Wednesday, March 27, 2013

2013 Preliminary Atlantic Hurricane Outlook

Hello everyone, and welcome to The Weather Centre's Preliminary Atlantic Hurricane Outlook. In this outlook, because it is just the preliminary outlook, we will only examine my current analog package and a batch of ENSO model forecasts.

We start now with the analog package. I utilized the Pacific-Decadal Oscillation (PDO) and Atlantic Multidecadal Oscillation (AMO) for this outlook's analog years. The PDO involves a positive and negative phase, where each phase depends on the prominent water temperature anomaly in the Northeast Pacific. A glance at the sea surface temperature chart for the northeast Pacific reveals a swath of above normal water temperatures in the offshore regions of the Gulf of Alaska, with below normal temperatures immediately offshore of North America and Alaska. This is the typical signature of the negative PDO. As for the Atlantic Multidecadal Oscillation, we also see a positive and negative phase with this index. The positive AMO signifies warmer than normal water temperatures across the waters off Greenland and in the far north Atlantic Ocean. In the same sense, the negative AMO allows below normal water temperatures to encompass the North Atlantic waters.

After choosing specific years from these two indices, I combined my chosen years and used the ones that had both the PDO and AMO in phases similar to what we are currently anticipating. As a result, I came up with the years 1951, 1952, 1955, 1956, 1999, 2000, 2008, and 2011. All of these years had a clear negative Pacific-Decadal Oscillation and positive Atlantic Multidecadal Oscillation, which is what we are experiencing now and what we should experience moving forward into the spring and summer months.

Let's take a look at an archive of hurricane tracks during the aforementioned years to see if we can see a trend in my analog years that could assist in helping us find a common track for the upcoming season.

In 1951, we saw a storm season that had many storms going out to sea rather than towards the US Mainland. At least 3 of those storms didn't even have a chance of making it to the United States. However, there were a few storms that did have a close brush with the Carolinas, as well as one landfall in Florida. A couple storms did make a threatening track towards the Gulf, but for one reason or another they did not hit the US. One storm (Hurricane Charlie) did hit Mexico, the Yucatan Peninsula, and Jamacia. All in all, the trend for this year was a slight Gulf threat with multiple close calls on the East Coast.

1952 was not unlike 1951 in terms of where tropical systems went. While the Gulf of Mexico was considerably quieter with no systems actually in the heart of the Gulf, there was one landfallig system in the Southeast US, another landfall in Florida, and at least 3 storms that ended up recurving out to sea. The system that struck Florida, which formed on Groundhog Day (yes, in February), was never named. Hurricane Able was the system that hit the Southeast. The trend in this season was a threat to the East Coast and a rather quiet Gulf of Mexico.

1955's Atlantic Hurricane season involved a pretty darn active season in terms of landfalling tropical systems. We saw multiple impacts on the Gulf Coast, especially in Louisiana and Mississippi. We also saw possibly more than 2 separate landfalls on the East Coast, all of which struck the Mid-Atlantic (one also sideswiped the Northeast). Tropical Storm Brenda was one of the systems that hit the Gulf Coast, and it was Hurricane Connie that slammed into the East Coast. Tropical Storm Five followed Brenda into the Gulf, while Hurricane Diane took a hint from Connie and sped towards the East Coast to make landfall. The clear trend here is an active landfalling season in the Gulf and East Coast.

1956 continued the idea of storms running dangerously close to the US Mainland, even making landfall a couple points along the way. We saw a rather unusual season in that the majority of the storms took almost a due-north path out to sea. Usually, storms will form off the coast of Africa and curve west before doing a 180 and going out to sea. In this season, the storms just formed and made their intentions clear as day. Tropical Storm One, the first named system of the season, impacted the Gulf with a landfall. Both Hurricane Anna and Tropical Storm Dora hit Mexico, while Hurricane Flossy did a double-take by hitting the Yucatan Peninsula and the Gulf Coast. An unnamed subtropical cyclone hit Florida and skirted into the Mid-Atlantic. The trend in this season is clear: Tropical cyclone threats were highest in the East Coast and Gulf Coast.

1999 was no different than previous years. We saw multiple tropical cyclones skirt immediately offshore of the Eastern Seaboard, and the Gulf of Mexico came under fire. We saw extreme southern Texas take a landfall from Hurricane Bret, while Florida was surrounded by several near-misses and at least two landfalling systems. It was Hurricane Dennis that took one of the most awkward tracks I have ever seen a tropical cyclone take. It began to curve out to sea after narrowly missing eastern Florida, but suddenly make a hard turn south, then a hard turn west before making landfall in the Mid-Atlantic as a weak tropical cyclone. The trend continues; East Coast and Gulf Coast were threatened in this season.

The year 2000 was a bit different in that the number of threats to the East Coast was reduced. We saw more of a tendency for storms to curve out to sea earlier than storms in, say, 1999. The state of Florida did bear the brunt of at least one landfall (courtesy of Hurricane Gordon), while Alabama and Georgia were affected by a tropical cyclone known as Tropical Storm Helene. A tropical depression also made landfall in the western Gulf Coast, although it is not shown in the map above. The Gulf Coast definitely made headlines as the most affected area this season, with the East Coast in a not-so-close second.

