Tag Archives: Weather

Weather Advance hurricane pic 3

Stratospheric Warming, Indian Ocean Dipole implications on hurricane season & next winter. 1996 now a hurricane season analog. Cape Verde storm for July?

Hi all, and I want to say first of all, I greatly appreciate all of the feedback and comments I receive in my posts, and if you are a viewer of these posts and are too shy or perhaps don’t know what to comment, feel free to do so. As I have discovered in my experience here at weatheradvance, that it’s the commenters, and even the simplest of questions which have helped to lead me to discover much more about the weather & to continue to learn and gain experience about weather forecasting, and it is an honor to have people such as all of you.


Now, looking at what is currently going on in the tropics and in the weather, the main story for the past few days has been Tropical Storm Andrea that developed in the eastern Gulf of Mexico. To some people, such a storm came as a surprise, but this pattern of development was foreseeable very far in advance. Look at the MJO (Madden Julian Oscillation, a generally eastward propagating region of rising and sinking motion in the global tropics, with 8 octants, and the MJO is important in the forecasts for winter storms, hurricanes, SOI, and equatorial Kelvin Waves, which have influence on the state of the ENSO index.), notice how the MJO has moved into Octants 2 & 3 recently, indicative of the rising motion from the MJO being centered over the Indian Ocean. When the MJO is in this region, it favors general upward motion in the Atlantic because a lot of the convection and upward motion that is produced goes into the Indian Monsoonal Circulation, and since activity in the tropics tends to move in an east-west fashion because of the effects of the Hadley Cell, convective activity over the Indian Ocean would help to enhance thunderstorm activity and moisture over Africa, and considering that around 90% of all tropical cyclones have their origins from African Easterly waves, it makes sense as to why even with the MJO towards the Indian Ocean that activity is still favored in the Atlantic. Also, with the MJO being a naturally eastward propagating wave of energy in the tropics, this usually means that any upward MJO pulse that is centered over the Indian Ocean is forced to move through the Atlantic,  or in close proximity, and any lingering effects of upward motion are likely still to be left present in the Atlantic until the atmosphere can adequately respond to the MJO’s passing. Thus, with upward motion favored in the Atlantic, this enhances thunderstorm activity that are the building blocks for tropical cyclones through processes of latent heat release & with increased thunderstorm activity, levels of moisture in the atmospheric column are also raised, and this is favorable to tropical cyclones by moistening their surrounding environment and limiting dry air, which tends to damper tropical cyclone development.


Look at the earlier GFS forecast for the MJO back on May 11th, if one just looked at this forecast you would have expected that the MJO would quickly return to the Atlantic



However, this was not the case, as the MJO did not move nowhere near as fast into the favorable tropical cyclone octants 8-3, thus, this is a good reason why we observed for several weeks preceding the formation of Tropical Storm Andrea, “ghost storms” on the GFS model.


Observed MJO with GFS verification forecasts, showing the large error the model had in predicting the MJO.



The ECMWF, although usually better at the MJO did not handle very well the progression of the MJO out of octant 4 (Maritime Continent), and perhaps this is due to the varied topography in the region which creates a “standing mountain wave” over the general area, a similar problem in this abnormally slow handling of energy by the ECMWF is also observed in the southwestern US, where a “standing mountain wave” due to the surrounding high terrain, particularly from the Rockies can cause issues with the model. Also, a relative lack of data in this region compared to other areas as well as a lack of physics input into the model to accurately assess this type of situation can lead to model issues and biases.

ECMWF MJO forecast from earlier in May



A “standing mountain wave” is generally a region where air on the windward side (front) side of the mountains is forced to rise up, thus air converges, but on the other side, air is generally forced to spread out, and this spreading out of air is indicative of warming, and this warming that occurs in the troposphere rises upwards because it is more buoyant than the cold air above and deposits its energy into the tropopause and lower stratosphere.

Stratospheric warming & mountains


The deposit of energy in these levels of the atmosphere, in particular during the winter, when the polar vortex is well entrenched over the northern hemisphere, can lead to major stratospheric warming events as the push of warmth into the lower stratosphere, where in this case, the stratosphere’s temperature profile is different from that of the troposphere, where in the stratosphere, temperature actually increases with altitude, thanks to the radiative effects of ozone in the lower stratosphere.

This picture below showing the levels of the atmosphere in relation to height above the surface and temperature profiles in the different levels of the atmospheric column, also shown are temperatures with decreasing values as you move towards the left side of the picture. You can clearly see that in comparison to the troposphere, the stratospheric temperature profile is reverse of that in the troposphere, where temperatures are coldest in the lower levels of the stratosphere and much warmer in the upper levels.



Thus, when you have a sudden push of energy upwards through the troposphere, with some of its energy deposited in the form of heat into the lower levels of the stratosphere, a “standing mountain wave” created naturally by the sudden changes in topography over a certain area can lead to major stratospheric warming events because with warmer air being forced into the lower levels of the stratosphere, this disturbs the natural vertical temperature gradient in the stratosphere, where the coldest temperatures are in the lower levels and warmer temps higher up, warming in the lower levels from standing mountain waves reduces this temperature gradient. A reduced temperature gradient creates major issues as it greatly disturbs the natural flow of air in the stratosphere as part of the Brewer-Dobson Circulation (an atmospheric circulation that moves ozone created in the tropics by the interaction of ultraviolet radiation with oxygen molecules, and this forces ozone to “pile-up” near the poles with significantly less amounts towards the tropics and highly variable amounts in ozone in what is referred to as the “Surf Zone”, which is located over the mid-latitudes.).



Brewer-Dobson Circulation, arrows in white indicative of the natural flow of air, orange showing radiative processes, and you can easily see the tendency for air to rise in the tropics, most of which flows back downward towards the surface near the “horse latitudes” in the presence of the natural subtropical high pressure zone, while some air makes it into the stratosphere and gets carried away towards the poles. where in this case, it contains ozone, which accumulates near the poles.



With this information in mind these standing mountain waves, which are in many cases triggered by large-scale “Rossby Waves” (features that generally refer to persistent and abnormally strong regions of atmospheric low and high pressure), which in turn must propagate upwards into the troposphere to rid themselves of excess energy, which is usually released in the form of heat, and knowing that the coldest part of the stratosphere is in the lower levels, excess energy released by Rossby Waves & Standing Mountain waves lead to some warming at this level of the atmosphere. This warming helps to slow down the natural flow of air from the tropics to the poles in the stratosphere into the polar vortex. Reduced air flow into the polar vortex, like any low pressure area greatly disturbs the region of low pressure and helps to lead to major stratospheric warming events.


This animation below shows the evolution of the stratospheric warming event in 2008-09, very interesting to watch the progression of the polar vortex and its demise as a major warming event takes place.


Just look at the stratospheric warming event that occurred this past winter, you can see how the warming event got going as what appeared to be a “wave-like feature” which rolled across the Pacific, (Kelvin Wave) forced a large push of energy across the global tropics, and as this wave of energy met up with the Himalayas, the wave of energy was forced upwards, and then on the backside of the mountain, the “standing mountain wave” forced any energy from the associated Rossby Wave feature to be forced upwards into the  upper troposphere and lower stratosphere. This led to significant warming in the stratosphere north of the Himalayas, and aided by substantial surface and tropospheric cooling underneath (leads to a general contraction of the lower levels of the atmosphere as cooler air naturally takes up less space than warmer air because of the limited molecule movement) from an abnormally high snowpack, this stratospheric warming event only continued to intensify.


This animation below shows the major stratospheric warming event that occurred last winter, at a level of 10 millibars, and you can easily see (shown in shades of deep red), how the stratosphere becomes very warm as a “standing mountain wave” on the north side of the Himalayas and a Rossby Wave combine to help promote this stratospheric warming event.

Here is a 10 millibar animation showing the stratospheric warming event.


Interestingly, at the 50 millibar level, much lower in the stratosphere, about a month earlier around the end of November, there were signs of an impending stratospheric warming event, however, this failed to occur. What could be taken away from this earlier warming is despite it being a failed attempt at a stratospheric warming event, it may have been a sign of things to come later in the fall & winter. More research should be conducted on this



Look at the 500 millibar pattern for November, and you can see the 500 millibar pattern over the northern hemisphere, with notable features such as a trough of low pressure near the southeastern US, the Bering Sea ridge, a trough in western Europe (teleconnects to the eastern US), and troughing over eastern Asia. What also seems to grab my attention is the very noticeable “cross-polar flow” that I’ll explain more in post below, where troughs and ridges in the jet stream pattern seem to orientate themselves meridionally (means more north to south in nature, or more poleward, as opposed to east & west). This type of signature is indicative of a fairly weak jet stream with very large undulations that have the capability of bringing fairly extreme weather to many areas of the globe.

Nov 2012 500 mb heights


What brought this to my attention was a very good post made by Andrew of The Weather Centre (link) http://theweathercentre.blogspot.com/2013/06/preliminary-2013-2014-winter-forecast.html

In the post, I noticed several things that are seeming to remind me of this pattern and the overall cycles of PDO & AMO. First of all, his selection of the years 1952 & 1963 are significant, in that they both fall within our current cycle of cold PDO and warm AMO.


What does seem to grab my attention though is how these analog years Andrew presented in his preliminary winter forecast seem to catch the central Pacific “modiki” el nino that’s being forecasted by many of the computer models to develop this summer (I’ll explain the reasoning behind that scenario later in the post).

1952 & 1963 water temperatures Jul-Sep Pacific basin, notice the region of warming over the central Pacific between 120 and 160 degrees west near the equator (shown in shades of green & yellow).


ECMWF forecast, interesting in how you can see, like the picture above, the cold PDO signature still in place with warmer than normal water centered to the south of the Gulf of Alaska, and cooler than normal waters near Alaska and off of the coast of northeastern Asia.



CFSv2 forecast, like the ECMWF, showing an el nino developing, although I think it is overly aggressive with the warmer than normal waters over the eastern Pacific, I’ll explain why later in this post.



This is a generalization of how a modiki el nino works to force cold over North America

modiki el nino=North America cold


The low pressure near Hawaii that is created by the warmer than normal waters over the central Pacific, acts like a rock in a stream, forcing air to slow down as it passes near the region of low pressure, essentially this trough near Hawaii acts as a road block for the Pacific jet, forcing air to slow down in its vicinity. In doing so, more air rushing in from the west continues to slam in, and over time there is a piling up of air that develops to the north & northeast of the Hawaiian low. This leads to regions of higher than normal pressure over the western US & western North America, and when you have high pressure over the western US this is indicative of a +PNA.

+PNAs in the winter favors a trough in the eastern US, and along with that trough of low pressure comes a favorable environment for the development of winter storms

A few examples of this Hawaiian low at work, notice in each picture below, how the low pressure near Hawaii (indicated in shades of blue) seems to teleconnect to a trough of low pressure (shown in shades of blue & green) in the eastern US, which in the winter, means cool & relatively stormy conditions that helps to provide a generally favorable environment for winter storms.






Remember the early February snowstorm that looked a lot like the 1969 Lindsay Storm?

1969 Lindsay Storm



2013 early February snowstorm




Well once again, look at this, you can see the trough near Hawaii, which argues for a major trough near the eastern US, thus, you should question this solution below offered by the ECMWF that only shows a modest trough, based on the Hawaiian to eastern US trough connection, it should be stronger, and that was certainly the case.



More similarity to this pattern presented by Andrew of The Weather Centre, in 1963, if we were take 1963 as a potential analog to this upcoming winter, and consider last year analogous to 1962, makes sense as if you recall the early March snowstorm that hit the northeast, looks a lot like the Ash Wednesday snowstorm of 1962, offering more evidence for Andrew’s potential analog years. Thus, if this pattern of following 1962 were to continue

March 5 1962 snowfall




March 7 1962 500 millibar heights


Track & snow map 1962 Ash wed storm


Early March 2013 snowstorm

NESIS snow March 4-9 2013



Going back to the failed stratospheric warming attempt in at 50 millibars in November 2012, the temperatures in the US that month looked like this, with a tongue of cold extending southeastward into the eastern US, warming farther to the west.


I’m quite astonished to see that this temperature pattern presented above seems to resemble in some ways the Decembers following the greatest hurricane landfalling seasons on the US coast, in that you can see the warmth over the western US, warmth towards eastern Canada (although farther north towards Labrador and the Hudson Bay), with a tongue of cold stretching southeastward into the eastern US.


Interesting in how also this temperature pattern of cold seems a lot like the winters of the 1960s



Just more evidence that this warm AMO pattern may be going into its cold mode, sooner than many think.



Here’s a graph showing Total Solar Irradiance (TSI) and the AMO, showing how the sun correlates nicely with the cycle of the AMO, and given recent trends in lowering sunspot cycles, one can conclude that the AMO will soon enter its cool phase.


As I’ve shown in previous posts, this evidence of the pattern heading towards a more 1960s look is also shown in recent European winters which bear resemblance to the winters of the early 1960s, before the cold AMO flip around 1965.

Europe winter temps 1960-1964


Last 4 European winters

Also, there are signs that tropical activity may be shifting back into the Gulf (at least in a relative sense) from the eastern seaboard, a similar shift was observed in the 1960s when the AMO cooled, while the PDO remained in its cold state. You can see how, given this evidence above, we may be currently nearing the edge of the 1950s pattern of east coast hurricane landfalls.

Starting in the 1940s, look how this landfall pattern of hitting mainly Florida & some of the Gulf of Mexico is analogous to the 2000s hurricane hit pattern.

1940s US major hurricane hits

1940s Major Hurricanes US coast


2000s US major hurricane hits

2000s Major Hurricanes US coast


1950s Major hurricane hits

1950s Major Hurricanes US coast


Irene (2011)




Hurricane Sandy (2012)




1960s US major hurricane hits

1960s Major Hurricanes US coast


Although this pattern may resemble the 1960s in some aspects, we are not completely into a 1960s pattern as the east coast hurricanes continue, the plains drought (in a relative sense) continues, much like the 1950s, thus with conflicting factors and signals for a 1950s or 1960s pattern, it makes sense as to why in my hurricane forecast, I prefer a blend of the 1950s & 60s landfalling patterns with the East Coast and the Gulf Coast both threatened this year.

