Forecasting Helene: Using Satellite Imagery

Some of you can name every hurricane from the last 50 years or have an honorary relationship with your local weather station. Others might be just dipping their toes in, having only spent, say, a grand total of 10 hours trying to figure out why everyone’s so obsessed with spinning clouds on satellite loops (and where Helene is going). And that’s totally fine! Whether you’re a seasoned storm-chaser or just learned the phrase “low level circulation is displaced” yesterday, the goal for these series of articles is to essentially provide the ‘hurricane forecasting toolbelt’. I’ll cover various products, terminology, and the types of analysis we can do with them.

---

Getting everyone on the same playing field

One of the first things to cover is the general configuration of a tropical cyclone. For now, we’re just going to focus on the vertical structure—I often like to describe it as a three layer cake. We have our: 

1. Bottom layer, or low-level circulation (LLC)

2. Mid layer, or mid-level circulation (MLC)

3. Top layer, or upper-level circulation (ULC)

 (image credit: when life gives you hurricanes, Publix gives you cake!)

We can assess these different layers of rotation, or relative vorticity, at various levels of the atmosphere (typically 850mb for LLC, 700mb for MLC, and 500mb for ULC). CIMSS provides excellent products that allow us to analyze these different layers separately.

Now let’s get into the fun part, satellite imagery! I like to do what’s called the top down approach (or starting with the top layer of the cake) so that we can assess the major synoptic features. Let’s do this with upper-level water vapor imagery.

---

How we can use water vapor imagery

In addition to identifying dry/convective air masses, upper-level water vapor imagery allows us to identify other key features. In the case of Helene, we can see the main features that have impacted track/strength over the last 24-48 hours:

We can identify what has limited strengthening (upper level low)

1. Our upper-level low (ULL) has been located to the northwest of Helene, and this has induced southwesterly flow at the upper levels 

2. While we can’t analyze surface flow with UL water vapor imagery, we know from the trade winds that our flow through the Caribbean is dominantly easterly

3. Since our wind vectors are in different directions (from the southwest at the upper levels and from the east at the lower levels) this creates vertical wind shear (specifically, directional shear)

4. With Helene remaining weak, this displaced the convection to the east

We can identify track guidance (ridging)

1. A blocking ridge at 500mb has largely prevented Helene from gaining significant latitude

2. For a weak system, the low-level flow dominates, and this led to gradual westward adjustments in the short term track. We can again use the mean layer steering products from CIMSS to analyze the steering flow for a weaker system:

In the loop below, we can see our ULL has become rain filled and asymmetrical. This indicates weakening of the ULL, and this drops the shear values over Helene. In the next section, we’ll look at visible satellite imagery and how this changes the convective pattern.

---

Using visible satellite imagery

With visible satellite imagery, we can begin to assess what’s occurring at the bottom layer, or surface. The image below was from 3:07 PM ET on Tuesday, and we’ll begin to see that as the ULL weakens, the convective pattern begins to change:

1. We now have persistent convection in the western quadrant

2. Decrease in southwesterly shear means convection can now rotate upshear (or towards the southwest when in the northwestern quadrant)

3. While the LLC remained displaced in this image from the MLC 

4. (slight side note from the weather discussion earlier: the short term southward component to motion indicated the LLC was likely oscillating on the periphery of the MLC, until they became colocated/stacked).

If we play the full visible loop below (from 3PM to ~7PM), we can see the LLC ‘tucks’ under the expanding convection, and we begin to see clear banding features (i.e., we have inflow bands that help the system become more symmetrical over time):

As we can clearly see above, Helene has entered the formative stages of building an inner core Tuesday night. We typically see pulse phases, or periods where we have convective bursts, followed by periods of the system needing to internally restructure (so to speak). Expect a few of those phases over the next 12-24 hours as Helene begins to ramp up—this is modeled to be an extremely expansive system (the NHC is forecasting it to be in the 90th percentile in terms of size for this region).

In the next posting I’ll cover the different types of models (ensembles, global/dynamic, and hurricane/mesoscale) and how they can be applied effectively!

- Matthew Weiberg