2008 brought an absolutely chaotic Atlantic hurricane season. We saw over half a dozen tropical cyclones hit the Gulf Coast, with one cyclone hitting the East Coast. Tropical Storm Cristobal affected the East Coast, while Hurricane Dolly and Tropical Storm Edouard hit the Gulf Coast. Tropical Storm Fay zig-zagged through Florida, while Hurricane Gustav ravaged the Gulf. Hurricane Hanna hit the Eastern Seaboard, and it was Hurricane Ike that dominated the Texas coast. The trend is probably the clearest we have seen in all of the analog years: The Gulf experienced the largest threat, with the East Coast also getting in the action.

My last analog year, 2011, echoed previous analog years, but also raised the recurring theme of storms curving out to sea. We saw the heavy majority of tropical cyclones curving out to sea in 2011, with little to no damage coming between them. However, there were two Gulf Coast landfalls that were weak, as well as one Mid-Atlantic landfall. The final analog year continued the idea that the US mainland was threatened.

The next piece of evidence I want to show you is a chart of the latest observed sea surface temperature anomalies over the Atlantic Ocean. I outlined three regions of anomalies- an above-normal sea surface temperature (SST) area in and just south of Greenland, a below normal SST area to the east-northeast of the Mid-Atlantic coastline, and a large swath of above normal SST values from the coast of Africa and west into the Caribbean. This type of arrangement of SST anomalies is called an Atlantic Tripole. Years that had an Atlantic Tripole in place had a positive correlation with the monthly hurricane total from June to July and Accumulated Cyclone Energy (ACE) values. This positive correlation means that, since the Atlantic Tripole is in its positive phase right now, we could reasonably expect more tropical cyclones than normal to form this season, and those cyclones could have more energy than normal, which would then raise the ACE index.

I am monitoring the El Nino-Southern Oscillation at this time, and current sea surface temperatures suggest we are in an El Nino. I don't really want to touch on that subject, as model forecasts have the Nino fighting an uphill battle as far as being able to survive during the hurricane season. Until the model forecasts clear up, I don't have enough confidence to lean one way or another on the presence or lack of an El Nino.

Based on my current analog package, combined with the Atlantic Tripole and a few other factors that greatly contribute to the Atlantic hurricane season, I believe that the threat to the Gulf Coast is 'High'. The 'High' threat means that the risk of a tropical cyclone of ANY strength hitting the Gulf Coast is pretty significant. I cannot go in-depth into strength of the cyclone until my official outlook (which will be out in Late April or Early May), but I cannot rule out the potential of a hurricane hitting the Gulf. As for the threat to the East Coast, I have classified the threat as 'Medium-High', meaning that there is a pretty decent chance of a tropical cyclone of any strength affecting the East Coast. Based on the analog package, the threat of a tropical system hitting that region is pretty solid.

If you have any general questions (no specific timing, strength or location questions, please), don't hesitate to ask. I will try my best to get back to you ASAP.



Mark said...

Any thoughts about Atlanta, GA?

Andrew said...

Mark: At the end of the post I asked for no location questions- this is only a preliminary and general outlook.

Eric said...

Andrew, I like the outlook and it was fairly detailed, although I do think something that will need to be taken very deep consideration in following outlooks and is something I outlined is the 400 millibar temperatures over the deep tropics, they are a much better indicator, at least from what I can tell, compared to using ENSO because even if you do see el nino, with the air aloft warmer than normal in the deep tropics, it promotes thunderstorm growth as tropical cyclones, unlike mid-latitude systems, feed off of pure latent heat release, thus when you have cooler 400 millibar temperatures like last year, even despite la nina, activity is very quiet in the deep tropics, and the same concept can be applied to 2004 (an analog year in my forecast) when weak el nino conditions were in place, yet deep tropical activity was very prominent because 400 mb temps were warmer than normal that season. Just food for thought

Anonymous said...

Do you think the tropical systems will be brought INTO the eastern seaboard, or will the just skim the coast like Hurricane Earl? I didn't know if this was a comment/question I could ask but please answer!

Andrew said...

Eric: My head was spinning from all the info I must have missed it. Good connection, there is definitely a similarity between 2004 and the current year as far as 400mb temps.

Anonymous at 4:16: Good question. My analogs suggest the storms would be able to do both, but the majority of them would most likely be offshore of the mainland.

Eric said...

Andrew, that is completely ok, lol, it may be easier just to read that post one bit at a time, or read it again, because there truly is a lot of information you can derive from it, but you had a good hurricane season forecast above anyways, although I do not see any mention of 1969. Although it was in the overall state of cold AMO, cold PDO, I made very clear recognition in my post that in 1969, which is my top analog for this year, that the AMO briefly spiked to warm, which would at least temporarily, bring a similar 1950s-like pattern back into place in that time period, which draws yet another connection to that hurricane season, but we'll just have to wait and see how that analog holds up.

