There may be ambient howling throughout parts of the Northeast this upcoming Halloweekend, but those ghastly wails won’t be coming from werewolves. For many with Halloween plans, the sound could be even more harrowing. It could be galling winds from an intensifying Nor’Easter, the first of the season. The most bone-chilling feature of the potential storm won’t be wind, rain or even snow. Instead, it would be the fact that this potential Nor’Easter could be the final stage of an undead hurricane. Who said Halloween and weather couldn’t mix?

The threat was first detected by the long-range GFS last weekend as a recurving extra-tropical cyclone making landfall in New Jersey. Storm threats that far in advance are not usually realized. Every small computational and procedural error resulting from observations, interpolations, and approximations compounds with each successive calculation time step in numerical weather prediction (NWP). That means errors grow exponentially further into a numerical forecast such that at long lead times model forecast solutions change dramatically. One week later though, this potential nor’easter is still being detected by the GFS, which now has support from other global models, including the ECMWF.

Though the coastal storm has grown in support among NWP, the changes between forecast cycles remain quite dramatic.  There is still roughly one week before potential impacts from this potential Nor’Easter, after all. These dramatic changes between forecast cycles are due to  small differences in the initial conditions of subsequent model runs, analogous to the small differences between the starting parameters for a forecasting ensemble. The large uncertainty resulting in these run-to-run changes is best appreciated via an ensemble spread, visualized below in the locations of low-pressure centers from each individual member of the 12 UTC Sunday, October 21 ECMWF ensemble (EPS) for the morning of Sunday, October 28. The plot essentially says that the coastal storm could be located anywhere within a large area encompassing the North Carolina Blue Ridge Mountains,  the North Atlantic Ocean between Nova Scotia and Bermuda, and the coastal plains of the Mid Atlantic and New England.

Focusing on surface features like the low-center so early would clearly lead to trouble. Practically speaking, the track and position of extra-tropical systems should hardly be acknowledged so far in advance due to the associated uncertainty. Why is there so much uncertainty with cyclones compared to sensible weather? Like any dynamic system in the atmosphere, extra-tropical cyclones depend on a plethora of factors.

The most important factor is upward motion. Low pressure systems cannot intensify  if air is not increasingly evacuated at the upper levels of the troposphere with time. Unfortunately, vertical motion is impossible to accurately represent in weather models. Minute errors in the wind field compound to become exorbitantly large issues when used to directly calculate synoptic-scale (large-scale) vertical motion. Vertical motion from small-scale and micro-physical processes is itself an estimation, so errors here are even larger. Given the magnitude of these errors, small-scale features that are generated by large-scale features such as the location of the center of a low-pressure system are impossible to determine more than several days in advance.

Large-scale features like ridges, troughs, and temperature gradients are much simpler to represent than vertical motion. These features are all related to vertical motion and can give a forecaster strong clues about the eventual location and intensity of a cyclone. Crucially, they can be accurately represented as sinusoidal waves by NWP, a framework that greatly simplifies calculations and inherently minimizes errors. This representation fails at local-scales, but such scales are not necessary to evaluate the evolution of an upcoming cyclone.

There is actually a lot of consistency between global models regarding the evolution of the potential Halloweekend storm when looking at the synoptic picture. There are three primary factors that will influence the storm’s development. The first is a shortwave trough that will prompt the development of a deeply amplified trough. The shortwave of concern is the third and strongest of a series of three shortwaves traversing the Northern Plains this week. It will send a reinforcing blast of cold air southward into the Central US toward the Gulf Coast. The shortwave will intensify along the left exit region of a jet streak rounding the top of an amplifying ridge of the West Coast. The right entrance and left exit regions of jet streaks are associated with upward motion. The former two shortwaves will have no such source of additional upward motion and will gradually wither away. The jet stream will intensify to compensate for the strong temperature gradient between the warm airmass of the western ridge and the cold airmass of the intensifying shortwave.

500 mb Geopotential Height and Vorticity chart with three important features relevant to the coastal storm labeled. Graphic courtesy of Pivotal Weather.

The second primary factor is Hurricane Willa. As of 12 pm MDT, Willa was a 105 mph Category 2 Hurricane with a central pressure of 971 mb traveling northwest at 7 mph before it recurves northeastward toward Central Mexico Monday. Willa will make landfall in western Nayarit, Mexico early Wednesday morning. Simultaneously, a weak wave of low pressure will emerge along a frontal boundary along Texas’s southwestern coast. That low pressure system will largely idle until the remnants of Willa eject into Texas. The circulation of the former hurricane will overtake that of the developing low over the Texas coast and straddle the frontal boundary eastward toward the Southeast coast by Saturday morning.

The third primary factor is a cold-core high pressure system in southeastern Canada. This high pressure system will continuously draw cold air into the Northeastern US this week.  The cold is pivotal in intensification of extra-tropical lows, since they are fueled by temperature gradients.

Little intensification will occur between Friday morning and Saturday morning. The shortwave will still be in the Northern Plains, too far away for its cold airmass to reach the Gulf Coast. Zombie Willa, a former tropical depression undergoing extra-tropical transition, will not have an antecedent front to increase a temperature gradient. As time advances however, the shortwave’s cold advection will intensify a jet streak over the Rocky Mountains. The jet will increase the shortwave’s southeasterly trajectory, thereby increasing the transport of cold air towards the Southeast. The cold transport will concurrently lower thickness between atmospheric layers and lower pressure in the Southeast, forming a broad and highly amplified upper-level trough.

The middle and upper-levels of the troposphere will maintain the strongest temperature gradient Saturday morning. A jet streak will form around the base of the newly formed upper level trough in response. The jet will build toward the coastal low, which should situate itself near the left exit region, where the greatest source of lift will be located.

It’s at this point where NWP solutions diverge and the confusion unravels. If the trough is too broad and extends too far to the east, the jet streak and the coastal low will follow suit. Zombie Willa would track harmlessly out to sea. This solution would be more likely if the cold from the shortwave was weaker and spread across a large area. Alternatively, if the cold is more intense but focused over a narrower range, the trough will be more amplified, pulling the jet streak and associated coastal low toward the East Coast. As the low barrels northeastward, it will eventually interact with the separate cold airmass gripping the Northeast, fueling an intense temperature gradient. The storm would roar to life in this scenario, intensifying rapidly Saturday afternoon and Sunday as it pounds the East Coast with heavy rain, howling winds, and possibly snow for higher elevations of New York and Northern New England.  Guidance currently favors this latter scenario, so it is the most likely possibility at this time.

Ultimately, the Halloweekend storm will follow the jet streak. Model solutions depicting otherwise as was the case last week and is likely the case with many ensemble members should be discounted. It is too soon to say whether outdoor plans for the festive weekend should be canceled, since the strength and position of a future upper-level trough could either direct the storm out-to-sea or it could cause it to barrel toward the Northeast US. The latter is favored at this time, but that could soon change as the three features involved have yet to form. Be sure to return during the week, as we will be providing updates throughout the week as confidence increases.





Josh is a lifelong nature and weather enthusiast as well as the Head Meteorologist at WeatherOptics. He began regularly forecasting for New Jersey, Long Island and New York City in 2014 on social media, contributing to community pages such as SBU Weather. He holds degrees in Physics and in Atmospheric and Oceanic Sciences from Stony Brook University, from which he graduated in 2018. In the Fall of 2018 Josh will start graduate school for his M.S. in Marine and Atmospheric Sciences at Stony Brook, continuing his research on approaches to non-convective wind gust forecasting.

Comments are closed.