TROPICAL METERLGY II
TROPICAL METERLGY II MET 5534
Popular in Course
Popular in Meteorology
verified elite notetaker
This 11 page Class Notes was uploaded by Mrs. Brandon Eichmann on Thursday September 17, 2015. The Class Notes belongs to MET 5534 at Florida State University taught by Staff in Fall. Since its upload, it has received 80 views. For similar materials see /class/205395/met-5534-florida-state-university in Meteorology at Florida State University.
Reviews for TROPICAL METERLGY II
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 09/17/15
East Atlantic Tropical Cyclone Track Variability and Relations to the Continental United States John L Sullivan Jr MET5534 Tropical Meteorology 11 April 10 2006 1 Introduction In recent years there has been pressure to provide more advance warnings of severe tropical weather Obviously people corporations and governments alike need to know of risks well in advance of landfall East Atlantic cyclones are known to be the most prominent storms in the Atlantic and most of the storms from the basin are born in this region Being that these cyclones are the farthest from the continental United States one would surmise that if we were able to warn of a hurricane ve days in advance it would be from this region of the basin This study rst examines the track variability of tropical cyclones in the Atlantic Basin mainly those storms from the Eastern region An attempt is made to understand the preferred paths of storms impacting the United States and even three separate coastal regions the Gulf Coast Florida and the East Coast Paths of storms that do not landfall in the United States are examined as well providing insight to the other side of the argument Finally an attempt is made to understand whether there is a largescale environmental pattern at one point in the track that would steer a storm in the direction of the United States and its speci c coastlines Again the same aspects for those storms that recurve north are also provided for insight 2 Methodology All plots and analyses were created using ArcGIS The data for this research comes from two primary sources The best track dataset is provided in shape le format from the NOAA Coastal Services Center httphurricanecscnoaaguvlruu39 39 39 39 htmlj 391 39 quot of the data were done using queries and database joins to extract speci c parts of the track dataset Raster calculations were done to normalize densities The best track data extend back to the 1800 s however this study focuses on the years following the usage of satellites in tracking cyclones in order to provide the most accurate results A cutoff year was chosen after researching previous papers on track quality Only data from 1968 on is used in the study primarily due to studies by Elsberry 1987 and his notes that many biases exist in years prior He mentions that storms were often concentrated along shipping paths and track lengths were often shorter than would be expected There were other suggestions for using 1968 as the year seemed to be a common theme in many datasets even the reanalysis data used in this study the climatological means were determined between 1968 1996 The NCEPNCAR Reanalysis Data is provided by the NOAACIRES Climate Diagnostics Center CDC Daily and summed daily composites are used form their website httpwwwcdcnoaagovCompositesDay Data were extracted using the CDC website where a plot would be made of which an ASCII text le could be exported The array of data in the text les were saved and written to a latitude and longitude grid 25 x 25 in comma separated values CSV format via a FORTRAN program These les were then easily imported into ArcGIS for further manipulation with the track data The variables on the grid were interpolated using a spline interpolation scheme The variables chosen from reanalysis are geopotential heights at 925 and 500 millibars The 925 millibar level was chosen as a factor in determining near surface features especially the subtropical ridge The 500 millibar levelwas later chosen at an attempt to nd a more consistent pattern in relation with the higher levels and stronger storms Both means and anomalies are plotted in the gures Daily composite means are de ned as an average from the day the storms or storm depending on whether the plot is of multiple storms or a mean member crossed out of the East Atlantic box later de ned Anomalies are de ned as the mean minus total mean The total mean is calculated by CDC as a climatological mean based on years 1968 through 1996 It is important to note that there are many other factors that determine the path a storm will follow Harr 1991 Elseberry 1987 These variables are only provided as a simple means of pattern recognition Further studies may incorporate other variables or oscillations Some examples might include seasonal variability ie the El Nino Southem Oscillation amp Atlantic Multidecadal Oscillation and intraseasonal variability ie the North Atlantic Oscillation which alter the environment in speci c ways There are also of course the internal