Why Create Automated Paragliding Forecasts?

There were several reasons why I created my algorithm and its resultant forecasts for all paragliding sites all over the Western Cape. It’s evolved to become even more useful for paragliding pilots in Cape Town and the Western Cape.

Early Warning for Paragliding Conditions

There was a day in 2019 when I noticed that Piketberg was looking promising for paragliding the next day. I shared this late at night in various WhatsApp groups but didn’t get much interest.  It was early in the season yet the day showed potential – at least for a thermic battle down to Renosterhoek. Somebody questioned my rationale as xc wasn’t yet on the radar for most. This experience highlighted the need for an early warning system to flag options, especially the ones we aren’t necessarily looking at. An early warning tool could help more paragliding pilots get motivated earlier and join in on and share flying opportunities.

Abundant Paragliding Sites in Cape Town

Cape Town boasts numerous paragliding sites. While this is a blessing for flying, it adds significantly to the amount of time for weather forecasting for each location. As paragliding pilots, we tend to narrow our focus on flying sites that we’ve flown before, or ones that are “hip” at the time. It’s funny how sites seem to grow and wane in popularity.  A narrow focus means we’re not looking at the other sites therefore limiting our flying options. By creating an automated algorithm that analyses all potential flying sites, more locations would come into consideration. This algorithm has personally benefited my flying, at the time of creation one of the first revelations for me was the potential for thermic flying at various sites around the peninsula.

Time Spent on Weather Forecasting

As a paragliding instructor, I spent a huge amount of time weather forecasting. Deciding whether to fly tandems at Signal Hill, take students to Porterville, fly solo recreationally, take an admin day, or just take a day off can be challenging. Additionally, RASP updates around 7 pm, which is often dinner time, extending my work hours into the evening and getting in the way of normal life. By writing an algorithm to analyse all the flying sites I’ve managed to save a lot of my time. Now all I do is wait to get a message from my telegram bot and then do quick in-depth analyses as required.

Other Apps and Tools Didn’t Align with My Needs

I found it frustrating to check different times and weather variables across various websites, pages, and pictures. Each time committing to memory what one variable said about a particular site. I wanted a simple table showing all potential flying sites for the next few days, broken down by time periods, allowing for quick evaluation. By writing my algorithm I could dictate exactly how the output would look making it as easy as possible for my brain to digest. Overtime emojis were added to the forecast to bring additional value without overloading the tables with too much info.

The Logic Behind The Algorithm

The logic behind my forecasting is fairly simple. Essentially, I’ve written code to automate my manual analysis of a site’s flyability—considering factors like rain, fog, surface winds, inversions, thermal strength and so on. While RASP is incredibly effective, other tools such as MeteoBlue offer (arguably) better insights on specific variables like fog. Over time, the algorithm has evolved, integrating a range of variables from RASP and other providers to enhance accuracy. I’ve also built in custom logic to make certain estimates, similar to intuitive guesses and deductions you would come to naturally in your head. The majority of the data used in these forecasts is sourced from the RASP model available at www.rasp.org.za.

The Negatives of This Tool

I’d like to say that weather forecasting is a passion of mine, truthfully it’s more a necessity. It amazes me that some people engage in life-risking activities without investing effort in understanding the air mass they’re flying in. Beyond accident risks, understanding the weather helps avoid wasted time and enhances performance. Knowing the forecast helps manage fear and make strategic decisions. A major concern of mine is that pilots might rely solely on my forecast as the final decision to go flying. This is a flawed approach. The purpose of this forecast is to highlight likely good flying sites. Pilots should still conduct thorough weather analysis.

How to Use This Forecast

The tables list the Boundary Layer tops (m ASL), Surface Wind speed (kph), and Surface Wind Direction. For example, "1579m | 9WNW" means a Boundary Layer top of 1579m ASL and a surface wind of 9kph from the west-northwest. If a value appears alongside a site, it means the site has passed multiple checks for flying potential. If a site is not listed, it’s failed one of my checks and is therefore unlikely to be flyable. However, listed sites should still undergo detailed weather analysis using tools like skew-Ts, regional wind analysis, etc.

