Will there be a tropical cyclone that forms in the Atlantic and travels all the way to the West Pacific?

[Team Ghost]

There have been tropical cyclones that crossed from the Atlantic to the East Pacific; in recent memory, Hurricane Bonnie did just that. There have been tropical cyclones that crossed from the East Pacific to the West Pacific, which Hurricane Dora had done just 2 years ago. Theoretically, there is nothing preventing a tropical wave from developing into a tropical cyclone in the North Atlantic, crossing through Central America into the East Pacific, and continuing to travel westward to the West Pacific; yet no tropical cyclone on record has done so.

[Hurricanehink]

Eventually, yes. Let's look at the two main factors: the crossover storms from Atl to EPAC, and the storms going from EPAC to WPAC.

There have only been 16 storms that officially crossed from the Atlantic to the EPAC in 149 years. But before the satellite era, it would likely be difficult to track them continuously. Restricting it to the satellite era (1966) leaves 13 in 59 years, or every 4.5 years. But even then, several of the crossover storms quickly dissipated upon hitting the EPAC. If we limit it to storms that re-intensified after hitting the EPAC, that leaves only 6 storms in 59 years, or every 9 years. 6 divided by 59 is 0.10169

During the same satellite era, there have been 8 EPAC storms that made it to the WPAC, or once every 7 years. That equals roughly 0.13559

Combining these probabilities gives us 0.013788, or once every 73 years. That doesn't mean we're overdue for one, but statistically, any given year is very unlikely for this to happen. There are a number of independent factors that are needed, such as a strong enough ridge to keep a storm south enough in the EPAC to not succumb to dry air. But it's possible.

[Cyclenall]

This reminds me of a muse I had years ago as to what the return rate of a TC that started near Cape Verde in NALT, and then circumnavigated the globe to be intact on the other side of Africa in the northern Indian basin. I think we possess the data to make a statistical probability of it including how the ridge configuration would be along its journey.

[Teban54]

IMO, the most plausible scenario is "Bonnie-Dora":

1. In July or early August, an Atlantic storm forms in or enters the SW Caribbean, crossing through Nicaragua or Costa Rica into EPAC.
2. Ridging is strong enough in EPAC and CPAC to prevent it from gaining much latitude. Chances are, either the storm becomes a classic EPAC Cat 4 and "pumped the ridge", or it stays weak or even dissipates before reforming.
3. The storm passes south of Hawaii and continues west into WPAC.

This is exactly what Bonnie 2022 and Dora 2023 did collectively. Their tracks even intersect with each other!



Is this easy? No, far from it -- and in fact, simply multiplying the number of ATL-EPAC and EPAC-WPAC frequencies doesn't paint the full picture. If anything, conditions that promote one type of crossover may cause the other to be less likely!

Time of the year:

• The EPAC-WPAC crossover basically requires July and August, when CPAC is warm enough and ridging is strong enough. In fact, all 8 known EPAC-WPAC crossover storms formed in these two months.
• But having a Caribbean storm in the early season is already rare -- and one in the SW Caribbean is even rarer! Most July and August storms that survive the Eastern Caribbean graveyard end up in the NW Caribbean, not SW. (This includes early MDR/Caribbean storms like Elsa 2021 and Beryl 2024.)
• SW Caribbean actually sees more landfalls in October than the earlier months: 10 out of 16 ATL-EPAC crossovers formed in September or later. But by then, the Pacific has ceased to cooperate.

Longitude in EPAC:

• Most ATL-EPAC crossovers didn't go deep into EPAC: many of them dissipate very soon after emerging from Central America. Only 5 out of 16 storms made it past 110 W (south of the Gulf of California). Only two made it west of 130 W: Bonnie 2022 and Cesar-Douglas 1996. They're also the only two ATL-EPAC crossovers in July.
• On the other hand, most EPAC-WPAC crossovers didn't even form until west of 110 W! Only three did: Dora 2023, Dora 1999, and John 1994. Of these, John has the easternmost formation at 97.1W, which is still almost 10 degrees from Nicaragua.

ENSO and other conditions:

• EPAC-WPAC crossovers heavily favor +ENSO years: of the 7 years with such activity, only one didn't have an El Nino by the end of the year (Dora 1999). But such years often shut down the Atlantic, especially the Caribbean. +ENSO years are also often correlated with weaker ridging (see 2023), making waves less likely to enter the Caribbean in the first place.
• Likewise, an early-season Caribbean storm relies not just on -ENSO, but favorable intraseasonal forcing passing through the Caribbean. Both can make conditions harsher in the Pacific.