oscillations that were discussed in detail during this course ie the relation of storm size to Beta and intensity to steering level depth 3 Investigation In order to de ne boundaries for storms coming from the East Atlantic the overall track densities were examined Figure 1 Figure 1 shows the density of track segments with tropical depression or greater winds over the entire Atlantic basin during the 1968 through 2004 dataset Three hotspots for locations where cyclones track most often stand out one in the Gulf of Mexico another off the east coast of the United States over the Gulf Stream and one in the main development region of the East Atlantic From this it is basically noted that storms tend to either continue on into the Caribbean or recurve north by about the 60 degree west longitude Storms included further in the study are basically all storms forming south and east of 20 degrees north latitude and 60 degrees west longitude The box is drawn for display purposes but one storm was not included in all further analyses as an outlier 1973 Christine which formed over Africa All further mention of the East Atlantic storms is with these boundaries in mind Also general mention of the coast or coastline refers to that of the continental United States from Southeast Texas through Maine 1 Track Density Variability Normalized density variations are shown for all storms from the East Table 1 Number Of Storms forthe decades of Figure 3 Atlantic those that landfall in the continental US and those not making 19682004 Number East Atlantic of landfall in the US in Figures 2a 2b and 2c respectively Table 1 shows US Landfall the number of storms in each gure The majorlty of storms d1d not strike 21 28 the US yet the track density variation is very low on two distinct paths Not U39S39 2 101 to the coast This suggests that there are preferred paths for East Atlantic storms Storms not striking the US do not seem to take a preferred path in any speci c direction though they do tend to recurve north rather than continue due west Figure 2c suggests that not many storms will make their way from the East Atlantic to the Northern Gulf of Mexico and not make landfall in the US Storms striking the coastline seem to often follow similar paths either south or north of Cuba Hispaniola and Puerto Rico The southern track leads right up into the Central Gulf across the Yucatan Channel whereas the northern track takes storms up into the Gulf Stream skirting the Florida Coast with a high concentration of storm paths crossing the Carolinas region The best assumption for the split in storm paths is that the storms that did run into the Hispaniola and Eastern Cuba region likely were torn apart mostly due to the mountainous terrain This is only an assumption on my part from hurricane tracking experience Also in looking back through the storm tracks of the 1990s it seems storms would often take the northern route whereas more recently storms have often tracked more along lines with the southern route This leads to the further investigation of variability in storm tracks across the decades East Atlantic storm track density variability normalized is shown for the Table 2 Number 0f510m18 for the decades of Figure 3 1970s 1980s 1990s and 20002004 in Figures 3a 3b 3c and 3d East Number of Atlantic Storms respectively Many interesting patterns could be examined more quot 39 39 1970 1979 25 from these gures as they would probably consist of a paper or two in 1980 themselves however only a br1ef analys1s 1s app11ed here Table 2 g1ves the 1133 47 number of storms for each plot The 1970s had more storm tracks 2000 2004 29 concentrated through the Car1bbean and up into the Gulf of Mexrco as opposed to the 1980s and 90s Storms also had the tendency to pass in a region up through the north northeast of the box A classic recurving pattern is not seen per se possibly having something to do with the high concentration of storm paths that extend far outside of the box and to the west From the 1970s to the 1980s there is a higher concentration of storms inside of the box and spread further east The high concentration of storms extends further west and north from 20N 60W bringing the higher concentration of tracks closer to the east coast The 1990s seem to be a very interesting decade with a larger quantity of storms but many of them recurving closer to the coast hence teasing the US coastlines and most often passing around the Cape Hatteras coast For the half decade of best track data for the 2000 era an interesting initial look into its pattern is shown The pattern has a striking resemblance to the 1970s decade with many storms following a similar path continuing up through the Caribbean Sea and into the central and northern Gulf Coast The higher concentration of storms to the north and east is also shown in the 2000 era however the maximum concentration of storms is shifted further east It will be interesting to nd what the rest of the decade brings the US The next step in the study was to categorize the preferred East Atlantic