Boundary Layer Tops

Why Use It

For a better understanding of the boundary layer and other RASP parameters, check out the RASP parameters description here. In simple terms, the boundary layer in paragliding refers to the height at which a thermal will reach. This could be the inversion layer, the height where the thermal temperature meets the environmental temperature or the top of a convective cloud. The boundary layer is useful as a rough indicator of inversion height. While soarable height is generally a better evaluation of how high you’ll get with a paraglider, boundary layer tops provide a quick, insightful glance which is also applicable in understanding the limitations of the forecast.

Why Not Publish Soarable Heights?

The main reason for not publishing soarable heights is to maintain a quick table-style overview. Too much information can make the tables unwieldy and decrease user experience. Additionally, the boundary layer top is perhaps more important because it is higher, making pilots like you more excited about flying. The boundary layer average wind is one of the important checks in my algorithm. The RASP model specifically predicts boundary layer information, so it is essential to consider its limitations. Publishing soarable heights instead of BL heights limits the info available to you. I’d rather you get excited by the BL top height and then dive deeper into lapse rates, soarable heights, and all the other fun stuff you love to look at.

Limitations

For sites like Signal Hill, boundary layer tops can appear quite low, even when thermals reach above takeoff altitude. This often occurs when the wind direction comes from the ocean, causing the model to extrapolate the boundary layer down to zero meters. When the wind direction is from inland, with a longer stretch of land upwind, the BL top values are often higher. Low BL top values, especially below takeoff altitude, can hide a strong wind gradient which is often the case on the prefrontal north westers for Signal Hill. Pay attention to the BL top and you’ll better understand the RASP forecast.

Regional BL Tops

Regional is a term I’ve given to the forecasts generally across the region. Experience shows that regional boundary layer tops are better indicators of actual inversion layer heights. Therefore, the forecasts include the afternoon (2 pm) values for Cape Town International, Porterville, and Sedgefield. Practically I’ve noticed that these regional values are better indicators of inversion heights that we experience. In addition, I’ve got there to always remind both me and my fellow tandem paragliding instructors of the potential to fly elsewhere. For example, if you see a 3500m BL top forecast for Porterville tomorrow, you might want to reconsider your work plans as you could be missing out on what could be the best paragliding flight of your life.

What Do the Emojis Mean?

Emojis have been added to score forecasts in terms of various thermal-related factors and the site’s historical data. Here’s a breakdown of what each emoji represents:

  • 🤔: The site is deemed flyable by passing a series of checks. Analyze the winds with altitude to make your call.
  • 👍: The site is flyable, and the soarable height* is higher than the takeoff height.
  • 🦅: The site is flyable, the soarable height is higher than the takeoff height, and the average thermal strength forecast is 3m/s or higher.
  • 🔥: The site is flyable, the soarable height is higher than the takeoff height, and the soarable height is in the top 25% of historical forecasts for this site.
  • 💥: The site is flyable, the soarable height is higher than the takeoff height, and the soarable height is in the top 5% of historical forecasts for this site.

Detailed Example

When you see the eagle emoji  🦅, it means that:

  • The site is deemed flyable by passing a series of checks.
  • The soarable height is higher than the takeoff height.
  • The average thermal strength forecast is 3m/s or higher.

*Soarable Height

The parameter known as “soarable height” is called hcrit in RASP. It is the height at which thermal strength is forecast to drop to approximately 1m/s (225fpm). Practically, this is the altitude at which you stop gaining altitude because the glider descent rate (~1m/s) equals the thermal’s upward velocity. For more information on this and other RASP parameters, click here.

If you love thermal flying like I do and see the 💥 emoji, think seriously about what you will do with your time – these days are rare!

A Note of Caution: 🦅, 🔥, and 💥 days indicate that the average thermal strength forecast on RASP is 3m/s or more. This is strong and recommended only for those with appropriate thermal experience.

Wind Speeds: Forecast vs. Reality

Experience shows that wind speeds at surface level tend to be approximately 10kph higher than forecast values. In general, a forecast wind speed of more than 19 kph is considered strong in my opinion – landing and takeoff will be in the region of 30 kph. Anecdotally, this corresponds with RASP colours changing from green to yellow on a normal wind day. Note that the RASP scale colours adjust with the maximum value for the map, so they can change significantly when there is very strong wind anywhere on the map.