Storm structure: For an Atlantic storm to survive the land interaction (even if just the narrow strip of Nicaragua), it likely needs a solid core. But this implies a much better organized system emerging on the Pacific side compared to the waves typical in this region (even if its wind speed has been knocked down), which would then be more likely to intensify and lift north.


Here's a map of all 8 EPAC-WPAC crossovers (including regenerations), and the 5 ATL-EPAC crossovers that survived past 110 W (Bonnie, Cesar-Douglas, Joan-Miriam, Gert 1993, Irene-Olivia):



Observe that

• 4 out of 5 selected ATL-EPAC crossovers did so through Nicaragua.
• 4 out of 5 ATL-EPAC crossovers made landfall as hurricanes. The weakest landfall was Bonnie, which was also quickly developing a core just before landfall despite being only at 50 kt.
• Dora 2023's continued WSW motion through the Pacific is very atypical, even among the successful EPAC-WPAC crossover storms. This means while three ATL-EPAC crossover storms passed immediately next to where Dora formed (14.8N, 103.7W), they would need very anomalous southward motion to be on Dora's track. But in reality, the 3 storms were all hurricanes moving WNW, not WSW.
• Aside from Dora 2023, all other EPAC-WPAC crossover storms formed south of 13N (including the other eastern formations, Dora 1999 and John). Nicaragua is at 12N, so such minimal latitude gain seems very difficult.
• Joan-Miriam 1988 is the only ATL-EPAC storm that lost substantial latitude to ~12N, conducive for Pacific long-trackers. But it occurred at the wrong time in October.

With an infinite time horizon, I do think that the stars would align for the conditions that resulted in Bonnie 2022 and Dora 2023 to align. But given their contradictory nature, they probably have a much longer return period than the mathematical statement of "once every 73 years", which assumes independence.

[Category5Kaiju]

In short, it's very hard to have such sprawling, persistent ridging that can steer a storm from the Caribbean all the way to the WPAC without recurving early and hitting land/fizzling out over sub-optimal water temperatures.

First, you need an Atlantic storm that survives the crossover from the Caribbean Sea to the EPAC (let alone a storm that doesn't get lifted to the north beforehand or doesn't really get going due to proximity to South America). This is rare alone, as storms normally greatly and rapidly weaken over Nicaragua and surrounding areas. It would take a very robust circulation and/or very fast movement to survive a crossover over Central America.

An EPAC storm ending up in the WPAC is a bit more common, but the major limiting factor in this entire situation is the Atlantic to EPAC crossover.

[Ptarmigan]

Probably it has happened in the past before satellite. Could it happen? Possibly.

[kevin]

Eventually, yes. Let's look at the two main factors: the crossover storms from Atl to EPAC, and the storms going from EPAC to WPAC.

There have only been 16 storms that officially crossed from the Atlantic to the EPAC in 149 years. But before the satellite era, it would likely be difficult to track them continuously. Restricting it to the satellite era (1966) leaves 13 in 59 years, or every 4.5 years. But even then, several of the crossover storms quickly dissipated upon hitting the EPAC. If we limit it to storms that re-intensified after hitting the EPAC, that leaves only 6 storms in 59 years, or every 9 years. 6 divided by 59 is 0.10169

During the same satellite era, there have been 8 EPAC storms that made it to the WPAC, or once every 7 years. That equals roughly 0.13559

Combining these probabilities gives us 0.013788, or once every 73 years. That doesn't mean we're overdue for one, but statistically, any given year is very unlikely for this to happen. There are a number of independent factors that are needed, such as a strong enough ridge to keep a storm south enough in the EPAC to not succumb to dry air. But it's possible.

In my opinion this calculation gives too high a chance to such crossover storms since you're calculating what the chance is (assuming no correlation) that the same season has both a ATL -> EPAC crossover as well as a WPAC -> EPAC crossover. However, these could also be different storms. It's basically the probability of Bonnie 2022 and Dora 2023 happening in the same season, but not the possibility that those two crossings correspond to the same TC. And if they are randomly selected, which is what we're assuming now, there is a very big chance that they will be different storms. The chance that the season has both crossovers as well as the requirement that both crossovers correspond to the same TC is much smaller.

