Subject: CSU's Atlantic Hurricane Forecast From: landsea@enso (Chris Landsea) Date: 7 Jun 1995 14:03:40 GMT Message-ID: <3r4bjs$1i9f@yuma.ACNS.ColoState.EDU> FORECAST OF ATLANTIC SEASONAL HURRICANE ACTIVITY FOR 1995 By William M. Gray*, Christopher W. Landsea**, Paul W. Mielke, Jr.***, and Kenneth J. Berry*** * Professor of Atmospheric Science ** Post-doctoral Researcher of Atmospheric Science *** Professor of Statistics Department of Atmospheric Science Colorado State University Fort Collins, CO 80523 Internet: landsea@downdry.atmos.colostate.edu As of 7 June 1995 This forecast is based on ongoing research by the authors at Colorado State University, together with recent April--May 1995 meteorological information. (A more detailed version of this report will be available via WWW at CSU within a week. A notice will be posted.) ******************************************************************************* ABSTRACT This paper presents details of the authors' forecast for the amount of tropical cyclone activity expected to occur in the Atlantic Ocean region including the Caribbean Sea and the Gulf of Mexico during 1995. This forecast is based on the authors' ongoing research relating the amount of seasonal Atlantic tropical cyclone activity to five basic physical parameters. These are: 1) the Quasi-Biennial Oscillation (QBO) of equatorial stratospheric winds; 2) the El Nino-Southern Oscillation (ENSO); 3) West African Rainfall (AR) anomalies during the previous year, 4) anomalous west to east gradients of surface pressure and surface temperature (Delta PT) in West Africa during February through May, and 5) Caribbean Basin Sea Level Pressure and Upper Level Zonal Wind Anomalies (SLPA and ZWA, respectively). Information received by the authors up to 6 June 1995 indicates that the overall 1995 hurricane season should be an above average season with about 8 hurricanes (average is 5.7), 12 named storms (average is 9.3) of at least tropical storm intensity, a total of about 35 hurricane days (average is 23), 65 named storm days (average is 46) and total Hurricane Destruction Potential (HDP) of 110 (average is 68). It is also expected that there should be 3 intense or major hurricanes of Saffir/Simpson intensity category 3, 4 or 5 this season (average is 2.1) and about 6 major hurricane days (average is 4.5). These parameters represent an overall measure of total hurricane and tropical cyclone activity which is about 140 percent of the last 45-year average. The amount of hurricane activity in this forecast has been increased somewhat from that in the first author's 30 November 1994 forecast and is above the 13 April 1995 assessment given at the Atlantic City National Hurricane Conference. This early June enhancement of the hurricane forecast is due to the dissipation of the long running El Nino event and the observed return of cold water conditions in the equatorial East Pacific, to new estimates of African Western Sahel rainfall (+0.27 S.D.), and to warm eastern Atlantic sea surface temperature conditions which have become established the last two months. This is favorable for hurricane activity. This forecast will again be updated on 4 August 1995, at the beginning of what is climatologically the most active part of the hurricane season. The updated August forecast will make use of June and July data and should provide a more reliable forecast, particularly with regard to the African rainfall as it relates to prospects for intense hurricane activity during the most active part of the season. The updated August forecast will also provide a much better gauge on the extent of expected development of a (cold water) La Nina event in the equatorial east Pacific. INTRODUCTION The Atlantic basin, including the Atlantic Ocean, Caribbean Sea and Gulf of Mexico, experiences more season-to-season variability of hurricane activity than any other global hurricane basin. The number of hurricanes per season in recent years has ranged as high as 12 (as in 1969), 11 (as in 1950) and 9 (as in 1980, 1955), and as low as 2 (as in 1982) and 3 (1994, 1987, 1983, 1972, 1962, 1957). Until recently there has been no objective method for determining whether a forthcoming hurricane season was likely to be active, inactive, or near normal. Recent and ongoing research by the author and colleagues indicates that there are surprisingly skillful 3 to 11 month (in advance) predictive signals for Atlantic basin seasonal hurricane activity. FACTORS KNOWN TO BE ASSOCIATED WITH ATLANTIC SEASONAL HURRICANE VARIABILITY This early June Atlantic seasonal hurricane forecast is based on the current values of indices derived from various global and regional scale predictive