Hurricane Forecast

Looking back at the original topic of discussion that helped to trigger this further analysis, the significance of stratospheric warming events are that they help to disturb the polar vortex and lead to major arctic outbreaks in the mid-latitiudes. The polar vortex itself is a feature that is created as the changing axial tilt of earth leads to areas in the polar region that go into permanent darkness following the fall equinox, and in doing so, the lack of sunlight energy leads to dramatic cooling not only at the surface, but throughout much of the atmospheric column. This cooling over time grows stronger, while other areas farther to the south continue to experience ample sunshine, thus in this scenario, which is something that is observed every fall, a major difference in temperature over a relatively short distance creates a large difference in pressure, and it is a pressure gradient that leads to the creation of wind. Since, this pressure gradient is strongest also where the temperature gradient is strongest, this means that the jet stream is most likely to form along the boundary between the completely dark areas of the arctic and the sunlight receiving regions farther south, which is a good reason why this jet stream is also known as the “polar night jet”.


This picture below shows 500 millibar heights over the northern hemisphere preceding this major warming event, and you can easily see the “cross-polar” flow in which, the flow of air, instead of circulating around the pole as what typically occurs, it appears to go across the pole, indicative of the how weak the jet stream is, thus it will have a much stronger tendency to promote upper-level ridging and blocking over the arctic, which favors a -AO & -NAO regime.


A similar pattern set-up was observed in advance of the major 1985 stratospheric warming event that led to one of the worst arctic outbreaks in US history that January.

500 millibar heights over the northern hemisphere December 30, 1984 through January 5th, 1985.



Look at the temperatures for the month of January 1985, very cold throughout much of the US.



Interestingly, like this year, the winter of 1984-85 got off to a late start, with December much above normal across much of the US, especially over the eastern US.


December 1984 temperature anomalies.



This past December temperature anomalies

US temp anomaly Dec-Jan 2012-13


The cold in the winter of 1984-85 hung around too, February 1985 temperature anomalies



Temp anomalies from January 15th 1985 through the rest of the winter reveal that once the pattern turned cold it stayed cold for the rest of the winter.



This past winter, like 1984-85, although the stratospheric warming event got going a little later, the results were the same, once the pattern turned cold as a result of major stratospheric warming, it never looked back.

Temperature anomalies from the beginning of February to the end of this past winter



Interestingly, the following 1985 hurricane season on the US was extremely destructive, with a record-breaking number of hurricanes making landfall on the US coast.




Also, like this year, 1985 was near the extremity of the solar cycle, which I have shown to have correlations to US hurricane landfalls, in that years falling near the extremes of the cycle (minimum of maximum) tend to lead to increased numbers of tropical cyclones making landfall on the US coast, and this is especially the case for hurricanes and major hurricanes.



Also, stratospheric warming events overall are more common towards the extremities of the solar cycle (maximums and minimums), and last year and 1985 are certainly not exceptions to this rule. With the current solar cycle anomalously low, (one of the weakest in about a century) based on the correlation between the solar cycle and major stratospheric warming events, this suggests that the likelihood for stratospheric warming events will be significantly higher than usual for the next several years.

Stratospheric warming, sunspot cycle & QBO relation


This video below, courtesy of the BBC helps to visually explain what I have talked about above, in case there are some concepts that seem hard to understand.


Another video from the UKMET Office also offers some information on stratospheric warming events.


Going back to what was being discussed earlier in the post regarding MJO forecasts, in general, the GFS model is notorious for being far too progressive with the handling of the MJO, especially over the Pacific, and this reflects upon the model’s skill in forecasting and handling tropical cyclone energy. Errors the model has in being too fast with tropical energy also have effects upon its performance in forecasting winter storms and other features because given the fact that the oceans have 1100x the energy capacity of the atmosphere…….

Heat Capacity of air & Ocean

Also, the difference in the amount of energy per degree increment increases as temperature increases point to the tropical oceans as being the “ultimate driver” of the weather patterns & oscillations, thus, if a model can not handle tropical energy, it will not understand the feedback tropical energy has into to the mid-latitudes & polar regions, thus forecasts for arctic outbreaks, winter storms, and even severe weather outbreaks may be off. The ECMWF on the other hand, although it is the best computer model when it comes to weather forecasting, it seems to struggle in that it holds energy too long over the Maritime Continent region (near octant 4) of the MJO, and this may be linked to standing mountain waves, which also are linked to stratospheric warming events, and this evidence suggests that standing mountain waves are a potential factor to the ECMWF’s issues in handling the MJO over the maritime continent.


Looking into this upcoming hurricane season, there are some things to note concerning the conditions at hand such as the 500 millibar height anomaly, 400 millibar temperatures over the deep tropics, the Indian Ocean Dipole, and ENSO among many other things


Looking at sea ice check out what is going on at the current time, notice how much higher sea ice is this year than the past several years, despite the peak being lower this year in the winter.

JAXA sea ice extent



DMI sea ice extent



NORSEX sea ice extent



NORSEX sea ice area way above many recent years



Why is the arctic sea ice stronger this year? Look at the temperatures, and you can clearly see (with the green line being “normal”, red being actual temperatures), that from 80 degrees north to the pole, temperatures for at least for the last few months have been significantly below normal, not burning away like some climate change alarmists want you to think.



Compared to 2007, the ice at this current time is holding together much better, 2007 was breaking up near northern Asia & towards Alaska, however, the cold PDO has helped to alleviate some of those issues on the Pacific side of the arctic. the Atlantic side has melted more, and that makes sense as the AMO continues to be well entrenched into its warm state.



However, even with some recent gains this year, still overall trend in declining sea ice (although certainly no “death spiral”) since the AMO flipped into its warm cycle in 1995, since then ice has decreased dramatically, before then from 1979 to 1994, ice relatively steady.


The southern hemisphere on the other hand doing great, near the top of the pack (in yellow), and ahead of last year, which broke many daily records (almost the all time record) for sea ice in the freezing season.

(link) http://arctic.atmos.uiuc.edu/cryosphere/antarctic.sea.ice.interactive.html

Even the NSIDC shows the Antarctic sea ice well above normal



Antarctic sea ice since 1979 continues to rise, with it especially rising since 1995 when the AMO flipped to warm and all-time high satellite records have been set as the PDO flipped to cold in 2007



Reason for the higher arctic sea ice can be attributed to the +NAO



Remember in my last post where I mentioned cold over the Canadian & Greenland Arctic this year in many of the models as further evidence along with the sunspot cycle years & the worst hurricane landfalling years, which all had +NAOs as evidence that based upon the conditions at hand, this would force a +NAO, which in turn would lead to more arctic sea ice (comparably to the last several years), with the minimum this year likely ending up around the mid 2000s average, still low, but certainly not a “death spiral”.


CFSv2 temperature forecast shown in my last post, showing my ideas for colder than normal conditions over the Greenland & Canadian Arctic, indicative of a +NAO, which spells trouble for the upcoming hurricane season.



Look at this temperature anomaly map over North America, and you can see (in shades of blue & purple) the colder than normal conditions towards the Greenland & Canadian Arctic. Amazingly, even despite the recent record warmth in Alaska, much of the state is still below normal since the start of May.



Thus, with a +NAO in place, this will naturally strengthen the Bermuda-Azores High, forcing storms in the Atlantic, and also helps to force the infamous blocking pattern over Atlantic & eastern Canada that blocks the natural northeasterly progression of tropical systems in the Atlantic, forcing them into the US coastline.



500 millibar heights show my predictions from over a month ago for a +NAO verifying, and with blocking over the North Atlantic, this is very concerning as the higher than normal pressures (shown in green & yellow) reduce the frequency of troughs of low pressure that help to steer storms out to sea, thus this is yet another negative sign for the upcoming hurricane season.

500 millibar heights since the beginning of this May show a very concerning pattern with a lot of blocking over the North Atlantic, limiting the chances for tropical cyclone recurvature before the US.



This is the opposite of the last several years, which had a lot of troughing over the North Atlantic and a negative NAO, with high pressure up & down the US coast helping to act as a shield against most tropical cyclones that approached the US coast and the trough over the north Atlantic had the ability to pick most storms up & push them out to sea. As shown by the picture above, this pattern isn’t  in place this year to help save the US coast once again.


Also, notice my predictions for a wetter than normal southeastern US in relation to the hurricane season in that, when conditions are wetter than normal over the southeastern US, this implies that there’s tropical activity present because if you have for example 5 years of rainfall data, 4 of those years ‘normal” rain falls, yet the 5th year, you have a tropical cyclone that produces flooding, it raises the average rainfall substantially high, thus, when conditions are actually wetter than normal, it should be an indication that tropical activity is present or heavily favored.

Last 90 days precipitation anomalies, notice how wet conditions are near the southeastern US.



Looks a lot like my new hurricane season analog package for Mar-Jun (1960, 1969, 1979, 1996, 2004, and 2010)

Palmer Drought Index of hurricane season analogs


Precipitation, overall, not too bad, hint of wetter than normal look from the upper midwest & Great Lakes into the southeastern US, dry farther south & west.


Now, look at 400 millibar temperatures over the tropics, notably warmer than normal, which promotes tropical development by enforcing latent heat release that tropical cyclones rely upon, and in the pre-season, warmer than normal temperatures here help to moisten the atmospheric column, which helps tropical systems later in the season.


In fact speaking of the warm 400 millibar temps promoting tropical development, looking at 92L, although only briefly an invest area in the central Atlantic, makes quite a statement in the fact that we already observed activity in the deep tropics this early in the season. Remember last season how inactive it was in the deep tropics, no hurricanes east of 70 west and south of 20 north.



In fact, the most powerful storms and all of the major hurricanes (except Sandy) occurred outside the tropics.

Including Hurricane Gordon

GOES infrared satellite image of hurricane Gordon on August 18th last year as the storm was moving generally east-northeastward on its way towards the Azores.



The track of hurricane Gordon last year showing it becoming a category 2 hurricane at its maximum intensity achieving maximum sustained winds of 110 mph and a minimum central pressure of about 965 millibars, making Gordon nearly a major hurricane.



Interestingly, the last hurricane Gordon in 2006 also made a strike on the Azores like hurricane Gordon last year.



Hurricane Kirk

Image of hurricane Kirk on August 30th, as it just became a minimal hurricane and was on its way to becoming a category 2 hurricane, despite forecasts and predictions for the storm to remain significantly weaker.




Hurricane Michael interestingly despite becoming a major hurricane, was hardly even detected by much of the computer modeling forecasts, and there’s plenty of reason for this as Michael was such a small and compact system, reminiscent of a “vortcane” that the models have major issues picking up on the system, and with all of the other various much larger scale systems in the vicinity of Michael such as mid-latitude troughs, upper level lows, etc, these systems can “overshadow” Michael, thus leading to inaccurate model forecasts, which sometimes even hours before the storm’s formation, show little if any change in the environment where hurricane Michael lies. Given its small size, intensity forecasts were greatly underestimated, thus given the conditions and factors at hand, Hurricane Michael “slipped under the radar” per say.



This infrared satellite animation courtesy of the GOES satellite shows Hurricane Michael near its peak intensity on September 6th last year.



Hurricane Nadine, although it officially formed in the deep tropics, it struggled for a large part of its life with dry air issues that were a common theme last year, and it was not until Nadine passed north of 30 degrees north that if finally began to intensify into a hurricane and performed a performed a cyclonic loop due to collapsing steering currents over the storm. Eventually, the storm’s slow motion over the relatively cool waters over the northeastern Atlantic led to significant upwelling, which combined with Nadine eventually moving farther north due to the presence of a mid-latitude trough that also imparted strong westerly wind shear into the storm, to bring about the death of Nadine. In the process, hurricane Nadine was the 4th longest lived hurricane on record in the Atlantic and it racked up 25.6 points of ACE (Accumulated Cyclone Energy, which is calculated by squaring a storms estimated intensity in knots (35 knots or greater is used as that is the required maximum wind speed for a storm to be classified as a named storm) and then dividing that number by 10,000. This calculation is done every 6 hours, thus storms with long durations and higher intensities will accumulate significantly larger sums of ACE), which gave Nadine about 20% of the entire’s season’s ACE.



Track of hurricane Nadine, from its formation on September 11th to its death all the way on October 4th. Such a storm must have been a headache to track and predict, especially for weather forecasters in the Azores, given the very erratic nature of the storm and the varying steering currents in the storm’s environment.



Going back to 92L, an OCEANSAT pass on June 6th shows (in shades of red & brown) winds in excess of 30-35 knots associated with 92L, in which you can see, according to satellite that there was a well defined surface circulation near 13 degrees north, 47 west, denoted by the change in wind direction with winds completely wrapping around 92L’s circulation. (in the bottom right hand side of the picture below). The winds indicated by this satellite pass generally equate to around 35 MPH winds at the surface, thus winds were sufficient enough to classify this system as a tropical depression.



Satellite of 92L shows a rather sheared system with most of the convection weighted towards the northern and eastern sides of its general center of circulation, which actually becomes exposed for a brief period of time. However, with the lack of thunderstorm activity over the center of circulation to promote the enhancement of this center through the process of latent heat to lower surface pressures, which is being caused by a strong trough of upper level low pressure that inflicts an unfavorable 40-50 knots of westerly wind shear, essentially ripping the storm apart, forcing it to go into a more tropical wave-like structure and dissipate.


This image in the early afternoon of June 6th helps to better depict 92L, and you can easily see near the large cloud mass just to the bottom left of the center of the picture below what appears to be some westward movement in the lower cloud mass, which suggests along with satellite evidence that there would indeed be a well-defined, closed low level circulation, and that along with satellite estimates for winds in excess of 30-35 knots would imply this system is indeed a tropical depression. If there is any circulation with this system it is weak in nature given the overall structure of the storm, which looks reminiscent of an easterly tropical wave with convection orientated north to south, the only real difference in this case is there’s a circulation embedded within this wave-like feature.