Anonymous said...

Any idea at what you think SST's will be in the Gulf for hurricane season? Will they be below average or above average do you think?

Eric said...

Andrew, something I said on, that may interest you, just appears in my mind to be just another comparison that can be made to 1969, among all of the vast other similarities, including the relation to the Lindsay Storm and early February snowstorms, which occurred virtually within hours of each other, just 44 years apart from each other. Also, the overall northern hemisphere 500 millibar pattern in 1969 from the Dec-Feb period is also similar to what has been observed, with a prominent Greenland block and a large trough over Europe and the US, with some ridging over the north Pacific, classic in a cold PDO, with a trough that hangs back underneath the ridge all the way to Hawaii. 1969 although was not in the overall period of warm AMO, cold PDO as the AMO had cooled in the early-mid 1960s, there was a notable spike in the AMO briefly to its warm mode in 1969, and with a cold PDO in place, this would just continue to enforce the 1950s-like pattern of warm AMO, cold PDO. Also 1969 was at the peak of the solar cycle, like this year, and the solar cycles of both what is being observed now and in 1969 are even more similar as both were at the peak of relatively flat and weak cycles in comparison to the surrounding solar cycles. Then, we just continue to go further into 1969 and we see that in the pre-season leading up to the hurricane season, the temperature profile distribution over the Atlantic is very similar to what is being observed now with a classic warm AMO signature in place, with warmer than normal waters over the deep tropics, cooler waters in the middle latitudes and warmer waters again towards the arctic. WIth such a distribution of water temperatures with cooler than normal waters over the mid-latitudes and warmer than normal waters over the arctic, this helps to enforce and create a classic +NAO signature in the summer, as the cooler than normal waters present from the eastern coast of North America to the shores of the European coast help to enforce the Azores high naturally prominent in this region during hurricane season, and with a stronger than normal Azores high, this forces the natural pressure gradient between the Azores high and the Icelandic low to increase, helping to tighten and confine the jet stream towards the arctic and strengthen the Icelandic vortex.

Eric said...

As noted in the hurricane seasons since 1950 with the highest number of hurricane landfalls on the US, there was a very prominent icelandic low pressure and general low pressure in place near Greenland, which would help to entice the landfalling pattern more so than the Pacific as it is more directly related to the high pressure block usually found over eastern and maritime Canada during high landfalling years. This high pressure block over eastern and maritime Canada, which was also found in hurricane Sandy helps to force storms that are over the western and southwestern Atlantic inward towards the coast as they block the natural northeasterly progression of tropical cyclones from the deep tropics and into the mid-latitudes that naturally occurs with the shifting of the wind belt from the tropical easterlies to the mid-latitude westerlies under the Hadley and Ferrell atmospheric cells. The region of high pressure also plays a role in that since air spreads outwards in all directions in regions of high pressure, this forces convergence and pressure falls over the SW Atlantic, which only helps to intensify tropical cyclones in this area by enticing rising motion which aids in the development of thunderstorm activity which is the basis and acts as the building blocks for tropical cyclones. Another aspect I noted that gave even further relation to 1969 was the 400 millibar temperatures, when you complied a list of analog years to this current set-up and stacked those up against 1969, it is absolutely stunning the profound similarity between the two. I could go on and on about the absolutely stunning relationship between this year and 1969, and with this temperature pattern in Europe, which seems to suggest that perhaps 1969 was a year, that was at least close where in the temperature during the month of March was colder than the month of January, which was also below normal, thus this may be one of the years you have been searching for to find this interesting temperature relationship. It also concerns me at the fact that 1969, clearly without a doubt, the closest analog year, featured the deadly and very destructive hurricane Camille as category 5 onto the Gulf of Mexico coastline, thus seems to suggest to me that the streak for no hurricane landfalls on the US coast could come to an abrupt and violent end. Also, let's remember that 1969 also had a significant east coast hurricane with Gerda that moved into eastern Maine as a category 2, one of the few hurricanes to do ever do so. Plus 1969, like what is being expected this year, 1969 came off of the 1968 hurricane season, which like last year, was active (at least in respect to the cold AMO pattern of the 1960s), but many of the storms in 1968 formed out of the deep tropics like they did last season and farther to the north and west, where also the highest ACE was found last season, and 1968 somewhat like last year had at least one significant storm to threaten the east coast (Abby & Gladys). Also, the ENSO index in 1969 is very similar to what is being observed currently and what I anticipate for this hurricane season with either ENSO neutral or weak el nino conditions, which even though usually means less tropical activity for the Atlantic, due to the overwhelmingly warmer than normal waters over the deep tropical Atlantic, means that the majority of tropical energy and rising motion will be focused towards the tropical Atlantic anyway. Thus, with all of this information considered, I honestly hope you have 1969 in your analog package, because if you don't, I think you're making a VERY BIG mistake, but that's just my two cents on this.