factors which the author and his colleagues have previously shown to be statistically related to seasonal variations of hurricane activity. Successive sets of values for these predictive factors are obtained by late November of the previous year, by early June of the concurrent year, the official start of the hurricane season and by early August (at the start of the most active portion of the hurricane season). These predictive factors include the following: (a) The stratospheric Quasi-Biennial Oscillation (QBO) influence. The QBO refers to variable east-west oscillating stratospheric winds which circle the globe near the equator. On average, there is nearly twice as much intense Atlantic basin hurricane activity during seasons when equatorial stratospheric winds at 30 mb and 50 mb (23 and 20 km altitude, respectively) are more westerly as compared to when they are more easterly directed. During the 1995 season, these QBO winds will be from an westerly direction and are expected to be an enhancing influence for this season's hurricane activity. (b) El Nino-Southern Oscillation (ENSO) influence: ENSO characterizes the presence of either warm or cold sea surface temperature anomalies in the eastern equatorial Pacific. The effects of a moderate or strong El Nino (warm water) event in the eastern equatorial Pacific act to reduce Atlantic basin hurricane activity. By contrast, seasons with cold sea surface temperatures, or La Nina years, have enhanced hurricane activity. These differences are related to alterations of upper tropospheric (200 mb or 12 km) westerly winds over the Caribbean Basin and western Atlantic. These westerly winds are enhanced during El Nino seasons. This condition creates strong vertical wind shear over the Atlantic which inhibits hurricane activity. During La Nina (or cold) years, these westerly winds and the associated vertical wind shear are reduced and hurricane activity is typically greater. It is expected that the unusually long lasting 1991-94 El Nino has finally run its course and cold water conditions are settling in. This, in turn, should be a factor enhancing this year's Atlantic basin hurricane activity. New April-May data indicates that distinctly cold conditions are developing for this year. (c) African Rainfall (AR) influence: The incidence of intense Atlantic hurricane activity is typically enhanced during those seasons when the Western Sahel and Gulf of Guinea regions of West Africa had above average late summer and fall precipitation during the previous year (in this case during the fall of 1994). Hurricane activity is typically suppressed if the prior fall rainfall in these two regions was below average. Rainfall amounts in the Western Sahel in August-September 1994 was +0.08 S.D.; Gulf of Guinea rainfall was +0.24 S.D. during August through November 1994. From these slightly wetter than average rainfall amounts we anticipate that Western Sahel rainfall will be slightly above normal in 1995 and distinctly above the very low rainfall which has occurred in the recent years of 1990-1993 and also greater than the drought years of the 1970s and 1980s. This Western Sahel rainfall forecast is based upon our early June forecast of rainfall. (d) West Africa west-to-east surface pressure and surface temperature gradients (Delta PT) influence. Recent project research has shown that anomalous west- to-east surface pressure and surface temperature gradients across West Africa during February through May are strongly correlated with the hurricane activity which follows later in the year. We find that Atlantic hurricane activity is enhanced when the February to May east minus west pressure gradient is higher than normal and/or when the east minus west temperature gradient anomaly is below average. These pressure and temperature gradients during February through May 1995 indicate an about average West African monsoon and an about average amount of seasonal hurricane activity. (e) Caribbean Basin Sea Level Pressure Anomaly (SLPA) and upper tropospheric (12 km) Zonal Wind Anomaly (ZWA) influence. April and May values of SLPA and ZWA have a modest predictive potential for hurricane activity during the following season. Negative anomalies (low pressure and easterly anomalies) imply enhanced seasonal hurricane activity while positive values imply suppressed hurricane activity. April-May 1994 values of SLPA and ZWA were both slightly below average, indicating a slight enhancing influence on this season's hurricane activity. Our seasonal forecast scheme has the following general form: Predicted Amount of TC Activity = Ave. Season + (QBO + EN + AR + PT + SLPA + ZWA) Per Season where QBO = 30 mb and 50 mb Quasi-Biennial