This type of system threatening development in the deep tropical Atlantic this early in the season is rather unusual as the Cape Verde hurricane season typically does not get going nor favor such systems until later in the hurricane season towards August and September when the ITCZ naturally lifts farther northward allowing for increased thunderstorm activity in the deep tropical Atlantic, which allows for higher levels of moisture to support tropical development, and the increased thunderstorm also help to warm the atmospheric column up to the tropopause, thus “priming” the atmosphere for tropical activity for later in the hurricane season.


Here is the position of the ITCZ (Intertropical Convergence Zone) in the month of January and the associated positions of the 500 millibar heights (red higher than normal pressures, blue lower than normal pressures.). In this picture below, I inserted arrows to help to denote the amount of movement in the Intertropical Convergence Zone over the course of the year, as the natural changing axial tilt of the earth along with other more localized factors help to determine the changing position of the ITCZ.



What is very interesting about looking at the natural movement of the ITCZ over the course of the year is the fact that I take note of how the ITCZ seems to change in position greatest in the presence of a significant continental landmass, and hardly ever moves at all over the eastern equatorial Pacific, although this is not true on a year-year basis as changing states of the ENSO dramatically affect the location, placement and strength of the ITCZ in this region. La ninas generally weaken the ITCZ and force it towards the northwards more towards central America and the Caribbean, where as el ninos keep it held in check and strengthen its influence over the region With a generally large region of latent heat going off in this area due to the el nino helping to induce rising motion over this area of the tropics, this induces large-scale divergence aloft, and thus areas to the east, downstream of this upper-level anticyclone in the deep tropical Atlantic experience higher than normal westerly wind shear as a result, which is one of the reasons as to why el ninos hinder the development of tropical cyclones in the Atlantic.

What I find interesting in looking at this picture below, the greatest movement in the ITCZ is over towards the Indian Ocean in association with Indian Monsoonal Circulation. This large seasonal variation in the ITCZ over this area can be partially attributed to the Himalayas and the effects of standing mountain waves (as earlier discussed in this post) which force air to converge on the windward side of the mountains, and air to spread out and sink at the surface on the leeward side of the mountains, and this sinking of the air at the surface is associated with compression, and this forces the air at the surface to warm as a result of the compression. A similar process occurs from the summer to winter in this region, where large amounts of air coming off of the ocean as result of the land heating up more than the oceans because of how the oceans have 1100x the heat capacity of the atmosphere, (which does not allow them to warm up or cool down as quickly as the land surrounding them) forces air to rise on the windward side of the mountains, thus enforcing a general region of low pressure throughout this region and over the Asian continent. Then, in the winter, air descending down the leeward side of the himalayas forces the air to compress and warm, forcing regions of high pressure as shown in the picture above in shades of orange. You should also notice how in the northern hemisphere that much of the low pressure is focused over the continents, while areas of high pressure dominate the Atlantic and the Pacific Oceans, and this is due to the fact that the landmasses naturally warm faster than the oceans, thus warm air is favored. Considering warm air naturally rises, this forces areas of low pressure over the continents, and since the oceans are naturally cooler, this forces air to sink, thus forcing areas of higher pressure over those areas. Also, you should notice in the picture below that the greatest region of low pressure is over the Asian continent, and this makes sense as it is the largest landmass in the northern hemisphere and the greatest area of rising motion in the summertime due to the land having the ability to warm faster than the surrounding oceans, would be associated with this region. It’s also why North America, although a significant landmass in the northern hemisphere still has relatively higher pressures than the other continents because it’s much smaller in comparison to Eurasia and Africa, thus any rising motion associated with North America would be less significant and also the fact that it is in much closer proximity to the Atlantic and Pacific Oceans means that surface pressures are generally higher, at least in comparison to Eurasia and Africa. Another interesting fact about the ITCZ is that it rises much farther away from the equator in the northern hemisphere summer than it does in the southern hemisphere summer because of the presence of more land in the northern hemisphere to ocean than the southern hemisphere, and this increased land influence results in larger temperature fluxuations that allow for greater movement of the ITCZ in the northern hemisphere than in the southern hemisphere.



This same concept of  how the seasonal variation of the ITCZ operates based on temperature differential and differences in relation to land and water in the northern and southern hemispheres on the global scale can also be applied locally, this time, we are going to look at the Indian Monsoon Circulation. The Indian Monsoon Circulation varies throughout the year with the summer generally featuring the ocean being cooler relative to the surrounding landmass, thus regions of higher than normal pressure develop over the ocean and the warmer air over land naturally rises inducing low pressure, and as a result, knowing that air naturally flows from regions of high to low pressure, this forces warm, moist air from the Indian Ocean over southern Asia, the Middle east, and eastern Africa. As the air rises due to a combination of orographic processes (which involve the forcing of air to rise due to sudden changes in topography, such as mountains, that naturally force air upwards) and Adiabatic cooling (which is the natural decrease in temperature of a parcel of air due to increasing altitude), it condenses and comes down in the form of rain, much of which is needed in this region as upwards of 70-80% of their rainfall comes from the summer monsoon.



The Indian Monsoon Circulation is also heavily related to the Indian Ocean Dipole Index, which is a measure of the difference in the water temperatures over the west Indian Ocean between 10 degrees north and south latitude and 40 and 70 degrees east longitude, and the eastern Indian Ocean between 0 and 10 degrees south latitude and90 to 110 degrees east longitude. When water temperatures are warmer in the western Indian Ocean, comparative to the eastern Indian Ocean, this is a positive (+) Indian Ocean Dipole, where as the opposite is true when the waters are warmest over the eastern Indian Ocean. This picture below helps to illustrate this relationship between the water temperatures in the western and eastern Indian Ocean and the Indian Ocean Dipole.




The Indian Ocean Dipole has fairly large significance in this region of the globe on influencing monsoon and rainfall patterns, which also in turn greatly affect temperature patterns over these areas. Australia, for instance heavily relies upon the effects of ENSO la nina or el nino conditions to help determine rainfall in a particular year or season, however, the Indian Ocean Dipole has great significance on rainfall, just take for instance the top 10 -IOD years (1958, 1960, 1964, 1971, 1974, 1975, 1989, 1992, 1993, and 1996) and look at the rainfall in Australia in a composite of those years. Clearly, a -IOD increases rainfall, especially across southern Australia. (indicated by shades of green & blue which are indicative of above normal precipitation)



What’s interesting about those top 10 negative IOD years, is when you compare a spring temperature composite of those years against this past spring, there is some clear resemblance, with cold centered over the plains that dives southeastward into the US, and with comparatively warmer conditions towards New England & the western US.

Top 10 negative IOD years US spring temperature composite


This past spring, interesting how similar this year is to the top 10 negative IOD years


Now, narrowing down these negative Indian Ocean Dipole (-IOD) events to now just neutral ENSO years, as we are currently experiencing ENSO neutral conditions over the equatorial Pacific, we are left with the years 1958, 1960, 1989, 1992, and 1996, and those winter and spring precipitation composites looked like this over Australia, notice wet to the south, drier to the north and west.



This year thus far, notice however, this time period goes from January to June, the picture above covers the southern hemisphere spring and winter, from June to November. Despite this, you do notice that there is some resemblance here with it wet farther to the south and over western Australia and drier than normal conditions farther to the northeast.



What I find fascinating about these neutral ENSO, negative Indian Ocean Dipole years (1958, 1960, 1989, 1992, and 1996) is that the associated hurricane seasons for the most part, were quite devastating on the US coast.


1958 hurricane season



Hurricane Helene passed within 10 miles of the eastern North Carolina coastline, saved in part by a passing trough that was able to pick up Hurricane Helene in time to prevent landfall, however, with Helene being a category 4 hurricane at its peak intensity and with its close proximity to the coast, it was still a destructive storm, with sustained winds reported near Wilmington, NC at 88 MPH with gusts reported up to 135 MPH.

Hurricane Helene (1958) track



The system that really grabbed my attention, oddly enough, was a subtropical storm that developed at the beginning of the season, in the Caribbean in May and came northward, being a close call for the east coast.



I couldn’t help but notice just how close this system’s track resembled Hurricane Gerda (1969) (in my #1 analog year for this upcoming hurricane season.)



1960 hurricane season (one of my top 5 analogs for this upcoming hurricane season) With Donna & Ethel being the most devastating storms of the season (Donna on the east coast, Ethel in the central Gulf of Mexico), also both systems were category 5s.



1989 hurricane season



Hurricane Hugo, interestingly was the natural disaster that helped to bring popularity to the Weather Channel and make them a household name

1992 hurricane season, despite being an el nino, resulted in category 5 hurricane Andrew on the south Florida coast.



Hurricane Andrew (1992)



Also, I did a very interesting tidbit on hurricane Andrew’s unusual strengthening as it approached the US coast and was able to attribute this to an equatorial Kelvin Wave. Here is that information once again (I had to review it again myself to fully understand the processes involved)

“You can get a better view of the overall conditions at hand in the US water vapor satellite, for about the same time frame showing Hurricane Andrew approaching the south Florida coast, and you can really see the big “hump” in the jet stream over the eastern US and into eastern Canada, indicative of ridging, and of course as mentioned in other times in this post, ridging in eastern Canada is closely associated with big hurricane landfalls on the US east coast, and Andrew certainly fits into this criteria. You can also see a swirling area of low pressure located over extreme northwestern Mexico and moving northward into the southwestern US along with a “bowling-ball” low pressure system evident over the eastern Pacific several hundred miles to the west of California, which will help to kick this trough quickly east and help to “catch” Andrew and veer the storm quickly eastward. Also of note is the large amount of thunderstorm activity present towards the ITCZ (Inter-Tropical Convergence Zone) towards northern South America and extending into the eastern Pacific. This would potentially give an indication to me that there perhaps is an upward MJO pulse over this area, because after all, upward MJO pulses are associated with rising motion over the tropics (which usually means more thunderstorm activity over the tropics and the ITCZ) and perhaps in a hurricane season like this with el nino conditions in place, this may be exactly what Andrew needed to have in order to intensify.”


When I actually went and looked back into the SOI daily data files, available at http://www.longpaddock.qld.gov.au/seasonalclimateoutlook/southernoscillationindex/soidatafiles/DailySOI1887-1989Base.txt I made a monumental discovery. Since I have always wondered in awe at the rapid intensification of Hurricane Andrew over the Atlantic, I’ve always pondered how this came to be, perhaps I may have found the underlying cause to this storm’s intensification. I don’t know if any of you know this, but there has been significant research into the study and effect of Kelvin Waves on hurricane intensification. Generally, as depicted by the picture below, equatorial Kelvin Waves, and their propagation eastward across the global tropics is unique in that as they move into a certain region, they have a tendency to lower the atmospheric pressure and enhance thunderstorm activity over the tropics. In doing so, this forces air to rise upwards and outwards in the upper levels of the troposphere to compensate for the rising motion near the surface, and as noted in previous posts, this increased thunderstorm activity during the winter can help to reverse the natural cyclonic wind flow over the stratosphere as the latent heat release from the condensation of water vapor (an exothermic reaction), increases the vertical propagation of energy upwards into the atmosphere, and in doing so, can turn the QBO (Quasi-Biennial Oscillation) into its easterly state, and when extrapolated over the entire northern hemisphere, this gives the impression of anti-cyclonic motion, which severely disrupts the cyclonic polar vortex in the stratosphere. Well, on the frontward edge of such Kelvin Waves (would be in the middle of this picture below), note how as air is first being driven upward while the surrounding air mass in front is continuing to spread out. Well, if you actually think about it, in this precise region on the edge between the downward and upward Kelvin Wave, divergence is the greatest in the upper levels of the atmosphere, and when you consider that in hurricanes the upper levels of the storm must be able to properly “evacuate” air out of the lower regions of the atmosphere and spread out in all directions in the upper levels in order to allow more air to rush in near the surface, and strong divergence aloft is extremely favorable for hurricanes. Now, consider this point and look at a depiction of the equatorial Kelvin Wave so you can have some idea of what i’m talking about here. Notice in the middle, where the Kelvin Wave actually is, (lower pressures in the wake of the Kelvin Wave, with higher pressures in front) if a hurricane is at that point, air will spread out in both directions from that location aloft, giving extremely favorable upper level atmospheric conditions to allow for strengthening, potentially rapid at that, for hurricanes.


Look at how vital in this pictorial below in how divergence aloft is extremely vital and necessary for hurricane development, and it would only make sense that at the Kelvin Wave boundary, where divergence is greatest aloft that the most favorable environment for strengthening would be located there.