tracks to different regions of the United States Three separate regions are de ned as impact zones Gulf Coast states Florida and East Coast states Figure 4 For simplicity a storm is said to have impacted a state when the track of the storm brings its center to intersect the state s polygon while classi ed as a tropical storm or greater during that path segment This de nition allows for the storm to be considered an impact and not just a coastal event It is also important to realize that storms may impact more than one region For example a storm passing through a Gulf Coast state or Florida and continuing through an East Coast state as a tropical storm or greater would be grouped into both impact zones The East Atlantic track density variability plots for the three different impact zones are shown in Figures 5a 5b and 5c These gures show more of what one might expect for the Gulf and East Coast zones Storms tend to track further south when they continue west into the Caribbean and up into the Gulf It seems that the longer a storm will stay south of 20 north latitude the more likely it is to impact the Gulf Coast This is especially true as it passes 70 west longitude it seems that this is the point at which storms are making their turn to the northwest Storms tracking to the East Coast seem to often stay just north of the Caribbean and often come too close for comfort by Florida Figure 3b gives a lot of credit to the notion that Florida is somewhat protected by the Caribbean islands from these often powerful East Atlantic storms The Florida impacts seem to be the most telling Both the Gulf and East coasts have tracks that seem to pass very near each other at a constant width all the way to the coastline but the Florida impacts have a remarkable wider area of track density in the region of the Bahamas This suggests that tracks often come from more multiple angles to impact Florida as we have seen with some of the more erratic storm tracks to landfall in the US b Largescale Pattern Recognition With the realization that storms do tend to stay confined and pass close to each other s paths in a dense pattern it would be enlightening to determine a variable at one point in the storm s track that would steer the storm in its direction Pattern recognition is a common favorite tool among forecasters across the eld and here this technique is applied to the tropics The most obvious point to choose in the track data is when the storm leaves the defined East Atlantic box An attempt is made here to choose a variable that would help to understand the path a storm takes according to the environmental ow and possibly forecast a storm s likely impact on the United States As a first approach two variables were chosen for their relation to the overall environment pattern as described in methodology Figure 6 contours 925 millibar heights and shades anomalies for the daily mean of all storms on the day that each storm left the box This level was chosen since it often better depicts the subtropical ridge in the reanalysis data Figures 6a 6b and 6c are those storms that impact the Gulf Coast Florida and the East Coast respectively The gulf coast has the west side of subtropical high more pointed and extending slightly farther west and into the Southeaster US whereas the East Coast events seem more northsouth oriented on the western edge possibly signifying that the ridge is retreating to the east Also of notice whether of relation or not is broader low over Baja California in the Gulf Coast impacts Being that the Florida mix would sort of be an inbetween that can be found somewhat in the Southeastern US where there is a small perturbation but somewhat wore at farther south along the Florida coast At 500 millibars Figure 7 there is an anomaly that seems to stick just above the 20 north latitude almost mirroring the path of the majority of storms 1992 Andrew excluded This is the most interesting of the reanalysis plots because it marks a catalyst that might push to keep storms south of the 20 north latitude for a longer than a climatological average period of time Of worthy mention but farther from the storm are the higher heights across the northern region Midwestern US in the East Coast plots The Gulf Coast also has a noteworthy low in the Hudson Bay region The Florida impacts have a bit of a touch of each but more of the resemblance of the low that shows up over the Hudson Bay region of the Gulf Coast map Most of the Florida storms already have more of a northeastern component as compared to those storms that impacted the Gulf Coast states The Florida storms seem to already be feeling the ridge as they make their way out of the East Atlantic box Figure 8 expands upon nding a pattern recognition scheme but shows the storms that do not landfall in the US Here is shown all storms leaving the box and not landfalling in the US with their associated mean heights and anomalies The recurving variety is of particular interest since they occur most often and have the greatest likelihood of threatening sooner