Oscillation zonal wind influence. (Increased hurricane activity for westerly (or positive) zonal wind anomalies; reduced hurricane activity for easterly or negative zonal wind anomalies.) EN = El Nino influence. (Warm surface water in the equatorial East Pacific reduces hurricane activity, cold water enhances it.) AR = Western Sahel rainfall. (Increase activity if wet; reduce activity if dry.) PT = West Africa west-to-east gradients of surface pressure and surface temperature during February through May. (High values of west-to-east pressure gradient and high values of east-to-west temperature gradient indicate more hurricane activity; less hurricane activity with opposite gradients) SLPA = Average Caribbean Sea Level Pressure Anomaly (SLPA) for Spring and early Summer. (Reduce hurricane activity if SLPA is significantly above average; add activity if SLPA is significantly below average.) ZWA = Zonal Wind Anomaly at 200 mb (12 km) for five low latitude upper air stations in the Caribbean. (Reduce hurricane activity if positive; increase hurricane activity if negative.) SEASONAL FORECAST Based on data through the end of May 1995, the 1 June predictors in our forecast equations indicate an above average hurricane season for 1995. Table 1 shows our objective forecast from our forecast equations and our qualitative adjustment of our actual forecast. Observe that our prediction equations indicate an extremely active hurricane season for this year. We have added a downward adjustment to many of the values here shown. Last year our adjustment to our forecast equation actually gave a poorer forecast. TABLE 1: The 1995 seasonal forecasts obtained by substitution of the parameter values in Table 2 into Equation (1). The author's qualitative adjustments and actual forecast are shown on the right column. Qualitative Adjustment Forecast Objective and Actual Parameter Forecast Forecast ------------------------------------------------------------------- Named Storms (N) 12.6 12 Named Storm Days (NS) 77.9 65 Hurricanes (H) 10.4 8 Hurricane Days (HD) 49.5 35 Intense Hurricanes (IH) 2.3 3 Intense Hurricane Days (IHD) 5.7 6 Hurricane Destruction Potential (HDP) 135.4 110 Net Tropical Cyclone Activity (NTC) 153.5% 140% Western Sahel rainfall forecast +0.27 S.D. +0.27 S.D. ------------------------------------------------------------------- Table 2 expresses each parameter in this adjusted forecast as a percentage of the last 45-year average. Note that all forecast parameters are well above the long period average. Table 3 compares this early June forecast to the first author's late November 1994 forecast of last year. The November 1994 forecast anticipated cold ENSO conditions to be in place during the height of the 1995 hurricane season. As of early June, this late November, 1994 assessment appears to be on target. Table 4 gives a comparison of this year's seasonal activity forecast with the amount of hurricane activity which has occurred during past years. Note that the 1995 season is expected to be much more active than have the last four hurricane seasons and more active than most of the hurricane seasons since the late 1960's. TABLE 2: 1995 Atlantic basin seasonal forecast values as a percent of the long term (1950-1994) average. Named Storms (N) 129% Named Storm Days (NS) 141 Hurricanes (H) 140 Hurricane Days (HD) 152 Intense Hurricanes (IH) 136 Intense Hurricane Days (IHD) 133 Hurricane Destruction Potential (HDP) 162 Net Tropical Cyclone Activity (NTC) 140 ------------------------------------------------- TABLE 3: Comparison of current, early June 1995, seasonal predictions with the 1995 seasonal predictions made in late November 1994. Changes in Current 7 June 95 30 Nov. Qualitative 1995 Early Fcst from 30 Forecast 1994 Adjustment June Nov. 1994 Parameter Fcst. 13 April 1995 Fcst. Fcst. --------------------------------------------------------------------- NS 12 10 12 0 NSD 65 50 65 0 H 8 6 8 0 HD 35 25 35 0 IH 3 2 3 0 IHD 8 5 6 -2 HDP 100 75 110 +10 NTC 140 100 140 0 --------------------------------------------------------------------- TABLE 4: Comparison of early June 1995 seasonal prediction with activity in previous years. 