When I actually looked into the daily data for SOI in August of 1992, this is what came up 1992 229 1016.33 1013.65 6.45 1992 230 1015.74 1012.85 7.73 1992 231 1014.39 1012.35 2.52 1992 232 1013.11 1012.35 -5.20 1992 233 1012.20 1013.20 -15.90 1992 234 1012.27 1013.30 -16.06 1992 235 1011.77 1013.10 -17.93 1992 236 1012.08 1012.55 -12.69 1992 237 1013.99 1012.75 -2.33 1992 238 1014.51 1013.20 -1.86 In this data above, the column on the far right indicated the year, which if course is 1992, the next column, being the day of the year, in this case from day 229-238 (Aug 15-25), and the numbers in the middle are the measured surface pressures from Tahiti, French Polynesia, and Darwin, Australia, from which the SOI values are derived. The actual daily SOI values are shown farthest to the right, and what is indicative of a Kelvin Wave, is lower than normal pressures at Tahiti along with above normal pressures at Darwin, Australia, and when this is the case, the SOI is negative, so we are looking for -SOI numbers, the bigger the negative, the stronger the Kelvin Wave. Look what happens as we get towards day 232, the SOI plummets to -5.2, then by day 235 has bottomed out all the way down to -17.9. Now, knowing of course it does take a little time for the Kelvin Wave to move east in the tropics, give it a few days from about August 18th-20th or so, and then the Kelvin wave should have moved into the Atlantic. Now, compare this information to what was observed with Hurricane Andrew during this time, with Andrew at one point being an open tropical wave, as noted by hurricane hunters on August 19th which had failed to find a well defined center of circulation. However, by August 22nd Andrew became better organized and an eye began to form and became noted on satellite. By later on August 27th, Andrew underwent a period of rapid intensification, which stunned many forecasters and people alike as its minimum central pressure dropped 47 millibars in just 24 hours (very explosive intensification) and by literally the next day had went from a minimal hurricane to one that attained a minimum central pressure of 922 millibars and had maximum sustained winds of 175 MPH as it approached the Bahamas, and such intensification certainly caught people off guard and left many in the Bahamas and south Florida scrambling to get ready for the impending hurricane. Of course, knowing that the arrival of equatorial Kelvin Waves bring with them large amounts of divergence aloft, thus enhancing conditions for any developing tropical cyclone, and that the SOI crashed (indicative of a Kelvin Wave) at the exact same time Hurricane Andrew underwent dramatic intensification, perhaps I may have discovered the reason behind the seemingly mysterious and unexpected intensification of Hurricane Andrew on August 22nd, 1992, and of course thanks to the rapid intensification of the storm, with my thinking this was due to a Kelvin Wave, this day would forever change the history of not only hurricane forecasting, but also the lives and property of those across south Florida and the Bahamas who went through the deadly and very historic hurricane. – See more at: http://weatheradvance.com/2013/03/24/hurricane-season-forecast-late-winter-ramblings/#sthash.O43Bu5E6.dpuf

Great video showing the evolution of Hurricane Andrew as it happened, and you can see over the course of the video how the storm rapidly intensifies, likely as I suspect, from an equatorial Kelvin Wave (who knew?) You should also notice in the rapid intensification of Andrew on the satellite loop, the very noticeable increase in the expansion of the upper level clouds associated with Andrew, indicative of the very favorable conditions aloft I was talking about, and as I think this is most likely due to an equatorial Kelvin Wave (indicated by -SOI) enhancing upper divergence aloft in the environment the storm is in.

Going back to Australian rainfall for a better comparison of current conditions, here are the April-June precipitation anomalies over Australia in the top 10 -IOD years.



Look at this year since the start of April (shades of yellow and green indicate wetter than normal conditions, blues drier than normal), looks almost exactly like the picture above.



What I find very interesting about Australian rainfall in relation to the IOD, look at the highest hurricane landfalling years on the US coast since 1950 (1950, 1953, 1954, 1955, 1959, 1964, 1971, 1979, 1989, 1998, 1999, 2004, 2005, and 2008) and compare Australian rainfall for the Jan-June time period, it is very evident that a -IOD is in place, with the well below normal rainfall farther to the north and west, wetter than normal farther to the south and the east. (blue and purple drier than normal, yellows, greens, and oranges wetter than normal)


Interestingly, this looks very similar to -IOD and neutral ENSO year’s rainfall in Australia. (reds, yellows drier, blues and greens wetter than normal)



This is just another reason why I put plenty of focus on ENSO neutral years for US tropical cyclone landfalls, and patterns in Australian rainfall offer further evidence for my observation in the westward change in tropical cyclone trajectories on the US coast from la nina to neutral years, which means that storms in neutral seasons compared to la nina, have a tendency to come farther west, and in doing so, put a larger amount of the US in harm’s way from tropical cyclones.


East coast la nina hurricane hits

E Coast Hurricane Map La nina yrs


East coast neutral ENSO hurricane hits, notice how compared to la nina years, the general track of hurricanes that hit the east coast shift considerably to the west, putting a much larger area of the eastern US in the path of hurricanes.

E Coast Hurricane map Neutral yrs


Florida la nina hurricane hits

FL Hurricane Map La nina yrs


Florida neutral ENSO hurricane hits, again, notice the significant shift westward in hurricane tracks.

FL Hurricane Map Neutral yrs


Looking into what mechanism courtesy of a -IOD would force an increase in hurricanes in the Atlantic, which especially under ENSO neutral conditions like what is currently being observed, help to lead to more US hurricane landfalls. Take a look below at a graph of the  Indian Ocean Dipole, and in this I have circled the areas of interest in accordance to the time of the year and have labeled the graph with “warm” and “cold” to help denote the positive and negative states of the IOD respectively, and also the “warm” and “cold” correspond to the overall temperature compared to normal over the northwestern Indian Ocean near the east African coastline.

Indian Ocean Dipole US winter Atlantic Hurricane season


Notice in the picture above that there are two general peaks in the IOD towards its positive state in 2011 and 2012 and two minimums in the negative state in 2009 & 2010. If you recall what the beginnings of those following winters were like the last 4 years, it is interesting to note that in the years of 2009 and 2010 that experienced large crashes in the IOD into its negative state in the middle of the summer that the following Decembers were cold in much of the US.

December 2009 US temperature anomaly, notably cold over much of the US, especially over the Rockies and points westward.


Remember that year also featured a rare white Christmas over the southern plains with a blizzard all the way up into the upper midwest and Great Lakes.

Surface analysis animation of the 2009 Christmas snowstorm showing multiple areas of low pressure ejecting out of the south-central Rockies and southern Plains and moving northeastward towards the northeastern US and Great Lakes, bringing with it, very heavy snow & blizzard-like conditions on the northwestern side of the low pressure area, and even some tornadoes & severe weather towards the Gulf Coast states. Also notice the rather anomalous high pressure area that was established over Ontario in the wake of a frontal boundary and enhanced by a strong surface snowpack (snow reflects 85-90% of solar radiation, cools the surrounding air, in turn cold air sinks, forces pressures to rise), this high pressure block helped to force air to rise underneath it and force pressure falls over the central US where this low pressure intensified quite rapidly.


NWS Norman, OK snowfall totals

OK Christmas snowstorm 2009 snowfalltotals


NWS Kansas City snowfall totals



NWS Duluth snowfall totals

2009_ Christmas_ Snowstorm2


Watches and warnings issued for the US at the time of this storm show the large expanse and power of this storm, with blizzard and winter storm warnings from Wichita Falls and near Dallas TX all the way up to the Canadian border with Manitoba and Ontario. Ice was also an issue on the eastern side of the system with a cold air damning set-up on the eastern side of the Appalachians up the I-81 corridor from Virginia into western New York where warm air forced aloft rose over the cold surface air that was put in place by a previous frontal boundary.



The National Snow Depth Analysis of this event reveals the snow left by this storm system,  over the southern plains, with a rather extensive snowpack already in place over much of the US, including the eastern US east of the Appalachians, and it was because of this snowpack that this storm initially began as a wintry mix or ice for many of these areas as the snowpack helped to enhance cold air at the surface to allow for freezing rain & sleet (if the layer of cold air at the surface was deep enough).



December 2010, following yet another major drop in the Indian Ocean Dipole in the middle of the summer into its negative state, cold temperatures were once again a common theme over the US at the beginning of winter.

December 2010 US temperature anomaly


Interestingly just like the year before, another major winter storm affected the US near Christmas, this time it hit areas farther east, and this is likely due to the core of the cold being farther east in December 2010 than December 2009, which helped to force the storm track as a result farther east as well.

2010 Post Christmas snowstorm

Accuweather snow totals from the 2010 post Christmas snowstorm, many areas up the eastern seaboard from North Carolina, through Philadelphia, New York City, Hartford, and Boston experienced 1-2 feet of snow, with some areas approaching 30 inches



Interpolated snowfall for December 24-27 2010 over the eastern US showing an initial axis of snow over the Ohio & Tennessee Valley, with another rather large band of snow associated with a main area of low pressure that developed just off the east coast as result of a phase between a clipper system in the northern branch of the jet, and a storm system in the subtropical jet that developed off the southeast coast.



Another snow map of this event developed By Raymond C. Martin Jr.



That winter and the winter of 2009-10, gave many areas more than their share of winter snowfall, some areas picking up nearly a year’s worth in just one storm (this was true for areas farther south & east where yearly average snowfalls are generally lower)

Northeastern US average winter snowfall



The two Decembers (2009 & 2010) averaged together produced this for temperatures, a tongue of cold extending from the northern plains southeastward into the eastern US, with some residual water near New England.


Somewhat surprising how close these Decembers are to what the temperature pattern looked like for this past winter


However, referring back to the IOD, the next 2 summers (last summer and 2011) featured a large spike in the Indian Ocean Dipole well into positive territory, and interestingly, both of those Decembers were blowtorch warm over much of the US.

December 2011 temps following the first major DOI + summer spike. This month marked the beginning of a very warm and “boring” winter for many in the eastern US.

December 2011 US temps


The following December which came also on the heels of a positive Indian Ocean Dipole spike the previous summer was also once again blowtorch warm across much of the US like the previous December.

December 2012 US temps


Now that you see there is a clear correlation between the summer Indian Ocean Dipole and US temperatures, take a look at the effects the changing Indian Ocean Dipole has on the associative hurricane season.

First 2010, a strong la nina year, as expected should be very active in the deep tropics, and that certainly was the case, although, it got off to a slow start due to an abnormally strong Saharan Air Layer which is associated with Saharan Dust that promotes general sinking of the air and naturally dries out the atmosphere, making the environment generally more difficult for tropical cyclones, however, 2010 still was able to crank out Danielle, Earl, Igor, Julia and Lisa in the deep tropics.

2009 on the another hand, was an el nino year one would expect for activity to be lower than normal as el nino tends to increase westerly wind shear, and this certainly was the case in the 2009 hurricane season as only 9 named storms formed that year, below the average of 11 named storms in the Atlantic, and certainly well below the 1995-2012 warm AMO named storm average in the middle teens. Yet, despite conditions being unfavorable for tropical cyclone development, 2 major hurricanes formed in the deep tropics (from 60 degrees west longitude, to the west African coast up to 20 degrees north), hurricanes Bill and Fred.

Now, the reason why I specify the deep tropical Atlantic as being in between 60 degrees west longitude to the west African coast, to all the way up to 20 degrees north is because of the fact that this area is east of the Lesser Antilles, thus not being associated with the Caribbean, and the fact that when you look at regions for tropical cyclone formation, this area is the general breeding ground for tropical cyclone genesis as tropical waves travel westward from the coast of Africa, relying on baroclinic processes for development, have to transition into one that relies off of the release of latent heat energy. Also, this region is a transition zone between storms that form closer to Africa generally hitting the eastern seaboard and storms forming closer to the Lesser Antilles generally affecting the Gulf of Mexico.

To understand more visually what I’m talking about, here’s a picture that helps to outline the region I denote as the “Deep tropical Atlantic”



When looking at the 2009 hurricane season, you can clearly see in the area I defined as the “Deep tropical Atlantic” there were two major hurricanes, Bill and Fred.



Hurricane Bill on August 19th near its maximum intensity as a category 4 hurricane northeast of the northern windward, Virgin Islands & Puerto Rico.

Hurricane Bill Aug 19th 2009


Hurricane Fred near its maximum intensity as a category 3 hurricane in early September 2009



2010, being a la nina year made conditions overall generally more favorable for tropical cyclone activity, and with a negative Indian Ocean Tripole in place, this helped to further enhance African easterly waves by enhancing the Indian Monsoon and precipitation surrounding the Arabian Sea, which added more moisture and available instability for waves, and with a strong SAL that season, this made for a powerful African easterly jet that was able to produce several hurricanes in the deep tropics.

In all, 6 hurricanes were present in the “deep tropics” in 2010, including Danielle, Earl, Igor, Julia, Lisa, and Tomas.



However, compared to 2009 & 2010, the 2011 hurricane season in the tropics was much different, despite a la nina, once again being in place, and this can be attributed to the Indian Ocean Dipole turning positive. In doing so, this lowered the Monsoon towards the Arabian Sea, thus limiting the capability of African easterly waves, and with limited storm activity, dry air would be a large issue as well as SAL. In fact, what is interesting about 2011 is it bears some resemblance to the following hurricane season, in that many of the season’s storms formed outside of the deep tropics.


2011 in the deep tropics to some was surprisingly inactive with numerous tropical storms, yet only one hurricane, (Katia) and no majors. The only major hurricanes (Irene, Katia, and Ophelia) all occurred outside of the deep tropics (also like 2012), indicative that conditions within the deep tropical Atlantic were relatively unfavorable.



The 2012 hurricane season, although not fully entrenched in a la nina ENSO state, was anticipated to be an active hurricane season, however, there were signs given the colder than normal 400 millibar temperatures in the deep tropics that activity would be limited there. The reason for the 400 millibar temperatures being cold preceding the 2011 season can be attributed to the Indian Ocean Dipole flip to positive in 2011, and we observed in that hurricane season the beginnings of what appeared to be an increasingly unfavorable environment there, which only worsened as the Dipole remained positive. With the IOD positive, this shuts down on the monsoon over the Arabian Sea and African continent, thus less moisture & instability in a general sense is available for African easterly waves as they traverse the continent. The reduced storm activity as a result of the Indian Ocean Dipole into its positive state leads to a cooler troposphere as the release of latent heat into the atmosphere is reduced, this forces the troposphere to shrink, and thus the cooler than normal conditions not only have the ability to hold less moisture than warm air and promote dry air in the atmosphere of the deep tropics, but the cooler than normal conditions in the troposphere also cut-off tropical cyclone’s main energy source of latent heat release. Thus, cool 400 millibar temperatures as a result of a positive Indian Ocean Dipole leads to an unfavorable environment for tropical cyclones in the deep tropical Atlantic.

Look at how there were no hurricanes in the deep tropics last year, only some tropical storms, but even those systems struggled immensely to develop.