Signal Hill and Lion’s Head Forecast

Given the tandem industry, these two sites are crucial for us. Decisions on flyability have a direct economic impact on many tandem pilots and businesses. Experience has shown that these are challenging sites to forecast accurately, but RASP does an amazing job. Given the level of pilots flying here and our deep understanding of these sites, I’ve added a detailed forecast for these two locations. This includes variables for takeoff, landing, Cape Town International (the regional “inversion”), and a point of interest for both.

Signal Hill

For Signal Hill, the point of interest is located 1km behind takeoff. This is helpful on wind shadow days where the shadow is small, and RASP shows westerly on takeoff but easterly right behind takeoff. From my observations, on wind shadow days where this point of interest is up to 10kph easterly, these are gambling days. More than that easterly tend to be a waste of time because the wind is over the back.

Lion’s Head

For Lion’s Head, the point of interest is located off the coast of Llandudno. This point helps assess whether the wind in the area is indeed SW or if it is S or even SE. RASP can forecast SW for Lion’s Head on days when the wind in the area isn’t SW. This is due to the local micrometeorology influenced by the 12 Apostles, Table Mountain and Lion’s Head itself. Experience shows that if the wind is not SW near Llandudno, then the flying conditions at Lion’s Head tend to be turbulent due to rotor turbulence coming off the 12 Apostles. As tandem pilots, we maintain a high safety standard, making this crucial information. Solo pilots willing to tolerate turbulent air may find it acceptable.

The detailed forecast for these sites simply states what the various variables are indicating and does not make any deductions on flyability. You’re an adult and a pilot, you can make your own decisions.

Morning, Midday, and Evening Updates

RASP updates three times a day, and here’s how it works:

  • Evening Updates (around 5 pm): The forecast for the next day starts updating. The update is usually complete by around 7 pm. For example, if it’s Monday night, then Tuesday’s forecast is updated. This is called “Tomorrow’s” forecast on the RASP website. The forecast includes results for 4km and 1.3km resolutions.
  • Morning Updates (around 6 am): The current day’s forecast starts to come out and is usually done by 8 am. For example, if it’s Tuesday morning, this forecast is for Tuesday and includes 4km and 1.3km resolution results. This is called “Today’s” forecast on RASP.
  • Midday Updates (around 10 am): A new forecast is run for the next few days. For example, if it’s Tuesday morning, the forecasts for Wednesday and Thursday are updated. These are called “The Day After Tomorrow” and “The Day After the Day After Tomorrow” on the RASP website. The resolutions for these two days are 4km and 1.3km for Wednesday and 4km only for Thursday.

Paragliding Site Gradings

Site gradings are categorized into three options: Basic, Sport, or GH (Ground Handling). This grading system is included here to help educate the community on the various grades as they stand today. If you hold a Basic License and are flying at Sport License sites, it’s good airmanship to consider the implications, especially for the rest of the flying community. There is a fine line between everything going perfectly and a serious accident. Sites are graded for good reasons, and if you believe a site’s grading is incorrect, you can get involved in the regrading process. For instance, we regarded Du Toits from Sport to Basic a few years ago after a fire wiped out the forest and made landing a lot easier.

Missing Flying Sites

Some sites have been excluded at the request of respected pilots in our community. This can happen due to concerns about damage to vegetation from increased popularity, which could result in the site being closed to flying. Other sites may be excluded due to safety concerns, such as crime. For example, pilots have been attacked at Wolfgat in Gordons Bay and therefore I’ve removed this site from my forecasts.

Of course, it is also possible that a site is not included simply because I am unaware of it.

Incorrect Forecasts, Questions, and Feedback

Firstly, it’s important to acknowledge that forecasts will never be 100% accurate. The provided forecasts are automated interpretations meant to flag potential flying sites. If RASP gets the forecast wrong and my results are incorrect, it is what it is. However, if you notice that I often do not call a site flyable when it is, this might be due to my site-specific parameters.

I encourage you to first check the RASP website to see if their forecast correlated with reality, indicating that the issue might be with my algorithm.

I’ve set up an open Telegram group where anyone can post questions, comments, requests, or feedback. Feel free to join and contribute to both questions and answers. I will participate where I can. You can join this group here.