7 June Average Average 45-Year Forecast ......Observed...... Season Season Ave. 1995 1994 1993 1992 1970-87 1950-69 (1950-94) -------------------------------------------------------------------------- NS 12 7 8 6 8.3 9.8 9.3 NSD 65 28 30 38 37.3 53.4 46.1 H 8 3 4 4 4.9 6.5 5.7 HD 35 7 10 16 15.5 30.7 23.0 IH 3 0 1 1 1.6 3.4 2.1 IHD 6 0 0.75 3.25 2.1 8.8 4.5 HDP 110 15 23 51 42.7 100.0 68.1 NTC 140% 37% 55% 62% 73% 123% 100% -------------------------------------------------------------------------- _US East coast and Florida Hurricane Landfall_. Hurricane Andrew and Bob have been the only US landfalling hurricanes of the last five years (Emily, 1993 only skimmed the outer banks as a minimal category 3 hurricane). All of these cyclones did not become of hurricane intensity until they were poleward of 25N. It is unusual to go five consecutive years without any low latitude origin hurricanes threatening upon the US coastline. These inhibiting influences are not expected to be present during the 1995 hurricane season, however. _Gulf of Mexico_. During the last five years (1990-94), there was only one hurricane (Andrew, 1992) to make landfall on the US Gulf Coast. This is much below the long term average. This reduction of hurricane activity in the Gulf was primarily a response to the long lasting El Nino event of 1991-94. Historical records show that Gulf hurricane activity is usually suppressed during El Nino years and enhanced during La Nina (cold SST) years. The probability of hurricane activity within the Gulf of Mexico will be higher during 1995 than it has been since 1989. _Caribbean Basin_. There has been no hurricane activity at all within the Caribbean during the last five years. This is a consequence of the long lasting El Nino event of 1991-94, Western Sahel drought conditions during 1990-93 and higher than average Caribbean basin surface pressures during the last five years. These inhibiting influences are not expected to be present during the 1995 season. Consequently, the probability of Caribbean basin hurricane activity will be greater this year than any of the last five years. DISCUSSION It is well known that Atlantic Basin seasonal hurricane activity is quite variable. Past records indicate that it is typical to have a number of suppressed or somewhat below normal years in a row which are then followed by a year of greatly increased hurricane activity. It appears that 1995 will be one of those seasons wherein a large upsurge in hurricane activity occurs. The El Nino, stratospheric QBO, West African rainfall, and Atlantic sea surface temperature anomalies are all coming together to promote the large- scale wind and thermal-moisture conditions which are associated with an active season. The first author now regrets his qualitative downward adjustment of the forecast in early April. The sea surface temperature pattern in both the Atlantic and Pacific have been more favorable for hurricane activity in the last two months. It is important to keep in mind that the early season hurricane activity has _no_ bearing on the entire season. Early June Hurricane Allison means nothing with regards to the hurricane activity to follow later in the season. The number of named storms (hurricanes) occurring in June and July correlates at an insignificant r = +0.13 (+0.02) versus the whole season activity. Actually, there is a slight _negative_ association of early season storms (hurricanes) versus late season - August through November - r=-0.28 (-0.35). Thus, early season activity, be it very active or quite calm, has little or no bearing on the season as a whole. SCHEDULE OF UPDATED SEASONAL HURRICANE FORECAST OF 1995 An updated forecast, to be made at the start of the most active part of the hurricane season, will be issued on Friday August 4, 1995. A verification report on the 1995 hurricane season and a forecast for the 1996 hurricane season will be issued in late November of this year. In addition, seasonal forecasts for 1996 ENSO conditions and 1996 Sahel rainfall will also be issued at that time. CAUTIONARY NOTE It is important that the reader realize that this seasonal forecast is based on a statistical scheme which will fail in some years. This forecast also does not specifically predict where within the Atlantic basin storms will strike. Even if 1995 should prove to be an active hurricane season, there are no assurances that hurricanes will necessarily strike along the US or Caribbean Basin coastline and do much damage. Or, if 1995 should prove to be an inactive season there is no assurance that no storms will come ashore. LIKELY INCREASE OF LANDFALLING MAJOR HURRICANES IN COMING DECADES There has been a great lull in the incidence of intense (category 3-4-5) landfalling hurricanes on US East Coast, Florida and Caribbean Basin during the last 25 years. We see this as a natural consequence of the slowdown in the Atlantic Ocean (thermohaline) Conveyor Belt circulation which has set off a variety of global circulation and rainfall pattern changes such as the Sahel drought, increased El Nino activity, Pacific and Atlantic middle latitude zonal wind increases, etc. Historical and geological records indicate that this lull in major