First 500km Flight, Pikes Peak Bagged

This past Thursday was likely going to be the best soaring day of the week.  Still, I wasn’t sure whether it would be a great day or merely a good day.

SkySight predicted a strong thermal flying day with attainable flight distances from Boulder between 500-600k for 18m ships (which have a somewhat better performance than our club’s Standard Class 15m Discus). Note that going east from Boulder into the dark red colored area is not practical since most of that area is part of restricted Denver Class B airspace.
Topmeteo was a lot more bullish and predecited outstanding soaring conditions for almost all of Colorado with potential flight distances of up to 1000 km.

Thermal height was going to increase throughout the day to approx. 20,000 feet MSL over the mountains with the highest cloud bases to the southwest.  Thermals were projected to be strong with climb rates of up to 10kts or even more.  There were going to be winds of 10-20kts aloft, mostly from the southwest.  These winds could complicate thermal flying:  increasing winds with altitude tend to distort thermals breaking off from the ground and the strongest lift is often limited to upper-level convection, i.e. close to cloud-base.  In addition, it is important to be careful on the lee side of taller mountains where such winds tend to cause significant sink and rotor turbulence. There was a good chance for cumulus clouds and a low risk for overdevelopment.

From the soundings it looked like the day might start as early as 10AM over the foothills and significantly later in Boulder – a common scenario due to the morning inversion down in the valley.

As I did the pre-flight checks on the Discus there was a slight easterly breeze on the ground – also a very common phenomenon – and an indication that there was likely a convergence line somewhere over the foothills where the easterly wind from the plains meets the westerly airflow aloft.

SkySight illustrates the convergence line running along the foothills (shown in dark red): this is where the eastern airmass from the plains meets the dryer western air mass spilling over the Continental Divide. Note that this forecast is for 1:30 PM MDT. The position of the convergence line tends to shift throughout the day – often it moves further west as the day progresses. I speculated about the reasons for this in my previous post.

Whenever conditions conform to this fairly typical pattern I have learned that it is important to get to the west side of the convergence line as quickly as possible:  thermals on the east side tend to be fairly weak and often top out at less than 1000 feet AGL – especially in the morning when they are trapped by the inversion layer.  Unfortunately this often means a long (and expensive) tow over the foothills is necessary to connect with the lift on the west side of the convergence.  (Note that usually you don’t have to tow all the way back to the convergence but you have to tow high enough to be able to glide west until you’re able to connect with rising air on the west side.  If you don’t steer decisively west and linger on the east side you might quickly find yourself too low to get far enough west.  This is especially true early in the day when the thermals on the east side are still non-existent or extremely weak. )

When I launched at 10:42AM MDT the air above Boulder was still as calm as it could possibly be – a sign that the inversion still suppressed any thermal activity over the plains. The first time the air started to stir just a little was after the towplane had turned west over the foothills (west of Altona).  However, the climb rate on tow remained constant until we had climbed to 10,800 feet over Jamestown and I decided that I was now high enough to push further west on my own as we still had not reached a thermal.

Flight path on tow shown in dark gray. Flight path off tow shown in light blue. Release point over Jamestown. My first attempt to climb off tow was on the east side of the convergence line. Lift was weak and uneven. A few miles further west everything changed and climbing from 10,000 to 16,000 feet was almost effortless.

I headed straight toward a cloud a few miles further west.  East of Hidden Lake I reached the east side of the cloud and found some weak lift.  I tried to climb in it but the air was rough and the lift weak and inconsistent.  The wind drift was from the SE towards the NW.  These were all indicators that I had reached the convergence line but that I was still on the east side.  After several turns (probably too many) I decided to push to the western edge of the cloud.  And voila, just as I had hoped, I had reached the west side of the convergence line and within 10 minutes I climbed from 10,000 feet to just under 16,000 feet.

As I started to head south I could easily see the position of the convergence line by looking at the curtain clouds underneath the next cloud in front of me.