In general, the positive Indian Ocean Dipole reduces tropical cyclone activity over the deep tropical Atlantic because of the fact that when you have warmer than normal waters over the northwestern Indian Ocean in comparison to near Indonesia, what this does, is knowing that the monsoon itself is a product of the difference in pressure created by a difference in temperature that naturally occurs in the summer time because of the different heat capacities of the atmosphere and the ocean, in which, the landmasses, naturally having 1100x less the energy capacity of the ocean, are susceptible to heating & cooling much faster than ocean, so in the summer, with increasing solar radiation, the landmasses surrounding the Indian Ocean warm-up much more in comparison than the water around them, and due to the fact that warm air has a tendency to rise, this forces lower than normal pressures over land. The higher than normal pressures now created over the water now help to establish a relationship in air flow from the ocean to the land, where areas of high pressure created over the ocean, due to their natural tendency to force air to spread out, force warm, moist air to pile up into the continental regions. In combination with orographic lifting and daytime instability, thunderstorms form, and these thunderstorms release latent heat into the atmosphere. Due to the natural progression of weather to move from east to west in the tropics, the thunderstorms, and the latent heat as well as the moisture associated with them also in turn, move west. To the west of the Indian Ocean is the African continent, thus these thunderstorms, forced to ring out their moisture near the Ethiopian highlands in eastern Africa, add more moisture, instability and latent heat into Africa to help fuel African easterly waves. Considering that African easterly waves are known to create about 90% of all major tropical cyclones in the Atlantic basin, the relationship between the African waves and the Indian Ocean Dipole is important to understand. Thus, when you have the Indian Ocean Dipole warmer than normal like it was in 2011 & 2012, instead of promoting rising motion over that region of the globe because of the warm waters in place, they actually have the reverse effect on continental areas that surround the Arabian Sea, like eastern Africa, because now with the ocean much warmer (although likely still colder than the surrounding land), pressures do not rise nearly as much and air flow into the land regions is limited, thus the monsoon is weak and sporadic. As a result, less latent heat is pumped into the atmosphere, and drier air surrounds this region of the globe, and knowing weather moves from east to west in the tropics, this has negative effects on African easterly waves by reducing their overall intensity. This leads to naturally weaker systems entering the deep tropical Atlantic, and with drier air promoted, this also enhances the Saharan Air Layer, which puts a further negative effect onto any developing tropical cyclone, which is why years like 2011 & 2012 were not very active in the deep tropics, even despite 2011 being a la nina year. It’s quite a statement when an el nino year, 2009, had more major hurricanes in the deep tropics than 2011 & 2012 combined, should be an indication to many that there are other factors than just ENSO, sea surface temperatures, etc. playing their hand in the hurricane season. So, with this year currently observing a major crash in the Indian Ocean Dipole, this certainly tells me that this year will be much more active in the deep tropical Atlantic, and without el nino in place to negate major hurricanes & with 92L earlier in the season having almost developed despite it being only early June, is a very clear indication that this season could turn out to be quite rough indeed, and with more activity closer to Africa, this naturally will put the US eastern seaboard under relatively high risk for hurricane landfalls this year.

As far as the winter goes, and the relationship between the winter pattern & the Indian Ocean Dipole, I suggest that perhaps this upward motion associated with the enhanced Indian Ocean Dipole does not immediately go away towards the end of hurricane season, rather, it continues to stick around right on into the winter, and as a result of all the enhanced upward motion as a result of the negative Indian Ocean Dipole, this attracts the upward pulse of the MJO. Looking at the MJO in the winter, when it enters generally octants 8-2, this helps to promote stronger than normal troughs over the US in general, especially the eastern US, and leads to colder than normal conditions.

MJO temp composite DJF, notice on the left side of the picture below with the temperature composites for the US, (reds above normal, blues below normal temps), look at how when the MJO enters octants 8,1, and 2, how conditions are generally cooler than normal, especially over the eastern US, and when you have a negative Indian Ocean Dipole in the summer, still with lingering effects into the winter, promoting upward motion over the Atlantic, which correlates to octants 8 & 1 (2 is near the Indian Ocean towards Africa), it makes some sense as to why a negative crash in the Indian Ocean Dipole in the summer can lead to at least a cooler than norma start to winter in the US. With this year experiencing a drop in the IOD like 2009 & 2010, it’s definitely plausible to conclude that this winter may get off to a fast start across the eastern US (unlike last year) & the threat will exist for major early season snowstorms much like December 2009 & 2010.



My start of winter forecast released back on April 21st, is supported by the negative Indian Ocean Dipole

start of winter


This is further supported by my observation in the highest hurricane landfalling years on the US coast (as I anticipate this year to do) getting off to colder than normal starts over the eastern US.


The CFSv2 also sees this fast start to next winter, amazing how close it appears to the highest hurricane landfalling years on the US coast, composite temperature map



Also when the natural progression of the Azores-Bermuda high lifts far enough to the north and weakens to reduce the northeasterly trade winds, which in the beginning of the season are naturally much stronger, and as result pumps a large amount of dust from the Sahara Desert in North Africa, and this dust not only dries the air out, but also given the weight of dust being heavier than average air, also leads to subsidence that enforces higher than normal pressures along with dry air, both factors which significantly hamper tropical development.

Saharan Air Layer



Looking at the ENSO, I have actually seemed to find a very interesting correlation between the ENSO index and the Sun’s plasma speed received to earth. Notice in the graph below the very large crashes in plasma speed observed near 1973, 1980, 1987, 1998, and 2010, which are denoted by red circles.



Look at the ENSO index for those exact same time periods, interestingly, there would appear to be a correlation between plasma speed and major el ninos



A reason as to why lower speeds in the sun’s plasma having relationship to the ENSO index and an increase in el ninos can be explained by the solar cycle relationship to the QBO, which I will explain further in upcoming posts as I read through some scientific papers this summer, but variations in the sunspot cycle and TSI (Total Solar Irradiance) can lead to significant changes in the length and strength of the QBO (Quasi-Biennial Oscillation), which is a measure of the direction and strength of downward propagating winds which start from the lower stratosphere and move downward and push into the upper troposphere where they dissipate because of how the temperature gradient is different in the troposphere compared to the stratosphere. The also QBO helps to enhance the trade winds in the tropics, in which the varying trade winds in the tropics also have dramatic effects on the state of the ENSO. If you recall back to a previous post I made earlier this fall, the effects the trade winds have on the ENSO are related to the Thermocline across the Pacific, where under “normal” conditions (la nina), the trade winds are stronger than normal. With the trade winds in the tropics naturally coming out of the east because of the Hadley Cell of energy in which tropical energy is naturally forced upward because of the fact that in the tropics, the greatest amount of solar radiation is found here, thus making the air naturally warm, and considering how warmer air is naturally less dense than the surrounding cooler air in the mid-latitudes, it has a tendency to rise. As the air ascends, it naturally cools with increasing altitude (Adiabatic Cooling) as it does “work” on its surroundings, and considering that air is naturally thinner with increasing altitude in the troposphere because of the force of gravity inducing higher concentrations of air molecules near the surface, the rising warm air decreases in pressure as well. As the air cools, it eventually reaches its saturation point, and thus any water vapor in the atmosphere condenses back into liquid water. This process is known as condensation, and it releases heat back into the atmosphere (latent heat), which is the primary fuel for tropical cyclones, and this is a good reason why warm 400 millibar temperatures promote tropical cyclone development. Once the rising warm, moist air reaches the tropopause, it spreads out in all directions, which generally means away from the equatorial regions, and over time as the air cools, it naturally sinks back towards the surface, and this occurs near the “Horse Latitudes”, where subtropical high pressure areas dominate. This air eventually sinking back to the surface eventually rises again, but this time in the mid-latitudes at the border of the “Polar Cell” of energy, the rising motion is not as strong as in the tropics because the air is naturally cooler here and the tropopause (boundary between the stratosphere and troposphere) is lower. This area of rising motion at the boundary of the polar cell also forms along the polar jet stream and variates depending upon the natural seasonal variation of the polar vortex.



Of course, when you look at the Pacific for the ENSO at the present time, it reveals a la-nina like signature in place, with a pool of cooler than normal waters just to the west of South America, and what appears to be a piling up of warm water towards the western Pacific.


The higher than normal pressures over the Indian Ocean as result of upward MJO pulse bombardment in comparison to the warmth over the west Pacific will force a major slowdown of the easterlies as large scale high pressure tries to take over. Since air spreads out in all directions from high pressure, air spreading out from Indian Ocean goes east, against the tropical easterlies, thus forcing them to slowdown. Slowdown of easterlies leads to el nino conditions as the Pacific warm pool moves east across the equatorial Pacific.

You can see this in looking at the Oceanic heat content cross-section, latitude shown on the time, with time in the vertical direction. Notice how late in the winter there was a large push of cold water that moved across the Pacific (promotes upwelling), thus leading to the observed cold pool anomaly over the eastern Pacific. However, since that time, look how much warmer the western Pacific (on the left side of the picture) has become and you can begin to see the first signs of a push of warming about to come across the Pacific thanks to the eastward moving MJO & associated equatorial Kelvin Wave.


However, given that the waters are already cold in the eastern equatorial regions, the oceanic cycle of the PDO (although briefly warm now) is still in an overall cycle of cold PDO, thus inflicting a significant push of cold water into the equatorial eastern & central Pacific. Also, knowing that by looking at the PDO in relation to the warm AMO, how difficult  is for both oceans to be warm or cold at the same time, meaning that since the Atlantic is warm, the Pacific will tend to stay in an overall state of cold to counteract this, thus any warming that occurs in periods of cold PDO & warm AMO (like what was seen in the 1950s) are relatively brief & weak. Until the Atlantic cools, like it did in the mid 1960s, el ninos that do develop will be weak & relatively transient in nature.

Notice how in looking at the ENSO index, when the PDO was cold from the beginning of the image all the way until about 1977, how until the 1960s, when the Atlantic began to cool, Pacific el ninos were weak in nature.


In essence, as many of you probably already know, the use of ENSO on long range weather prediction is very important and understanding its processes and how it works is key to help unveiling other aspects of weather.

El nino for the upcoming winter does not always mean it will be cold & snowy, in fact, it is the type of el nino that really matters, and when you see a “modiki” , central Pacific el nino coming, this is good news as it is correlated with significantly below normal temperatures over the eastern half of the US, compared to the strongest el ninos, which are actually blowtorch warm.



You can attribute this difference in the location of the warmest water relative to the eastern equatorial Pacific. Having the warmest water farther to the east does not allow for a high pressure break to build in between the Pacific & the Atlantic (at least when the Atlantic is in its warm mode), thus low pressures in the winter, means the continent in the middle of the two oceans (North America) is generally warm. The warm water being farther to the east also does not allow for the Hawaiian Low to form, at least very often, and in doing so, allows for troughiness to build over the Pacific, which in turn forces ridging to over take much of the eastern US, a blowtorch warm set-up for winter that does not tend to favor as significant nor as numerous winter storms.


The last time we had a modiki el nino like what is being forecasted was back in 2004, and of course that is one of my analog years for this upcoming hurricane season, and 2004 was quite devastating for Florida, 4 hurricanes hit the state.










Interestingly, also the winter pattern did bear at least some similarity to this past winter, look at January 2004 temperatures over North America, notice the -NAO look with warmth towards Atlantic Canada and Greenland, with a tongue of cold stretching southeastward out of Alaska into the Northern Plains, Great Lakes, and Northeastern US.



Although this past January was different from the 2004 pattern



It just looks like it was about 3 weeks late

North America temps Jan 21-Feb 21 2013



Another factor that I considered for this upcoming hurricane season was the sunspot cycles, and if you recall in a previous post this is what I said about that in relation to US tropical cyclone landfalls

“Here’s what I collected in the data since 1880, is that there were 452 tropical systems of tropical storm strength or greater to make landfall on the US coast, with a total of 224 hurricanes to make landfall and 78 major hurricanes that have hit the coast. In the sunspot cycle years of extreme maximum or minimum in the overall solar cycle, there were 52 years used including (many of which mentioned above) 1886,1887, 1888, 1889, 1893, 1894, 1899, 1900, 1901, 1902, 1906, 1907, 1911, 1912, 1913, 1915, 1916, 1917, 1921, 1923, 1926, 1927, 1928, 1929, 1933, 1934, 1935, 1938, 1944, 1947, 1948, 1949, 1953, 1954, 1957, 1964, 1968, 1969, 1976, 1979, 1981, 1985, 1986, 1989, 1990, 1991, 1996, 2000, 2002, 2008, and 2009. There were a total of 80 remaining years since then, and when you compare the total number of storms in minimum and maximum sunspot years to non-years you get 204 tropical storms in sunspot years to 248 in non-years. However, it is very significant to see that the number of landfalling hurricanes in the sunspot years at 117 & 42 respectively, is actually more than non-years which have 107 & 36 for hurricanes and major hurricane landfalls respectively. Now, such differences may not seem like much, however, considering that in non-years, there are 28 MORE hurricane seasons into that data as opposed to sunspot years, and there is still significantly less hurricanes and major hurricanes on the US coast, this is very telling that there is indeed some definitive trend in sunspot cycle minimums and maximums to US hurricane landfalls.

The results get even more surprising when you take the average number of landfalling tropical storms, hurricanes and major hurricanes of those years, with the 52 sunspot years which are significantly higher on average than non-years. Average number of tropical storms making landfall on the US coast in the sunspot years analyzed is 3.9 per year, with non-years only at 3.1 per year. Hurricane landfall average in sunspot years is 2.25 as opposed to only 1.3 in non-years, with major hurricane average frequency in sunspot years at .81 with non-years at only .45 on average. This tells you that given the data, tropical storms during extreme periods of solar maximum or minimum are 21% more likely to make landfall on the US coast, hurricanes 40% more likely and major hurricane landfalls are 44.3% more likely in years of solar extremes.