Assuming a storm crosses over from the Atlantic -> EPAC (~10% chance per season), what is the chance that this specific storm also crosses over into the WPAC, assuming no correlation? I won't consider a season with multiple crossings in a single year since those probabilities are even more negligible. Roughly 15 TCs/season form in the EPAC, so about 885 in those 59 years of the satellite era. Out of all those TCs, 8 have crossed over into the WPAC. So the chance of a random EPAC storm crossing over is about 0.90%. Combining those chances gives a combined chance of 10%*0.90% = 0.09% of a random season having a Atlantic -> EPAC crossover which then also crosses over into the WPAC. Or a return period of ~1000 years. Of course this assumes that there is no correlation between a storm crossing over from the Atlantic -> EPAC and then also from the EPAC -> WPAC. But I can imagine that the condition that allow crossovers would be linked at least somewhat. But either way, I expect such a track would be exceedingly rare not just based on these statistics but also based on metereological theory since a very special background state would be required to sustain a storm across this entire track.

[Teban54]

Eventually, yes. Let's look at the two main factors: the crossover storms from Atl to EPAC, and the storms going from EPAC to WPAC.

There have only been 16 storms that officially crossed from the Atlantic to the EPAC in 149 years. But before the satellite era, it would likely be difficult to track them continuously. Restricting it to the satellite era (1966) leaves 13 in 59 years, or every 4.5 years. But even then, several of the crossover storms quickly dissipated upon hitting the EPAC. If we limit it to storms that re-intensified after hitting the EPAC, that leaves only 6 storms in 59 years, or every 9 years. 6 divided by 59 is 0.10169

During the same satellite era, there have been 8 EPAC storms that made it to the WPAC, or once every 7 years. That equals roughly 0.13559

Combining these probabilities gives us 0.013788, or once every 73 years. That doesn't mean we're overdue for one, but statistically, any given year is very unlikely for this to happen. There are a number of independent factors that are needed, such as a strong enough ridge to keep a storm south enough in the EPAC to not succumb to dry air. But it's possible.

In my opinion this calculation gives too high a chance to such crossover storms since you're calculating what the chance is (assuming no correlation) that the same season has both a ATL -> EPAC crossover as well as a WPAC -> EPAC crossover. However, these could also be different storms. It's basically the probability of Bonnie 2022 and Dora 2023 happening in the same season, but not the possibility that those two crossings correspond to the same TC. And if they are randomly selected, which is what we're assuming now, there is a very big chance that they will be different storms. The chance that the season has both crossovers as well as the requirement that both crossovers correspond to the same TC is much smaller.

Assuming a storm crosses over from the Atlantic -> EPAC (~10% chance per season), what is the chance that this specific storm also crosses over into the WPAC, assuming no correlation? I won't consider a season with multiple crossings in a single year since those probabilities are even more negligible. Roughly 15 TCs/season form in the EPAC, so about 885 in those 59 years of the satellite era. Out of all those TCs, 8 have crossed over into the WPAC. So the chance of a random EPAC storm crossing over is about 0.90%. Combining those chances gives a combined chance of 10%*0.90% = 0.09% of a random season having a Atlantic -> EPAC crossover which then also crosses over into the WPAC. Or a return period of ~1000 years. Of course this assumes that there is no correlation between a storm crossing over from the Atlantic -> EPAC and then also from the EPAC -> WPAC. But I can imagine that the condition that allow crossovers would be linked at least somewhat. But either way, I expect such a track would be exceedingly rare not just based on these statistics but also based on metereological theory since a very special background state would be required to sustain a storm across this entire track.

Additionally, as I elaborated above, even the chance of two types of crossovers (ATL-EPAC and EPAC-WPAC) occurring in the same year, let alone the same month -- regardless of whether they're the same storm -- is likely lower than the two independent chances multiplied.

EPAC-WPAC crossovers heavily favor July and August in years with ongoing or developing El Ninos. On the other hand, ATL-EPAC crossovers are more common during late seasons in -ENSO years, as they typically occur in the Caribbean.

[Yellow Evan]

Dora 23 was arguably classifiable as a TC in the MDR.

[ljmac75]

How would the odds compare with that of a tropical cyclone forming in the West Pacific and traveling east all the way to the Atlantic?

[Category5Kaiju]

How would the odds compare with that of a tropical cyclone forming in the West Pacific and traveling east all the way to the Atlantic?

Although highly highly unlikely in any given year, an Atlantic to WPAC storm is much more likely than a WPAC to Atlantic storm. Going from the WPAC to Atlantic would require the system to basically go against the default steering flow aloft of east to west in the tropical EPAC. Even if the storm were to recurve in the WPAC and track a bit more north, it's going to hit the cold waters of the North Pacific and really struggle to get going there. There's a reason why it's rare for hurricanes to happen in the North Pacific Ocean, east of the International Date Line. And then, of course, there's crossing the North American continent that this hypothetical system would have to worry about. The Rocky Mountains won't be so kind.