The curtain clouds in this picture are a perfect indicator for the position of the convergence line. Good climbs can be found on the west (dry) side.  The nice clouds further south (on the right hand side of the picture) suggest a quick and easy flight route ahead.  My position in this picture is a few miles north of Ward above the Peak-to-Peak Highway, looking SSE.  The city of Boulder is on the left edge of the picture.

Two quick climbs above Nederland and south of Idaho Springs took me up to 17,000 feet – high enough to push through the potential sink on the lee side of Mount Evans.  The air above Evans was rough but I found two reasonable climbs southwest of Mount Bierstadt that gave me the altitude needed to fly further southwest into Southpark where I saw some promising clouds north of Como.  At this point my goal was Buena Vista and – if all went well – perhaps Salida.

I climbed in ok-ish lift near Fairplay and headed for some nice looking clouds near Antero Junction.  However, none of the clouds along the western rim of South Park worked well at all.  I could not make out a good reason for this but when the fourth or fifth cloud provided only mediocre lift I decided to change course and save Salida as a goal for another day.  A few miles to the north I saw several nice clouds right through the middle of South Park, spread about 10 miles apart.

These worked and within 20 minutes or so I had reached the east rim of South Park.  This area worked particularly well: it was now early afternoon and the combination of wind and sunshine from SSW provided for great lift above the SW facing slopes on the east side of South Park.

At this point I had my eye on Pikes Peak. There was a blue hole right around the mountain but it was not nearly as big as during my prior attempt.

Pikes Peak approx. 10 miles away and – somewhat surprisingly – free of any clouds. At the point of this picture I was above 17,500 feet and the summit at 14,114 feet seemed within easy reach. However, it wasn’t going to be that easy.

The wind was blowing at 18 kts from SSW so I decided not to fly straight toward the mountain but instead to pass it on the west side and then approach it from the SW.  If I wouldn’t find any lift along the slope, I would have an escape route to the south towards Fremont County Airport near Cañon City, which I figured was easily reachable even flying against the wind.  There were also some nice looking thermals in that direction giving me extra confidence.

As I approached Pikes Peak the air turned very rough with strong sink.  A few miles SW of the summit I reached a low point of around 14,500 feet – just a few hundred feet higher than the mountain itself – too low for comfort to fly over the peak.  Fortunately I found a turbulent thermal breaking off from the ridge line below me.  Although it was extremely disorganized and caused erratic movements of my vario from +10kts to -10 kts I was able to climb to 16,700 feet – high enough not only to fly over the summit but also to reach the next cloud, which was approx. 10-15 miles further north.

I was surprised to find strong sink on my approach to Pikes Peak. The climb above the ridge SW of the summit was quite rough.  Also note the sudden (and non-obvious) change in wind drift from SW to WNW after I crossed the little lake below the summit.  The air near big mountains can be very fickle.

Having “bagged” Pikes Peak, I decided to head back towards Boulder while remaining on the lookout for worthwhile excursions.

View of Pikes Peak from the West just as I was getting into strong unexpected sink during the approach. The climb that took me back up over the top was just above the ridge line near the little lake on the right edge of this picture.

The clouds in front of me looked solid.  However, when I got to the nearest one, north of Woodland Park, the climb rate was disappointing.  I wondered if I was too far east – perhaps the air from the prairie had penetrated over the foothills, suppressing thermal activity?

I decided to turn west again towards the east rim of South Park where I had experienced much stronger climb rates some 40 minutes ago.  However, my next three thermals were equally poor and each time I decided to push on without gaining much altitude.  Lee-side sink north of Buffalo Peak caused me to loose another 1,500 feet.  I was now below 13,000 and worried that I might lose my connection to the clouds.