Now, I do recognize that seasons like 2004 & 2005 are certainly exceptions, as are various other years to this overall trend in increasing hurricane landfalls, but when I remove just those 2 seasons from the data going back to 1880, considering them somewhat as outliers, the trend is increasingly evident. Of course, just removing these two years still leaves the non-sunspot years with 26 MORE hurricane seasons than sunspot years. The overall increased probability in the number of tropical storms or greater seems relatively unaffected with a still significant, but not a huge change in overall probability, up to 24% more likely in sunspot years. The hurricane probability seems moderately affected with an increase in difference of probability of about 6% from the data without these years. However, the number of major hurricane landfalls on the US coast is extremely affected without 2004 & 2005, a 11% difference in probability to now about 55% more likely to see a major hurricane landfall in a sunspot extreme year versus a non-year. In fact, with just 2004 & 2005 removed from the non-years there are only 29 major hurricanes in those 78 hurricane seasons to make landfall on the US, while when I looked at just high sunspot cycle years I analyzed like 1893, 1894, 1906, 1907, 1915, 1916, 1917, 1926, 1927, 1928, 1929, 1947, 1948, 1949, 1957, 1968, 1969, 1979, 1981, 1989, 1990, 1991, 2002 & 2002, there were in just those 26 years, 27 major hurricanes on the US coast, which makes both sets of data with 78 non-sunspot cycle years to just 26 sunspot years virtually equal in the amount of major hurricanes on the US coast. This means that when comparing these two sets of data that major hurricane landfalls in high sunspot years are about 280% more likely to occur than in non-years, absolutely amazing. The overall conclusion here is that my original suspicions of that there is a relationship between sunspot cycles and US hurricane and major hurricane landfalls is verified by looking at this data.”


Now, I realize in looking at this data, there are going to be years that correlate much better than others due to more year-year conditions, and variables such as the ENSO index, Indian Ocean Dipole, PDO, AMO, etc, can lead to such changes. Now, the question is why do some years correlate well to the susnpot cycle- US hurricane landfall relationship, while others do not? Well, I got my answer in reading through a scientific paper from Florida State University’s College Of Social Sciences by Robert Edward Hodges titled (link to the scientific paper, so you can see for yourself) http://diginole.lib.fsu.edu/cgi/viewcontent.cgi?article=2036&context=etd

“Evidence For A Solar Influence On US-Affecting Hurricanes”. In this paper on page 26, there was a chart made that distinguished between years that correlated and did not correlate to the sunspot cycle-US hurricane landfall relationship. 5 years were selected in each category that best fit correlating or opposing the sunspot cycle-US hurricane landfall relationship.

The top 5 years the correlated well with the sunspot cycle-US hurricane relationship were

#1 1999 hurricane season



#2 1938 hurricane season



#3 1929 hurricane seasontrack


#4 1950 hurricane season



#5 1871 hurricane season



Now, the top 5 hurricane seasons that DO NOT correlate with the sunspot cycle-US hurricane landfall relationship

#1 1978 hurricane season



#2 2001 hurricane season



#3 1957 hurricane season



#4 1908 hurricane season



#5 1981 hurricane season



In looking over these years, on page 26 of that scientific paper there was a chart which visually showed the actual data between the two sets of years, I will post this below


Years that correlated the most with the relationship of sunspot cycles to US hurricane landfalls

Year           U.S. hurricane landfalls             U.S. major hurricane landfalls

1871                           3                                                             1

1929                           3                                                             1

1938                           2                                                             1

1950                           3                                                             2

1999                           3                                                             1

Average                   2.8                                                          1.2

Years that did NOT correlate the most with sunspot cycle-US hurricane landfall relationship

Year            U.S. hurricane landfalls            U.S. major hurricane landfalls

1908                           1                                                             0

1957                           1                                                             1

1978                           0                                                             0

1981                           0                                                             0

2001                           0                                                             0

Average                   .4                                                            .2

Now, why would their be such a large discrepancy in hurricane and major hurricane landfalls on the US if both sets of years had the same general parameters in place, referring to the extremity of the sunspot cycle? Well, the answer appears to lie within the processes involved between the interaction of the solar radiation and the sunspot numbers in the May-July & November period compared to September on stratospheric ozone and its relationship to stratospheric temperature and the troposphere underneath.


Here’s a link to monthly sunspot numbers going back to the 1600s


To test out this theory to see if there is indeed a relationship between May-July, November sunspot numbers and US hurricane landfalls and to see what potential conclusions can be drawn from such information, look at the monthly sunspot numbers in 1871.

The number farthest to the left shows the year, the next number shows month (5 corresponding to May, 6 to June, 7 to July, etc..), and then the most important number, the third one, shows the actual sunspot number for the given month. It is interesting to note that the sunspot numbers are much higher in the months May-July & November, compared to September, (denoted by the number 9), notice the sudden drop in the sunspot number from 110.1 in August to 80.3 in September, and how it goes back up again towards October & November.


1871 May-November sunspot numbers

1871  5 145.5  21.9
1871  6  91.7  31.6
1871  7 103.0  31.4
1871  8 110.1  37.0
1871  9  80.3  31.2
1871 10  89.0  28.7
1871 11 105.4  29.5


Look at other years like 1999, 1929, 1938, and 1950 to see if this trend of high sunspots in May-July & November and relatively lower numbers in September continues.

1999 May-November sunspot numbers, notice the large fall in sunspot numbers in September (9) compared to the surrounding months

1999  5 106.4  20.0
1999  6 137.7  33.3
1999  7 113.5  26.0
1999  8  93.7  42.7
1999  9  71.5  24.0
1999 10 116.7  27.9
1999 11 133.2  36.2


1929 May-November sunspot numbers, once again, you can see the sudden drop in sunspot numbers near September.

1929  5  58.2  19.6
1929  6  71.9  20.5
1929  7  70.2  15.5
1929  8  65.8  30.2
1929  9  34.4  15.1
1929 10  54.0  22.2
1929 11  81.1  25.7


1938 May-November sunspot numbers, once again, a large drop in sunspots in September

1938  5 127.4  24.5
1938  6  97.5  19.7
1938  7 165.3  30.2
1938  8 115.7  26.9
1938  9  89.6  34.5
1938 10  99.1  38.0
1938 11 122.2  32.6


1950 May-November sunspot numbers, not surprisingly, another large drop towards September.

1950  5 106.2  31.2
1950  6  83.6  19.1
1950  7  91.0  20.6
1950  8  85.2  18.0
1950  9  51.3  19.8
1950 10  61.4  25.9
1950 11  54.8  21.6


Now, look at the years that did not correlate well with the sunspot cycle-US hurricane landfall relationship and had significantly less numbers in US hurricane & major hurricane landfalls

1908 May-November sunspot numbers, notice towards September there’s actually a rise in sunspot numbers, interesting.

1908  5  40.8  23.2
1908  6  48.1  28.8
1908  7  39.5  23.9
1908  8  90.5  29.6
1908  9  86.9  24.2
1908 10  32.3  23.9
1908 11  45.5  24.7


1957 May-November sunspot numbers

1957  5 164.6  33.5
1957  6 200.7  41.7
1957  7 187.2  48.1
1957  8 158.0  43.6
1957  9 235.8  44.2
1957 10 253.8  31.3
1957 11 210.9  27.6


1978 May-November sunspot numbers, look at the very large spike in sunspot numbers near September, exactly the opposite of years with large hurricane landfalls on the US coast

1978  5  82.7  13.2
1978  6  95.1  41.9
1978  7  70.4  32.7
1978  8  58.1  17.4
1978  9 138.2  29.2
1978 10 125.1  27.3
1978 11  97.9  21.5


1981 May-November sunspot numbers, once again, very large spike near September

1981  5 127.5  33.7
1981  6  90.9  31.8
1981  7 143.8  41.1
1981  8 158.7  43.4
1981  9 167.3  27.8
1981 10 162.4  42.7
1981 11 137.5  54.1


2001  May-November sunspot numbers, and just like the years with a lack of US hurricane landfalls during the peak of the solar cycle noted above, very large spike in sunspots near September.

2001  5  96.6  21.1
2001  6 134.0  36.7
2001  7  81.8  24.9
2001  8 106.4  14.2
2001  9 150.7  27.7
2001 10 125.5  22.8
2001 11 106.5  25.0


Out of curiosity, I looked at the sunspot cycle numbers for April-November the last 4 years because of our lack of hurricane landfalls to see if the lower sunspot number for September in relation to the surrounding months, theory for US hurricane landfalls to sunspot cycles worked.

2012 sunspots April-November, it worked last year, although the drop in sunspot numbers was not as substantial as in the high correlation years like 1999 & 1938, but what does really grab my attention is how the lowest sunspot numbers are actually in October, and this correlates with hurricane Sandy making landfall on the US.

2012  4  55.2  28.6
2012  5  69.0  11.0
2012  6  64.5  33.3
2012  7  66.5  22.8
2012  8  63.0  26.0
2012  9  61.4  22.1
2012 10  53.3  13.8
2012 11  61.8  22.5

2011 sunspots April-November, notice rising sunspot numbers throughout the time period through October, correlates well with the overall large amount of storm recurvatures observed in 2011.

2011  4  54.4  13.1
2011  5  41.5  17.1
2011  6  37.0  21.3
2011  7  43.8  15.9
2011  8  50.6  22.2
2011  9  78.0  21.2
2011 10  88.0  23.4
2011 11  96.7  14.6


2010 sunspot numbers April-November, once again, consistently rising sunspot numbers throughout the hurricane season, corresponding to the lack of US hurricane landfalls

2010  4   8.0   8.4
2010  5   8.7   8.9
2010  6  13.6   7.8
2010  7  16.1   6.7
2010  8  19.6  11.9
2010  9  25.2  11.7
2010 10  23.5  13.7
2010 11  21.5  10.1

2009, an el nino year, thus one wouldn’t expect much in the way of tropical activity upon the US coast, although you should notice towards November a drop in sunspots which corresponded to Hurricane Ida, which at one point was a category 2 hurricane in the Caribbean in November, but luckily weakened before landfall on the northeastern Gulf coast

2009  4   0.8   2.3
2009  5   2.9   4.1
2009  6   2.9   4.3
2009  7   3.2   5.4
2009  8   0.0   0.0
2009  9   4.3   6.4
2009 10   4.8   7.6
2009 11   4.1   4.9

2008, on the other hand, a very active year for hurricane landfalls, with Dolly, Gustav, and Ike, and Hanna almost a hurricane, had a drop in sunspots towards September, in comparison to May-July & November surrounding time periods, thus this offers further evidence in supporting this sunspot cycle theory on US hurricane landfalls

2008  4   2.9   4.4
2008  5   3.2   4.9
2008  6   3.4   3.9
2008  7   0.8   2.5
2008  8   0.5   1.9
2008  9   1.1   2.8
2008 10   2.9   4.3
2008 11   4.1   4.9

2008 hurricane season


Ok, so we see that years with a large number of hurricanes on the US coast during the peak of the solar cycle have a large number of sunspots before the peak and following the peak of the hurricane season, where as years that had lower numbers of landfalls had large decreases in sunspot numbers in the beginning and end of the season, with rising sunspot numbers near the peak of the season. What physical mechanism could be at play to produce these kind of results?

Well, thinking about it, I understand that solar radiation, in particular, ultraviolet radiation interacts with ozone in the lower levels of the stratosphere, that lead to radiation from above in the stratosphere that helps to create the different temperature gradient in the stratosphere, compared to the troposphere where we live, where instead of temperature falling with increasing altitude, it actually get warmer with increasing altitude in the stratosphere. The ultraviolet radiation interacts with normal oxygen molecules and helps to form ozone, and this ozone is created in the tropics where the most solar radiation is received.


Diagrams showing the reaction of oxygen & ultraviolet radiation showing how it becomes ozone in the stratosphere, and even how it becomes destroyed.




If you really think about it, consider how the ozone helps to warm the stratosphere, in how it gets warmer with increasing altitude. Also, know that the greatest amount of solar radiation received is in the tropics, which based off of the Brewer-Dobson circulation is also where ozone is created, thus you should conclude that the greatest effects felt from changes in the amount of UV radiation received to earth in relation to the sunspot cycles will be over the tropics.



Let’s say, the sunspot cycles suddenly decrease like they did in the high hurricane landfalling years, in the middle of the summer towards September. What happens in this instance is if you see a large drop in solar radiation, in particular UV on the stratosphere, it leads to significant cooling in that layer of the atmosphere due to decreased amounts of UV radiation interacting with ozone to help radiate that atmospheric layer. Thus, considering that cooler substances tend to take up less space, decreased radiation from the sun in the middle of the summer will force the stratosphere to shrink. In doing so, this allows for the tropical troposphere to expand in order to compensate for the decreased space taken up by the stratosphere. Expanding substances are associated with warming, thus, the troposphere over the tropics should warm. If you recall back to the 400 millibar temperature relation to deep tropical activity, remember that warmer temperatures in the troposphere promote tropical cyclone development, because unlike mid-latitude cyclones that rely on differences in pressures and temperature (warm vs cold) to create energy (baroclinic processes), tropical systems use barotropic processes, relying solely on the latent heat release of water vapor into its liquid state to create energy, most of which is converted into heat energy, but some goes into creating wind, that helps to determine the strength of a given tropical cyclone. Thus, in general, a reduction in sunspot cycles in the middle of the hurricane season leads to a cooler stratosphere that allows for the tropical troposphere underneath to expand & warm, creating a more favorable environment for tropical cyclones.

A comparison of the highest correlating to highest non-correlating years for the sunspot cycle-US hurricane landfall relationship at the 400 millibar level for September reveals that at least over the tropical Atlantic, that the lower sunspots help to lead to comparably warmer temperatures at this level for the tropical Atlantic. What strikes me in this picture is the fact at how cool the Pacific is at the 400 millibar level, indicating limited tropical activity there, and notice also how the Indian Ocean is cooler than normal as well, and the western Pacific is slightly warmer than normal. Also, noticeable is a band of warmer than normal temperatures over the far north Atlantic with a band of cooler than normal temps in the mid-latitudes in the Atlantic (sounds like an Atlantic Tripole).


Unfortunately, this sounds awfully familiar to the current pattern of sea surface temperatures over the global tropics, definitely not a good sign of things to come.


Hurricanes and sunspot theory


Looking at the current sunspot numbers, we are experiencing a spike in sunspot numbers from April to May

2013  1  62.9  30.0
2013  2  38.0  13.6
2013  3  57.9  20.1
2013  4  72.4  12.8
2013  5  78.7  21.5

When I took a closer look at years that did not correlate with the sunspot cycle-US hurricane landfall relationship, I noticed a very astonishing trend, in that from the month of April to May sunspot numbers fell.

