Guide toTropical Cyclone Forecasting, WMO/TC-No. 560, Report No. TCP-31, WorldMeteorological Organization, Geneva.
Houghton, J. T., G. J. Jenkins and J. J. Ephramus, Eds., 1990:ClimateChange: The IPCC Scientific Assessment. Cambridge University Press,New York, 364 pp.
Houghton, J. T., B. A. Callander and S. K. Varney, Eds., 1992:ClimateChange 1992: The Supplementary Report to the IPCC Scientific Assessment.Cambridge University Press, New York, 198 pp.
Houghton, J. T., L. G. Meira Filho, B. A. Callander, N. Harris, A. Kattenberg,and K. Maskell, Eds., 1996:Climate Change 1995: The Science of ClimateChange. Contribution of WGI to the Second Assessment Report of theIntergovernmental Panel on Climate Change. Cambridge University Press,New York, 572 pp.
Jarrell, J. D., P. J. Hebert, and M. Mayfield, 1992: Hurricane experiencelevels of coastal county populations from Texas to Maine.NOAA Tech.Memo. NWS NHC 46, Coral Gables, Florida, 152 pp.
Jarvinen, B. R., C. J. Neumann, and M. A. S. Davis, 1984: A tropicalcyclone data tape for the North Atlantic Basin, 1886-1983: Contents, limitations,and uses.NOAA Tech. Memo. NWS NHC 22, Coral Gables, Florida, 21pp.
Joint Typhoon Warning Center (JTWC), 1974:1974 Annual Typhoon Report.U.S. Fleet Weather Central, Joint Typhoon Warning Center, COMNAVMARIANAS,Box 17, FPO San Francisco, CA, 96630, U.S., 116 pp.
Jury, M., 1993: A preliminary study of climatological associations andcharacteristics of tropical cyclones in the SW Indian Ocean.Met. Atmos.Physics,51, 101-115.
Keen, T. R., and R. L. Slingerland, 1993: Four storm-event beds andthe tropical cyclones that produced them: A numerical hindcast.J. Sed.Petroe.,63, 218.
Knaff, J. A., 1997: Implications of summertime sea level pressure anomaliesin the tropical Atlantic region.Mon. Wea. Rev.,10, 789-804.
Lander, M., 1994: An exploratory analysis of the relationship betweentropical storm formation in the Western North Pacific and ENSO.Mon.Wea. Rev.,122, 636-651.
Landsea, C. W., 1999: El Niño-Southern Oscillation and the seasonalpredictability of tropical cyclones. Accepted as a chapter forEl Niño:Impacts of Multiscale Variability on Natural Ecosystems and Society,edited by H. F. Diaz and V. Markgraf (in press).
Lawrence, J. R., and S. D. Gedzelman, 1996: Low stable isotope ratiosof tropical cyclone rains.Geophys. Res. Let.,23, 527-530.
Lawrence, M. B., and J. M. Pelissier, 1981: Atlantic hurricane seasonof 1980.Mon. Wea. Rev.,109, 1567-1582.
Lehmiller, G. S., T. B. Kimberlain, and J. B. Elsner, 1997: Seasonalprediction models for North Atlantic basin hurricane location.Mon.Wea. Rev., 125, 1780-1791.
Lighthill, J., G. Holland, W. Gray, C. Landsea, G. Craig, J. Evans,Y. Kurihara, and C. Guard, 1994: Global climate change and tropical cyclones.Bull. Amer. Meteor. Soc.,75, 2147-2157.
Liu, K.-B., and M. L. Fearn, 1993: Lake-sediment record of late Holocenehurricane activities from coastal Alabama.Geology,21, 793-796.
Ludlum, D. M., 1989:Early American Hurricanes 1492-1870. LancasterPress, Inc., Lancaster, Pennsylvania, 198 pp.
Malmquist, D. L., 1997: Oxygen isotopes in cave stalagmites as a proxyrecord of past tropical cyclone activity.Preprints of the 22nd Conferenceon Hurricanes and Tropical Meteorology, Amer. Meteor. Soc., Fort Collins,CO, 393-394.
McBride, J. L., 1987: The Australian summer monsoon.Reviews of MonsoonMeteorology, C. P. Chang and T. N. Krishnamurti, Eds., Oxford UniversityPress, 203-231.
Neumann, C. J., 1993: Global Overview,Global Guide to Tropical CycloneForecasting, WMO/TC-No. 560, Report No. TCP-31, World MeteorologicalOrganization, Geneva, 1-1 - 1-43.
Neumann, C. J., B. R. Jarvinen, C. J. McAdie and J. D. Elms, 1993:Tropicalcyclones of the North Atlantic Ocean, 1871-1992, National ClimaticData Center in cooperation with the National Hurricane Center, Coral Gables,FL, 193 pp.
Nicholson, S. E., 1989: Long-term changes in African rainfall.Weather,44, 46-56.
Nicholls, N., 1979: A possible method for predicting seasonal tropicalcyclone activity in the Australian region.Mon. Wea. Rev.,107,1221-1224.
Nicholls, N., 1984: The Southern Oscillation, sea-surface temperature,and interannual fluctuations in Australian tropical cyclone activity.J.Climatol.,4, 661-670.