Figures 9a and 9b show those storms that recurve only after leaving the side of the box 60 W at 500 millibars and 925 millibars respectively Overall the images show much weaker anomalies if any especially over the US western Atlantic and speci cally the storm s direct path This suggests that the recurving nature of the storms set the climatological normal for the storms The recurving nature is the natural tendency of storms The higher anomalies seem to be a key ingredient for US landfall Finding a pattern difference among the various criteria for the storm impact zones future is straight forward Difficulty arrives when trying to determine the signi cance of these pattern differences In determining whether there is promise to these patterns members of each composite were taken out and analyzed individually After scrutinizing over many different member composites it is determined that further signi cance test should be done to fully justify one way or the other Often patterns will match up but other instances are completely off The fact that anything is here gives promise to these ndings however It will take quite a bit more data organization programming and computing to run these calculations against every member mean There are multiple images that can get across the point here but for brevity and redundancy sakes Figure 10 gives an example of a composite that fits well as a common theme one might nd in comparing a member to its composite There are somewhat striking similarities with the 500 millibar higher height anomaly that dangles down and pushes the Gulf Coast storms farther west but other patterns over the US for example do not show much signi cance There are pieces of the plots that show skill however dif cult to distinguish while analyzing multiple maps with the human eye alone The ability is there to plot signi cance as a raster but will require further downloading of the reanalysis data for all of the individual storm tracks Overall it is safe to determine that patterns can be identi ed but actual weaknesses as strength comparisons should be taken with a grain of salt or two 4 Summary East Atlantic tropical cyclones between 1968 through 2004 often exhibit a similar path or at least pass in the same vicinity when they do landfall in the United States Storms that do not end up impacting the United States as a tropical storm or greater tend to have more variable paths but even those can often be similar during shorter periods such as decades The variation in track density from the 1970 1980s 1990s and what has existed so far since 2000 show some surprising contrasts and similarities The 2000 era can be mostly compared to the era of the 1970s however with more dense regions further west in the earlier years The 1980s and 1990s do not show a favorable route through the Caribbean and into the Gulf as do the 1970s and 2000s To a large degree Florida seems to be somewhat protected by the Caribbean islands of Cuba and Hispaniola However Florida also shows an interesting inclination to be struck by more erratic or unfamiliar storm paths We remember the jokes about Hurricane Jeanne and the path it took in 2004 but perhaps storms striking Florida are the hardest to predict Even those that do not landfall in Florida often tease the state before heading up toward the East Coast Daily reanalysis means at 925 and 500 millibar geopotential heights give sometimes differing perspectives on the ability to recognize a pattern in the largescale ow as the storm is leaving the East Atlantic box Similarities were found but the logical next step in this research willbe to statistically quantify the areas of correlation using the Geostatistical Analyst extension in ArcGIS Once these are found relations hips of track density variability to other documented oscillations may be more closely analyzed such as the El Nino Southern Oscillation ENSO Atlantic Multidecadal Oscillation AMO or North Atlantic Oscillation NAO The work here is tedious but somewhat basic in its attempts In the science community hard numbers and quanti cation often make or break our results but sometimes it is necessary and often reassuring to take a step back and look at what we actually are observing Further research will go into more specific detail of the statistical aspects This project was both enjoyable and rewarding to my knowledge and interest of tropical meteorology Special acknowledgement goes to Clark Evans for input on analysis of the reanalysis gures 5 References Elseberry R L 1987 Tropical Cyclone Motion Chap 4 A Global View of Tropical Cyclones Of ce of Naval Research Arlington VA 22217 91171 Harr PA RL Elsberry 1991 Tropical Cyclone Track Characteristics as a Function of LargeScale Circulation Anomalies Mon Wea Rev 119 14481468 Hart R E 2006 class notes Me15534 Tropical Meteorology II Department of Meteorology Florida State Univesity Tallahassee FL 32306 Kalnay E and Coauthors 1996 The NCEPNCAR Reanalysis 40year Project Bull Amer Meteor 00 77 437471 6 Figures
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'