1908 April-May sunspot number

1908  4  57.6  32.6
1908  5  40.8  23.2

1957 Apr-May sunspot number

1957  4 175.2  41.4
1957  5 164.6  33.5

1978 Apr-May sunspot number

1978  4  99.7  18.3
1978  5  82.7  13.2

1981 Apr-May sunspot number

1981  4 156.4  45.2
1981  5 127.5  33.7

2001 Apr-May sunspot number

2001  4 107.7  33.9
2001  5  96.6  21.1

Now, look at this year once again, we are not falling, in fact we went up.

2013  4  72.4  12.8
2013  5  78.7  21.5

What could this mean? This means that given the data that we are not going to be like 1978, 2001, 1908, 1957, and 1978, which were dud hurricane seasons overall in terms of landfalls.

What’s worse is look at the years that were very destructive on the US coast and look at the Apr-May sunspots in those years

2005 (remember Dennis, Katrina, Rita, and Wilma), look at the Apr-May sunspots, they didn’t go down, they went up like this year.

2005  4  24.2   8.2
2005  5  42.7  17.2
How about 2004 with Charley, Frances, Ivan, and Jeanne? Once again, sunspots from April to May went up
2004  4  39.3  14.8
2004  5  41.5  15.9
How about 1999, with Irene, Floyd, and Dennis on the eastern seaboard?
Once again, a rise in sunspots from April to May
1999  4  63.7  18.0
1999  5 106.4  20.0
Even 1996 with Bertha & Fran on the eastern seaboard went up
1996  4   4.8   6.0
1996  5   5.5   6.4
1989 with hurricane Hugo on the eastern seaboard, sunspots from April-May
1989  4 130.6  25.0
1989  5 138.5  32.8
1979 (an analog year to this upcoming season) with David & Frederic on the US coast
1979  4 101.5  24.7
1979  5 134.4  30.5
1969, with the very historic Hurricane Camille on the central Gulf coast, once again a rise in sunspots from April to May
1969  4 106.8  30.6
1969  5 120.0  44.7
1960 & 1954, luckily did not correlate, a little relief, however I do note that the drop in sunspots from April to May is not comparable to the drops observed in the “Dud” hurricane years of 1978, 1908, 2001, and 1981
1960 Apr-May sunspots
1960  4 122.0  23.6
1960  5 119.6  19.5
1954 Apr-May sunspots
1954  4   1.8   3.8
1954  5   0.8   2.4
Thus, the lesson here is although the rise in sunspots correlates nicely with very destructive hurricane years on the US coast, and with the observed rise this year, this is not good news, however, correlation does not equal causation, thus, with years that observed drops but were still destructive on the US coast, there’s still potential that this correlation between April-May sunspots and the following year’s hurricane landfalls may not work. Given all of the other evidence at hand, if anything, this sunspot cycle connection to US hurricane landfalls just offers yet more support for my ideas for this to be a destructive year on the US coast, as our streak of years without a major hurricane landfall may come to an abrupt and violent end.
I’m going to leave you with this

Look at the sea surface temperatures in the top 10 hurricane landfalling years since 1950, you can see the Atlantic Tripole Signature, with a large region of warmer than normal SST extending from Africa through the Caribbean and into the Gulf of Mexico. What should really grab your attention though is the band of cooler than normal water around 20-25 degrees north that stretches from

Atlantic SST top 10 US landfall years


Now, look at this year, and compare this picture below to the one above and notice how similar they are, especially in that band of cool water, which seems to just about perfectly match the top 10 landfalling years in that it is centered north of the Greater Antilles. You can also see how the warmth stretches and is focused from Africa through the Caribbean and into the Gulf of Mexico and the warmth also farther to the north, indicative of the Atlantic Tripole signature.



Certainly a far cry from the top 10 lowest ACE (Accumulated Cyclone Energy) years, in fact, this year is the exact opposite, thus if you are looking for a quiet season this year, think again.

Atlantic SST 10 top low ACE years

Now I think you can see why I’m worried about this upcoming hurricane season, and I’m not just trying to hype things up here, I’m actually going back and showing you physical evidence for my reasoning and throughly explaining the conditions at hand.

In general, look for this hurricane season to be a very active one, and I’m not going to make any changes to my original hurricane forecast that I put out back on March 24th


13-18 named storms

7-11 hurricanes

3-5 major hurricanes

“Looking specifically at hurricane landfalls, the analog years featured, they were all high hurricane landfall years, except for 2010 with an average of 4.4 tropical storm or greater landfalls, 2.75 hurricane landfalls, and 1.2 major hurricane landfalls (that major hurricane landfall average is about double what is considered “normal”). Also these years featured 2 category 4 hurricane landfalls (Charley & Donna), and one category 5 hurricane landfall (Camille). Also, considering all of the similarities to 1969, it is very ironic given we are in the longest period known to go without a major hurricane on the US coast, now approaching 8 years since Hurricane Wilma’s landfall in south Florida as a category 3 hurricane in October 2005, and the analog years to this upcoming season are very active with hurricane and major hurricane landfalls (especially 1969’s category 5 Camille). Thus, my thinking is that this very historic period without a major hurricane landfall on the US coast is likely to come to an end soon, and could so quite violently, and unlike the last few years, I think this won’t just be an east coast hurricane year, as I expect the Gulf of Mexico to get involved as well.”

Hurricane Forecast


In saying this, I will make one change to my forecast. I’m going to now add the 1996 hurricane season to my analog package. As a matter of fact, way back when I was compiling my analog years for this season, 1996 was 6th place on my list, just missing the years I denoted as being in the top 5 (1969, 1960, 1979, 2004, and 2010) analogs for this year. It has become very hard to ignore all of the similarities this year shares to 1996, including the winter pattern.

Looking at the 500 millibar pattern, its not too far off, only difference is this past winter, the cold PDO signature with ridging in the North Pacific is stronger and there’s more blocking near Greenland.

1996 500 millibar pattern northern hemisphere Dec-Mar


This year Dec-Mar


US winter temperatures 1996



This past winter, although not as cold, certainly the general region affected by the chilly weather was the same, from the northern plains southeastward into the eastern US, some modification towards eastern Canada and New England



Also, like this year, despite the beginning of this year being warm, it was relatively cold late Dec & early January, much like 1996, although 1996 was colder further to the east.

1996 Dec 21-Jan 5 temps



This past winter Dec 21-Jan 5 temps



Also, remember near this time period in 1996, the Blizzard of 1996 struck, dropping historic snows in the eastern US, then what followed was a mid-winter thaw, much like what was observed this year.

Blizzard of 1996




This year too, had very large snowstorms, although they were centered much farther west as result of the core of the cold as being further to the west compared to 1996.

Take for example, the large midwest snowstorm that just passed to the northwest of Chicago in the week before Christmas, and led to all-time record snowfall in Madison, Wisconsin



Here are some more specific snowfall maps courtesy of the National Weather Service and weatherbell. You can really understand the scope as to how unpredictable snowfall predictions can be, with all of the factors at hand. Literally  within miles of a certain location you can see an increase or decrease in several inches of snow, and of course if you throw in the effect of sleet or freezing rain into the picture this makes the forecast even more difficult.


Then, the mid-winter thaw, which was observed in 1996, led to historic flooding in many areas of the northeast & the mid-atlantic

The town of Duncannon, a few miles north of Harrisburg, PA experienced major flooding issues from the snow melt.

The town of Duncannon, a few miles north of Harrisburg, PA experienced major flooding issues from the snow melt.

Here is a picture of the Harvey Taylor Bridge in Harrisburg in January 1996 along the Susquehanna River which experienced major flooding as result of the snow melt.

Here is a picture of the Harvey Taylor Bridge in Harrisburg in January 1996 along the Susquehanna River which experienced major flooding as result of the snow melt.


1996 winter temps Jan 15-25, notice the surge of warmth that went northeastward during this time period, leading to significant snow melt and flooding following the Blizzard of 96.



This past winter observed a very similar mid-winter thaw, of course which led to many going on to claim that “spring was here”, as a very large surge of warm temperatures engulfed the eastern US, pushing temperatures well into the 60s & 70s.



However, this didn’t last, and as I had warned back in December, that based upon the northern hemisphere snowfall anomaly for October that worst of winter was to come later, this was certainly the case as the rest of winter from February onward was quite cold from the plains eastward, with some residual warm air associated with blocking over eastern Canada residing in northern New England



1996 also was cold from February on through the rest of winter from the plains and points eastward



The spring pattern this year certainly was a lot closer to 1996

March 1996 temperature anomalies over the US



This past March’s US temperature anomalies



This past spring through the end of April was the coldest since 1996

Mar-Apr 2013 US temps



Mar-Apr 1996 US temps



Also, the sunspot cycle (extremities of the cycle, peaks & minimums) connection to US hurricane landfalls also seems to resonate with 1996, as it was towards the minimum following solar cycle 22.

1996 near minimum of solar cycle 22


Also, 1996 is within the overall cycle of warm AMO that began in 1995 and continues to the present day, I do like the fact at how it is towards the edge of the warm AMO, somewhat reminiscent of the early 1960s, which is a time period the current pattern appears to closely resemble.



Also, for the first five months of the year, 1996 has a Atlantic Tripole signature, much like this year, with warmth towards the Arctic and higher latitudes, cool in the mid-latitudes, and warmer than normal waters over the tropics.

1996 Jan-May Atlantic water temps


This year’s Jan-May Atlantic water temps, just like 1996, the Atlantic tripole signature is in place


The 1996 hurricane season also falls in line with the overall general pattern of warm AMO, cold PDO pattern that produces east coast hurricane landfalls (although the overall oceanic pattern in 1996 was somewhat different). The main reason why I didn’t include this analog in March during my hurricane forecast was the fact that there was a complete absence of tropical systems in the Gulf of Mexico, this is where I completely disagree with 1996, but all other things considered, it’s certainly a fairly good analog to this upcoming season.


Another interesting aspect about the similarities between this year and 1996 is how the track of Tropical Storm Josephine seems to resemble the recent Tropical Storm Andrea this year.


Tropical Storm Josephine (1996)



Tropical Storm Andrea (2013)



Speaking of hurricanes, if you recall in an earlier post the references I made to the 1938 hurricane season in that it had a similar pattern of coming off a previous hurricane season with limited tropical activity in the deep tropics, a series of rare, late season snows in Chicago that bear some similarity to this year’s late snows over the midwest, and the summer pattern that year near the east coast  (warm & wet towards New England, cool & wet to the southeast) seems to reflect my ideas for this summer. Well, looking at 1996, I am intrigued by the similarity in the track of hurricane Edouard to the Long Island Express of 1938, in that Edouard took an abnormal track of heading due north for several hundred miles, but luckily for the US coastline, Edouard veered out to sea at the last minute.

1996 hurricane Edouard




1938 Long Island Express



Even the temperature anomalies since the start of this June are reminiscent of 1996 in that there is a large surge of cool coming southeastward out of the northern Plains and into the eastern US, a pattern that has been persistent throughout much of the winter & spring.


June 1-17 1996 US temp anomalies



June 1-17 2013 US temp anomalies, although the cold is not as strong as 1996 for areas farther south & east, the overall pattern resemblance is still evident.




With all of this information considered about the close relationship 1996 has to this year, I’m going to be forced to put this hurricane season into my analog package, thus there are going to be some changes in the forecast, although they are only minor.

My 500 millibar height anomaly package for my original analog package of 1960, 1969, 1979, 2004, and 2010 looked like this, notice the ridge over the north Atlantic (helps to block storms from going out to sea), and the trough closer to the central US as part of a weaker “Texas Death Ridge”


With 1996, the 500 millibar height anomalies now look like this, noticeably, less ridging near the east coast, the ridge over the north Atlantic retreats eastward somewhat, but is much stronger, and the central US trough is slightly weaker but more spread out in nature. However, the tropics still very active, with a large region of generally low pressures through much of the deep tropics, indicative of higher than normal tropical cyclone activity there, the opposite of last season.


The summer temperature forecast, released back on April 21st, virtually no change, my analog package with 1996 included looks like this, main features being western warmth (especially warm towards California), slightly cooler towards the Pacific northwest, cool in the plains and into the southeast with some residual warmth towards parts of the northeastern US & New England.


My forecast from April 21st

Summer temp forecast

New analog package precipitation anomalies, virtually no change either as the new forecast still agrees with my summer precipitation forecast.

Analog package precip forecast, notably wet southeastern US & east coast, even wet into Texas, a sign of the weakening “death ridge”, slightly drier in comparison towards New England and the Great Lakes, & monsoon rains should help to bring at least conditions to near normal over the southwestern US, which is currently experiencing drought


My summer precipitation forecast from April 21st

Summer precip forecast


In addition, in seeing 1996 as a potential hurricane season analog, I have become concerned about a potentially long track Cape Verde system in July. This started with a look at the MJO in 1996

JFM MJO 1996



MJO April-June



Look at just how similar the April-June MJO looks to this year with very persistent movement into the Atlantic and especially the Indian Ocean (phases 2 & 3), with very little in the way of movement towards the western Pacific.



The constant MJO bombardment into the Indian Ocean has led to a dramatic cooling in surface waters, forcing the Indian Ocean Dipole, that was discussed earlier in this post, into its negative state, which promotes the African Monsoon & the monsoon surrounding the Arabian Sea, which helps to add more moisture and instability into African easterly waves, which account for around 90% of all tropical cyclones in the Atlantic. the persistent upward motion has helped to stir up the surface waters, much like when you blow on a bowl of hot soup, thus leading to the observed cooling.

Just look at the difference the MJO has made in just a few short months

Indian Ocean temps April 15th



Indian Ocean temps now



Notice though in the AMJ MJO composite the MJO diving into octants 2 & 3 by the end of June after being over the Maritime continent & western Pacific early-mid month


Now look at the current MJO forecasts, and look at the maroon line in the picture above, and you can easily see the resemblance between the two.