landfalling hurricane activity will not continue indefinitely. A return of increased major landfalling hurricane activity should be expected within the next decade or two. When this happens, (because of the large coastal development during the last 25-30 years), the US will see hurricane destruction as never before experienced. More research on the causes and the likely timing of this change-over to increased intense hurricane activity is desperately needed. This is a more assured and immediate threat to the US than that of greenhouse gas warming and other environmental problems which are receiving much greater attention in comparison to the hurricane threat. _Changes in the North Atlantic_. We may be seeing the early stages of the beginning speed-up of the Atlantic thermohaline (Conveyor Belt) circulation from its three decades long slow down. Dr. Aagaard has recently reported on a large decrease in ice flow through the Fram Strait (the North Atlantic passage between Greenland and Spitzbergen). This decreased ice flow reduces the introduction of fresh water and low salinity values into the North Atlantic. This ice flow reduction is leading to salinity increases in the North Atlantic. Other researchers have also recently reported recent salinity increases in the North Atlantic. Saline water weighs more than fresh water and is able to sink more readily to the bottom of the North Atlantic. These salinity increases that are now being measured in the North Atlantic may result in a speed-up of the Atlantic Ocean thermohaline circulation in the next few years. If this does occur, then we should see a general increase in West African Sahel rainfall, a decrease in Atlantic summertime upper tropospheric westerly winds and an increase of Atlantic basin intense hurricane activity. These new regional Atlantic measurements may be an ominous sign of future increases in US and Caribbean basin landfalling hurricane activity. The quarter century lull which we have enjoyed cannot be expected to continue indefinitely into the future. VERIFICATION OF PAST SEASONAL FORECASTS The first author has now issued seasonal hurricane forecasts for the last eleven years. In most of the prior forecasts, predictions have been superior to climatology, which was previously the only way to estimate future hurricane activity (see Table 5). The seven late May and early June seasonal forecasts for 1985, 1986, 1987, 1988, 1991, 1992 and 1994 were more accurate than climatology. The forecasts for 1984 and 1990 were only marginally successful and the two seasonal forecasts for 1989 and 1993 were failures. The 1989 forecast was a failure because of processes associated with the excessive amounts of rainfall which fell in the Western Sahel that year. Prior to 1990, our seasonal forecast did not include African rainfall as a predictor. We have corrected this important omission and forecasts since 1990 have incorporated Western Sahel rainfall estimates. The failure of the 1993 seasonal forecast is attributed to our failure to anticipate the resurgence and continuation of El Nino conditions through the whole of the 1993 hurricane season. In particular, the first author failed to anticipate the re-emergence of stronger El Nino conditions after the middle of August 1993. It is very unusual to have an El Nino last as long as the recent 1991-94 event has. This failure motivated us to develop a new extended range ENSO prediction scheme. TABLE 5: Verification of the author's previous seasonal predictions of Atlantic tropical cyclone activity for 1984-1994. Year & Late Late May/ Late July/ Category November Early June Early August Forecasted Forecast Forecast Forecast OBSERVED -------------------------------------------------------------------------- 1984 & NS - 10 10 12 1984 & NSD - 45 45 51 1984 & H - 7 7 5 1984 & HD - 30 30 18 1985 & NS - 11 10 11 1985 & NSD - 55 50 51 1985 & H - 8 7 7 1985 & HD - 35 30 21 1986 & NS - 8 7 6 1986 & NSD - 35 25 23 1986 & H - 4 4 4 1986 & HD - 15 10 10 1987 & NS - 8 7 7 1987 & NSD - 40 35 37 1987 & H - 5 4 3 1987 & HD - 20 15 5 1988 & NS - 11 11 12 1988 & NSD - 50 50 47 1988 & H - 7 7 5 1988 & HD - 30 30 24 1988 & HDP - 75 75 81 1989 & NS - 7 9 11 1989 & NSD - 30 35 66 1989 & H - 4 4 7 1989 & HD - 15 15 32 1989 & HDP - 40 40 108 1990 & NS - 11 11 14 1990 & NSD - 55 50 68 1990 & H - 7 6 8 1990 & HD - 30 25 27 1990 & IH - 3 2 1 1990 & IHD - - 5 1.0 1990 & HDP - 90 75 57 1991 & NS - 8 7 8 1991 & NSD - 35 30 22 1991 & H - 4 3 4 1991 & HD - 15 10 8 1991 & IH - 1 0 2 1991 & IHD - 2 0 1.2 1991 & HDP - 40 25 23 1992 & NS 8 8 8 6 1992 & NSD 35 35 35 38 1992 & H 4 4 4 4 1992 & HD 15 15 15 16 1992 & IH 1 1 1 1 1992 & IHD 2 2 2 3.2 1992 & HDP 35 35 35 51 1993 & NS 11 11 10 8 1993 & NSD 55 55 50 30 1993 & H 6 7 6 4 1993 & HD 25 25 25 10 1993 & IH 3 2 2 1 1993 & IHD 7 3 2 0.8 1993 & HDP 75 65 55 23 1994 & NS 10 9 7 7 1994 & NSD 60 35 30 28 1994 & H 6 5 4 3 1994 & HD 25 15 12 7 1994 & IH 2 1 1 0 1994 & IHD 7 1 1 0 1994 & HDP 85 40 35 15 1994 & NTC 110 70 55 37 DEFINITIONS Atlantic basin - The area including the entire Atlantic Ocean, the Caribbean Sea, and the Gulf of Mexico. Hurricane - (H) A tropical cyclone with sustained low level winds of 74 miles per hour (33 m/s or 64 knots) or greater. Hurricane Day - (HD) Four 6-hour periods during which a tropical cyclone is observed or estimated to have hurricane intensity winds. Tropical Cyclone - (TC) A large-scale circular flow occurring within the tropics and subtropics which has its strongest winds at low levels including hurricanes, tropical storms, and other weaker rotating vortices. Tropical Storm - (TS) A tropical cyclone with maximum sustained winds between 39 (18 m/s or 34 knots) and 73 (32 m/s or 63 knots) miles per hour. Named Storm - (NS) A hurricane or a tropical storm. Named Storm Day - (NSD) Four 6-hour periods during which a tropical cyclone is observed or estimated to have attained tropical storm or hurricane intensity winds. Hurricane Destruction Potential - (HDP) A measure of a hurricane's potential for wind and storm surge destruction defined as the sum of the square of a hurricane's maximum wind speed (in 10,000 knots*knots for each 6-hour period of its existence. Intense Hurricane - (IH) A hurricane reaching at some point in its lifetime a sustained low level wind of at least 111 mph (96 kt or 50 m/s). This constitutes a category 3 or higher on the Saffir/Simpson scale (a "major" hurricane). Intense Hurricane Day - (IHD) Four 6-hour periods during which a hurricane has intensity of Saffir/Simpson category 3 or higher. Millibar - (mb) A measure of atmospheric pressure which is often used as a vertical height designator. Average surface values are about 1000 mb; the 200 mb level is about 12 kilometers and the 50 mb is about 20 kilometers altitude. Monthly averages of surface values in the tropics show maximum summertime variations of about +/-2 mb which are associated with variations in seasonal hurricane activity. El Nino - (EN) A 12-18 month period during which anomalously warm sea surface temperatures occur in the eastern half of the equatorial Pacific. Moderate or strong El Nino events occur irregularly, about once every 5-6 years or so on average. Delta PT - A parameter which measures the anomalous west to east surface pressure (Del-P) and surface temperature (Del-T) gradient across West Africa. SOI -Southern Oscillation Index - A normalized measure of the surface pressure difference between Tahiti and Darwin. QBO - Quasi-Biennial Oscillation - A stratospheric (16 to 35 km altitude) oscillation of equatorial east-west winds which vary with a period of about 26 to 30 months or roughly 2 years; typically blowing for 12-16 months from the east, then reverse and blowing 12-16 months from the west, then back to easterly again. Saffir/Simpson (S-S) Category - A measurement scale ranging from 1 to 5 of hurricane wind and ocean surge intensity. One is a weak hurricane whereas 5 is the most intense hurricane. SLPA - Sea Level Pressure Anomaly - A deviation of Caribbean and Gulf of Mexico sea level pressure from observed long term average conditions. SST(s) - Sea Surface Temperature(s). ZWA - Zonal Wind Anomaly - A measure of upper level ($\sim$ 200 mb) west to east wind strength. Positive anomaly values mean winds are stronger from the west or weaker from the east than normal. Net Tropical Cyclone Activity (NTC) - Average seasonal percentage sum of NS, NSD, H, HD, IH, IHD. Gives overall indication of Atlantic basin seasonal hurricane activity. 1 knot = 1.15 miles per hour = .515 meters per second. ***************************************************************************** Chris Landsea Voice: (303) 491-3569 Department of Atmospheric Science Fax: (303) 491-8449 Colorado State University Internet: Fort Collins, CO 80523 landsea@downdry.atmos.colostate.edu ***************************************************************************** "The Florida straits were as dangerous as the Florida Indians. There were... the extraordinary danger of hurricanes in the tropic latitudes, that could blow up almost without warning from June to November, gray screaming whirlpools of wind more than a hundred miles an hour, dragging in their centers a mound of sea water and blowing before them the high ungovernable ships like dried leaves onto that deadly line of reef and rock." - _The Everglades: River of Grass_ - Marjory Stoneman Douglas - 1947