Nicholls, N., 1992: Recent performance of a method for forecasting Australianseasonal tropical cyclone activity.Aust. Met. Mag..40,105-110.
Nicholls, N., C. W. Landsea, J. Gill, 1998: Recent trends in Australianregion tropical cyclone activity.Meteor. Atmos. Phys.65197-205.
National Oceanic and Atmospheric Administration (NOAA), 1997: WSOM Chapter-41,Tropical Cyclone Program. W/OM12. NOAA, Washington, D. C., 59 pp.
Philander, S. G. H., 1989:El Niño, La Niña, and theSouthern Oscillation., Academic Press, New York, 293 pp.
Pielke, R. A., Jr., and C. W. Landsea, 1998: Normalized Atlantic hurricanedamage, 1925-1995. Submitted toWea. Forecasting.
Pielke, R. A., Jr., and R. A. Pielke, Sr., 1997:Hurricanes: TheirNature and Impacts on Society., in preparation.
Raper, S., 1992: Observational data on the relationships between climatechange and the frequency and magnitude of severe tropical storms. InClimateand sea level change: Observations, projections and implications.,R. A. Warrick, E. M. Barrow, and T. M. L. Wigley (Eds.) Cambridge UniversityPress, 192-212.
Reed, R. J., 1988: On understanding the meteorological causes of Saheliandrought.Pontificiae Academiae Scientarvm Scripta Varia,69,179-213.
Revell, C. G. and S. W. Goulter, 1986: South Pacific tropical cyclonesand the Southern Oscillation.Mon. Wea. Rev.,114, 1138-1145.
Ryan, B. F., I. G. Watterson and J. L. Evans, 1992: Tropical cyclonefrequencies inferred from Gray's yearly genesis parameter: Validation ofGCM tropical climates.Geophys. Res. Letters,19, 1831-1834.
Saunders, M. A., and A. R. Harris, 1997: Statistical evidence linksexceptional 1995 Atlantic hurricane season to record sea warming.Geo.Res. Letters,24, 1255-1258.
Shapiro, L. J., 1982: Hurricane climatic fluctuations. Part II: Relationto large-scale circulation.Mon. Wea. Rev.,110, 1014-1023.
Shapiro, L. J., and S. B. Goldenberg, 1997: Atlantic sea surface temperaturesand hurricane formation. Accepted toMon. Wea. Rev.
Simpson, R. H., 1974: The hurricane disaster potential scale.Weatherwise,27, 169 and 186.
Trenberth, K. E., and T. J. Hoar, 1996: The 1990-1995 El Niño-SouthernOscillation event: Longest on record.Geo. Res. Letters,23,57-60.
Velasco, I., and J. M. Fritsch, 1987: Mesoscale convective complexesin the Americas.J. Geophys. Res.,92, 9561-9613.
Wallace, J. M., 1973: General circulation of the tropical lower stratosphere.Rev. Geophys. Space Phys.,11, 191-222.
Watterson, I. G., J. L. Evans, and B. F. Ryan, 1995: Seasonal and interannualvariability of tropical cyclogenesis: Diagnostics from large-scale fields.J. Climate,8, 3052-3066.
Zehr, R. M., 1992: Tropical cyclogenesis in the western North Pacific.NOAA Technical Report NESDIS 61, U. S. Department of Commerce, Washington,DC 20233, 181 pp.
2 Whileformal U.S. armed forces aircraft reconnaissance was begun in the NorthwestPacific in 1945 (Guard et al. 1992), the U.S. Joint Typhoon Warning Center(JTWC) considers data only from 1959 onward as reliable (JTWC 1974). However,the aircraft data could and should be utilized in a tropical cyclone "reanalysis"to extend the trustworthy records back as far as possible for this basin.
3 Whilerecords are available for the entire Atlantic basin hurricanes back tothe late 1800s (Jarvinen et al. 1984) and for landfalling hurricanes alongthe United States coastline back to the 16th Century (Ludlum 1989), reliablyknowing the intensity of such systems extends for a much briefer periodof time. For the whole Atlantic basin, accurate intensity measures existback to 1944 at the commencement of routine aircraft reconnaissance (Neumannet al. 1993), but even these data have been arbitrarily corrected to removean overestimation bias in the winds of intense hurricanes during the 1940sthrough the 1960s (Landsea 1993). For U.S. landfalling hurricanes, observationsof minimum central pressure provide accurate records back to 1899 for nearlyall hurricanes (Jarrell et al. 1992). Before this year, records of intensityat landfall are incomplete and can only provide very rough estimates ofthe hurricanes' strength.
4 "Subtropicalstorms" are non-frontal low pressure systems comprising initially barocliniccirculations developing over subtropical waters with sustained one minutesurface winds of at least 18 ms-1 (NOAA 1997). Such nomenclaturehas been utilized since 1968, though it is likely that these systems weredesignated as tropical storms previously. Thus, failure to include thesubtropical storms into the climate record examined would introduce anartificial bias into the database (Neumann et al. 1993).
5 Thiscurious measurement of past hurricanes can be obtained due to the characteristicof primarily the most intense tropical cyclone rainfall having quite lowoxygen-18 isotope concentrations compared with other types of local rainfall(Lawrence and Gedzelman, 1996).