You can even see some resemblance in the overall MJO behavior in 2008, once again constant pushing into the Indian Ocean & Atlantic, very little in the way of action towards the western Pacific. Also, both 1996 & 2008 notably have the MJO diving into the Indian Ocean towards the end of June & beginning of July as many of the computer models show for the MJO in the coming weeks.

2008 April-June MJO2008_AMJ


NCEP GFS MJO forecast, as usual due to poor physics within the computer model, looks aggressive & way too progressive into the Atlantic, the MJO likely to be much slower & weaker in amplitude than is shown by the GFS below.



ECMF MJO much better, slower & less amplitude, yet, even by the ECMWF’s standards, this is quite an MJO pulse



ECMM MJO forecast quite reasonable in line with the ECMF


UKMET MJO forecast, usually like the GFS too fast & progressive with MJO, although usually not as bad as the GFS.


Based on the computer models, and analogs of 1996 & 2008, this is my general idea on the MJO, granted, we are talking about the MJO here, a still very misunderstood part of the weather, thus forecasting the MJO is a whole other animal compared to other weather phenomena.

In general, I agree with the ECMWF MJO forecast, yet, i am nowhere near as agressive as the GFS, which is obviously, as usual, way too fast, and the MJO forecast beyond 5 days with that model should be generally discarded. I think we should see at least some amplification into octants 8 & 1, and depending upon the model’s handling of an oncoming equatorial Kelvin Wave, this could help to slow down the MJO by enhancing convection, which has that kind of effect on the MJO. Also, with the Indian Ocean cooler, I am fairly cautious at this point in shoving the MJO deeply into octant 2 & 3 like my analogs, as there is some uncertainty as to whether in the longer ranges whether the MJO will continue to progress freely throughout the octants like it is known to do in neutral ENSO, or retreat back into the CDO (Circle of Death).



The MJO into octants 2 & 3 is very favorable for tropical cyclone formation over the Atlantic, in fact, look at this correlation

Octants 8 &1, although overall generally favorable, tend to focus more of the upward motion towards the eastern Pacific & Caribbean, southeastern US coast a place to look out, but the deep tropical Atlantic still relatively quiet until MJO gets closer to the Indian Ocean towards octants 2 & 3, octant 1 still not too bad though for tropical cyclone development.



Here’s all of the octants of the MJO, notice 2 & 3 are the development phases, octant 8 & 1 not too bad. Octant 4 & 5 where we were located usually means shut-down overall, although you do notice some upward motion ( in yellows & orange) towards the western Caribbean, which is exactly where recent tropical storm Barry formed. Octant 6 & 7 where we are now going into are very unfavorable, overall downward motion promoted, thus, the tropics given this information should remain quiet for at least the next week or so before we may have to start looking towards the Caribbean & southeastern US coast towards the very last few days of June.



Now, let me jog your memory on what occurred in 1996 & 2008 that had similar MJO behavior to this year. Both seasons featured long track, major Cape Verde hurricanes in the month of July, (quite a rarity based on climatology) both named Bertha, and both affected land. 1996’s Bertha hit eastern North Carolina as a category 2 & the 2008 Bertha hit Bermuda in its weakening cycle as a category 1 hurricane

1996 Bertha



Hurricane Bertha near its peak intensity as a category 3 hurricane as it passed to the north of the Virgin, northern Leeward Islands & Puerto Rico



2008 Bertha



Hurricane Bertha (2008) at its peak intensity as a low-end category 4 hurricane on July 7th.



A system as strong as either Bertha in July is quite rare and doesn’t occur all too often, however, I have seemed to find a correlation between the cycles of PDO/AMO and Cape Verde July tropical systems.

Now, I’ve done even more digging into this pattern, and here’s the storms (tropical storm or hurricanes) that have actually were long track Cape Verde storms, we’ll just say east of the Lesser Antilles because I realize that the Berthas of 1996 & 2008 were anomalies, and even formation that far east is quite an accomplishment this early in the season.
Hurricane Bertha (2008), TS Chris (barely in July), (2006) Hurricane Emily (2005), TS Alex (1998), Hurricane Bertha (1996), TS Arthur (1990), TS Cesar (1990), TS Barry (1989), Hurricane Dean (once again, barely in July) (1989), TS Claudette (1979), TS Anna (1969), TS Ella (1966), Hurricane Arlene (1963), Hurricane Anna (1961), (did form east of the Antilles, but quite anomalous in that it became a major hurricane during July), Hurricane Abby (1960), TS Two (1944), TS Three (1933), Hurricane Five (1933), Nassau Hurricane (1926), Hurricane Two (1926), TS One (1917), Hurricane Three (1916), Velasco Hurricane (1909), TS Two (1901), Hurricane Three (1901) Carrabelle Hurricane (1899), (The storm was initially detected as already a category 1 hurricane over the eastern Caribbean. Given the intuition that this region of the tropical Atlantic is naturally a dead zone because of higher trade winds enforced between the Columbia Low & Bermuda High that get squeezed in between South America & the Greater Antilles, promotes surface divergence and has a tendency to disrupt surface circulations, thus I suspect the actual formation of this storm was much farther to the east, likely east of the Lesser Antilles)

PDO & AMO since 1895

AMO flip to its cold mode near 1900 coincides with Carabelle Hurricane of 1899, TS Two of 1901 and Hurricane Three of 1901. By 1915, there was a very brief AMO spike into its warm cycle, the PDO turned cold, this correlates to TS One (1917) & Hurricane Three (1916). Around 1925, the AMO & PDO both flipped into their warm cycles, not surprisingly we observed The Nassau Hurricane in 1926 and Hurricane Two in 1926. Just before 1935, the PDO briefly went into its cold cycle, corresponds to TS in 1933 and Hurricane Five in the same season. Then, just before 1945, the PDO went into its cold cycle which lasted through the late 1970s, in 1944 near the PDO flip we had TS Two. In the late 1950s, near 1960, the PDO briefly flipped into its warm cycle, corresponds to Hurricane Abby & Anna in 1960 and 1961 respectively. In 1964, the AMO went into its cold cycle that lasted until 1995, but the AMO was rather slow to enter its cold phase in the mid-late 1960s, and as a result 3 storms ocurred in this period, Hurricane Arlene (1963), TS Ella (1966), TS Anna (1969). Then, the PDO flipped into its warm cycle in the late 1970s also known as the “Great Climate Shift”, during this period in 1979 TS Claudette occurred. Around 1990, there was a rather sudden, but relatively brief flip in the AMO into its warm cycle (more than likely the reason for the landfalls of Hurricanes Hugo (1989) & Bob (1991) during this time period), this led to several July Cape Verde storms in 1989 & 1990. Then, the AMO finally flipped completely into its warm cycle in 1995, around this time, Hurricane Bertha (1996) formed. After the “super el nino” of 1998, the Pacific went through several years of la nina, that led to a period of cold PDO in the very late 90s and into 2000, TS Alex occurred during this general time period in 1998. The PDO continued in its overall warm state through the mid 2000s, then flipped fully into its cold cycle around 2007. Around this flip was Hurricane Bertha (2008), TS Chris (2006), and Hurricane Emily (2005).


Given that the AMO appears to follow the solar cycles (TSI a very good indicator of the overall strength of the solar cycle),

And that the current solar cycle is anomalously low, I suspect that the AMO should soon tank into its cold cycle, thus, if this is indeed the case and a flip in the oceanic cycle flip is in order, then it makes a little more sense given other evidence and how favorable the conditions are at hand this year, to see a long track Cape Verde storm this July.



Look at the 1996 & 2008 Jun-September MJO, which give us a hint at where MJO may go overall later this season. Considering though, that conditions become naturally more favorable as we approach the peak of the season and the MJO is not as important in storm formation, as the environment and conditions at hand can help to sustain a tropical cyclone without much help from the MJO, it’s important to consider that aspect as well. However, after the large pulse in July, it appears that another large break in activity occurs until about mid August or so, when the MJO returns to help kick off the heart of the hurricane season, and in 1996, this helped to form Edouard & Fran, which both became large threats to the eastern seaboard of the US, and in 2008 this helped to jumpstart the season by giving birth to Tropical Storm Fay & Hurricane Gustav, which were later then followed by Hurricanes Hanna & Ike as the MJO continued to sit over octant 3, a favorable development phase for the Atlantic. I expect this year to follow an overall similar pattern, and uptick in activity towards the very end of June & into the first half of July, with at least a  3 week break or so, before the “meat” of the hurricane season arrives in August & September.

1996 June-September MJO



2008 June-September MJO



Given the similarities in the winter pattern this past year with 1996 (cold start, warm in the middle, cold ending lasting into spring), the cold spring pattern, Atlantic Tripole, sunspot cycle extremity, 400 millibar, etc.. I definitely think it is safe to say we are closely following 1996. Also seeing the MJO closely approach the western Pacific and seeing many of the computer models predicting a modiki el nino has me aware that we may see a major equatorial Kelvin Wave propagate eastward across the Pacific and even into the Atlantic in the relatively near future. I’ve also come across in some scientific readings that Kelvin Waves are possibly a factor to enhance convection associated with African easterly waves over the African continent. Taking into account how 92L almost became designated in the early stages of June and how the conditions appear to be favoring an active Cape Verde season, this information implies that we may in fact see the Cape Verde hurricane season come to life much sooner than normal, perhaps as soon as when this next upward MJO pulse enters the Atlantic. That would put the formation of any long track system in the very late stages of June or early July, about the time Hurricane Bertha formed in July 1996.
I was intrigued to see the latest CFS hint at a potential long-track Cape Verde hurricane in the early stages of July that seems to further support this notion for the first long track tropical cyclone of the hurricane season to come potentially in July. These images below show pressure spread, and want you to look towards the “Deep tropical Atlantic” between Africa & the Lesser Antilles for regions of lower pressure (shown in shades of yellow, orange, red, and brown), indicative of potential tropical cyclone activity or genesis. Areas I circled in black denote these regions of interest being shown by the CFSv2 model pressure spread


Tuesday’s run for July 20th



Monday’s run for July 19th



Sunday’s run for July 13th



Saturday’s run for July 17th



Friday’s run for July 17th about the exact same time frame as the systems being shown as in the other runs it hints at a long track tropical system, this time a very large tropical wave emerging off Africa, at least by this model’s standards is shown.13071712_1400


The run from Wednesday just continues to show relatively high activity in a general sense over the eastern Atlantic, at least in this run, it shows a series of rather strong series of tropical waves emerging from Africa, although the model does not do anything with the disturbances, (model runs shouldn’t be taken literally anyways), it is an indication that the model is “seeing” the MJO pulse coming across the tropics, and with the other conditions at hand & relations made to 1996 & 2008, this implies that the environment will become favorable for tropical cyclone genesis in the deep tropical Atlantic as we get past the first week of July or so.


July 12th CFSv2



July 16th



July 21st



The newest run now makes it at least the 7th in a row that this model shows development towards the eastern deep tropical Atlantic after July 10, which agrees with my forecast map at the bottom of this post. Only this time, this model is very aggressive and shows 2 systems coming off Africa and making runs at the east coast, although this can’t be taken literally, this shows that the environment in place, despite climatology, will be favorable towards long track Cape Verde storms.

Notice the timing hasn’t really changed from a week ago, that’s very good model consistency.

July 10 powerful African wave moving off the coast as the upward MJO likely is approaching by this time



By July 14, it shows conditions continuing to grow more favorable, a series of waves in the pressure spreads from the CFSv2 indicative of how over time, compared to previous model runs, this model is becoming more confident in some type of long track Cape Verde storm in the earlier stages of July.



July 20, the CFS implying that a potential hurricane would be near the southeast coast of the US, with more activity coming off Africa.



I really like the model’s handling though of activity towards the end of July & going into August, notice how conditions quiet down once again, this in line with the upward MJO pulse moving away by this time.

July 29th



MJO forecast ECMWF


In general, there are several key points to take away from this post.

#1-The error in progression of the MJO by the ECMWF in the Maritime Continent region can be linked to standing mountain waves, among other things, which relate strongly to stratospheric warming events & the Brewer-Dobson Circulation


#2 These warming events bring about (at least to me) a reminder of patterns in similar cycles of PDO in the 1950s & early 60s & by looking into a post from Andrew of the Weathercentre, I was able to determine the analogs he mentioned had a central Pacific el nino (modiki), which is much like what’s being forecasted in the computer models given the MJO progression & how the Pacific will have trouble warming the already cold east equatorial waters.


#3 Earlier reasons as to why I expected this hurricane season to be active & particularly dangerous, like the +NAO, 500 millibar pattern, warm 400 millibar theory, and the Atlantic Tripole are all coming to pass.


#4 A new oceanic oscillation I learned recently, the Indian Ocean Dipole, has connections to the hurricane season in affecting activity in the deep tropics & even implications on how next winter will start. Given this oscillation is in its negative phase, this promotes tropical development in the MDR & hints at a cold & snowy start to next winter, especially east of the Rockies.


#5 Specification in my sunspot cycle theory was needed, and in discovering the May-July, November & September sunspot cycle relationship to US hurricane landfalls & comparing those years to April & May sunspots offers more hints at a potentially destructive & violent hurricane season at hand.


#6 I reviewed also my hurricane analogs, and in looking once again at the conditions at hand, I deemed it was necessary given all of the similarities, that I include 1996 as part of my analog package, however, this hardly changes my original hurricane season ideas from March.


#7 Seeing 1996 as analog and the highly anomalous Hurricane Bertha in July made me connect this to a potential upward MJO pulse, and by also adding in 2008, finding supoort in the AMO/PDO cycles as well as looking to the CFS super long range computer model for further support, this implies for a potential long track Cape Verde storm in July.

This my thinking on how I think tropical activity may play out the next several weeks, notice how the timeframe moves up in time as you progress from east to west and confidence decreases from east-west as there remains some uncertainty in timing & the MJO as we get deeper into July.

Weather Advance hurricane pic 3


I hope you all enjoyed reading this, and for future posts, as I will try to keep them much shorter (we all know how that usually works out, lol), and I expect this to be the case especially as we get deeper into hurricane season & more immediate posting & storm analysis is required.