Our scheduled guest blogger, Mr. Ken Johnson, chief meteorologist at KAUZ 6 in Wichita Falls, TX will join us later this week, discussing what the transition from a strong El Nino to La Nina might be like. With model guidance, observations, and climatological reasoning all indicating a transition to La Nina, that blog will touch on topics that could prove very useful later this year.
In the meantime, isolated active severe weather along the northern tier of states continues as forecast by wxchasing.com. The Storm Prediction Center (SPC) has highlighted an area stretching from near Chicago toward the east as having a moderate chance for severe weather, including tornadoes, on Wednesday. The primary threat that day appears to be the risk for strong, damaging winds over a large area. As our last article discussed, this is the time of year for that area to take seriously any severe thunderstorm warnings that might be issued.
In the tropics, a weak tropical storm Danielle has come and gone in the Bay of Campeche, tracking into eastern Mexico. This storm was very short lived, and produced mostly rain and breezy conditions as its primary impact.
In the extended range (10-14 days) an interesting feature has been showing up on the Global Forecast System (GFS) model. After around 7 days, any forecast specifics become very vague and uncertain, but we can draw some useful knowledge at this range if we look at trends and consistency. This article’s primary focus will be to describe how best to utilize these very long range models when discussing tropical activity.
Often, models at very long ranges look somewhat similar to a climatological expectation. It’s not at all unusual for these models (particularly the GFS, which regularly forecast to 384 hours out) to churn out hurricane after hurricane that simply never verify. These storms are colloquially known as “Ghost Storms”. Ghost storms are the bread and butter of clickbait artists and are almost always available for an armchair meteorologist to point at and say “Hey, you guys, Katrina is coming back!”. I assure you that is not the purpose of this article.
Most often, these apparent ghost storms are fleeting. Here one run, gone the next. Other times, these storms seem to “hang out” at a certain timeframe beyond initialization, never actually getting closer to occurring. Often, a storm will continue to crop up at 12-14 days out. It may appear on one run of the model, but as days pass, it somehow is still 12-14 days out. When either of these two things happen, you can be fairly confident that the solution is pure fiction. Even rarer, sometimes these storms appear at the extended range and come closer as you would expect in a “real” weather forecast. These are the most important to watch and the trickiest to forecast. These are the ones where the chance for a real storm may be possible.
So, let’s look at the GFS’s forecasts. Remember to look for consistency, and an approaching timeframe as discussed above:
Consistency: The GFS has trended stronger for the most part through time, but shows at least troughing in the Gulf of Mexico on each run. That’s good consistency at this range that there will at least be some general storminess in the Gulf at that range.
Timeframe Concerns: From the 16th through the 18th, the GFS struggled with timing. The storm was seemingly drifting between 348 hours (14.5 days) and 384 hours (16 days). By 00z on the 19th, the timeframe became accelerated. From yesterday into today, the storm stepped forward in time further, from 324~ hours (13.5 days) to around 288~ hours (12 days). This isn’t the scenario we want in order to become really concerned about a system, but it’s enough to spark interest.
Once we’ve decided that a storm is at least plausible (though certainly not likely at this stage), it’s prudent to compare to other guidance. Most extended range models (GEM and ECMWF) only go out to around 240 hours, so that doesn’t give us much help at around 288 hours typically. However, with tropical systems, development occurs on the order of days typically so if we look at the end of the run and compare it to the GFS at 240 hours, we can see if the impetus for potential tropical development exists across guidance.
While the GFS is the most bullish at this stage, showing strong cyclonic curvature near the Yucatan, the GEM (CMC) and ECMWF (Euro) both show cyclonic curvature in the same general area as well as increasing wind speeds. This is indicative of a tropical wave. While that is no guarantee of development, a majority of tropical storms and subsequently hurricanes develop from the influence of these waves. These waves are also long lived, propagating from northern Africa. This increases the confidence in the forecast of wave positions compared to other storm modes because the feature already exists that is being tracked, even at 10 days out. So, we can say with some degree of confidence that in 10 days’ time, a tropical wave will be centered somewhere from the far western Caribbean to as far west as the Bay of Campeche. For now, we have little else to go on as far as tropical development, but we can look at the synoptic pattern to see where a storm located in that region might go were it to develop. Below, the ECMWF, GEM, and GFS 500 mb anomalies are plotted. This helps illustrate potential steering currents.
All three models show significant ridging extending from the Southwest into the northern Plains, with additional ridging off the east coast. Troughing is shown across the Mississippi Valley and eastern states, which would act as a pathway for a storm to drift northward. This trough would probably lift north and east beyond 10 days, potentially closing that path, but should a storm drift northward and become captured by the trough or should the trough linger longer than is typical, impacts in the United States would be possible.
Beyond this, we should recall the underlying conditions for development. Are the seas warm enough? What is the expected shear across the basin forecast to be like this year? Is there significant dry air already ongoing? How have storms in that region fared thus far?
In the post discussing the hurricane season, we saw that seas were warm, wind shear should be below average, and that significant dry air intrusions were so far absent. We also have seen 2 storms (Colin, Danielle) develop in that same general vicinity. And as we discussed in the seasonal forecast blog, this can be an indication of genesis and track of later storms. This informs us that the underlying conditions for development should be more favorable than is typical.
After analyzing the data, my conclusion would be that the first 5 days of July are at a heightened risk for tropical impacts across the Gulf Coast. We can be confident that lowering pressures in the Gulf are likely during this time frame. Do note, however, that a tropical storm or hurricane landfall is still NOT likely. With only 1.7 average hurricane landfalls per year (http://www.aoml.noaa.gov/hrd/tcfaq/E11.html) and the assumption that the vast majority of these occur within the 6-month hurricane season, that means that any 1-week stretch (assuming equal chances through the season, a poor but workable assumption) has around a 7% chance of seeing a hurricane landfall. That being the case, with the data we have available from observations, climatology, our seasonal forecast, and model guidance there is around a 10% chance that a tropical storm or hurricane impacts the US during this time frame. And due to that a 90% chance that the week goes landfall-free.
In short, don’t believe the hype, but do be paying attention! Should circumstance change, I will either edit a note below or issue a new blog update. The post above will remain in its original form as an educational tool on the analysis of “ghost storms”.
Be sure to check back later this week for our guest blogger! I’ll have a new post of my own up sometime next week, or sooner if the weather dictates. Have a great week!
Meteorologist Logan Poole