Just as my mind was working on fall-back options: Perry Park airfield was still within reach (though barely) and an off-landing field near Roxborough Park (which I recently checked out on the ground) was more easily accessible, I found a modest climb southeast of Bailey.  When you’re down to 12,700 feet over unforgiving terrain and with another 50 miles to go to Boulder you can’t be picky.  I carefully centered the thermal and happily spent the next 12 minutes climbing back up to 17,500 feet.

A succession of frustratingly weak climbs led me to a low point of 12,700 feet SE of Bailey. Perry Park Airfield was 30 miles away (requiring a 35:1 glide ratio). The nearest comfortable out-landing field was just over 20 miles away, NE of Roxborough Park.

It only takes one climb for the world to look totally different again. I steered to the SW side of Mount Logan (just SE of Mt Evans), climbed again to just under 18,000 and decided to fly north.  Maybe I would find lift above the Continental Divide so I could extend the flight to the north towards the Wyoming Border?

However, when I got to the Divide near Mt. Eva the route to the north along the Divide did not look promising.  Rain showers and virga obscured the sight and the area appeared over-developed.  To my left however, there was a nice-looking cloud street towards the SW.  I followed this line until Berthoud Falls when the route ahead also seemed be over-developing.

I made a 180 degree turn.  Looking north from my new vantage point I could suddenly see what was going on:  the convergence line that had sat over the foothills in the morning had crossed over the Continental Divide.  What had looked like rain showers and virga when viewed from the East side were in fact curtain clouds marking the position of the convergence line.  Further to the west the sky was completely blue.

Beautiful view of the convergence line on the west side of the Continental Divide. I took this picture after my northern turn-around point. The curtain clouds marking the convergence are clearly visible.

I decided to follow this line to the north, always staying to the west of the curtain clouds. It worked amazingly well.  I followed the line of curtain clouds that paralleled the divide.  Flying between 80 and 100 kts IAS I was able to maintain my altitude of around 17,500 feet over a distance of 38 miles in just 16 minutes.

The power of a convergence line: 38 miles in 16 minutes, that’s an average ground speed of 142 mph. Without losing altitude.

North of Longs Peak the line became less clearly defined and the cloud base dropped so I decided to turn around.  I followed the line south past Apache Peak and then decided to return towards Boulder.

Grand Lake and Lake Granby viewed from the west side of the Continental Divide near Longs Peak.

My final glide took me past Gross Reservoir towards Golden where I checked out the model plane airfield near Arvada from the air.  I recently visited this field on the ground and determined for myself that its perfectly paved 700 foot runway is a viable out-landing option (land on the N side of the runway to minimize the risk of hitting a fence and watch out for the power lines on final).

From there I followed the ridge line of the Flatirons on my way back to Boulder.  The winds at Boulder airport were 3 kts from the NE.  I briefly considered landing on G26 but ultimately opted for Runway G8.  I’m glad I did: as soon as I climbed out of the cockpit to get the dolly a freakish 27mph gust hit the airfield straight from the east. The gust lasted for about 2-3 minutes, then total calmness returned.  I have absolutely no explanation for this gust.  I just know that landing with a 27 mph tailwind would have been quite troublesome.

It was an interesting end to a challenging and rewarding first 500km flight. The full flight track is here.

Lessons Learned:

  • If there is a convergence over the foothills get to the west side before you are too low. This is especially true in mornings with strong ground inversions when there are still no viable thermals on the east side.
  • Always have a viable landing field in mind.  I caught myself a bit by surprise when I was down at 12,700 feet and noticed that getting to Perry Park would have been doable but already a bit of a stretch.
  • Approach big mountains with respect.  They can make their own weather and it’s not always what you might expect.  E.g., Pikes Peak surprised me with sink on the west side, which had direct sun exposure and was facing the wind. (It reminded me of a video from Bruno Vassel confidently approaching the Tetons from the west only to find unexpected sink.)
  • Keep in touch with the clouds when you can.  Throughout the flight I noticed that the climb rates improved the closer I got to the clouds.  This phenomenon seems to get more pronounced throughout the day.  Thermals down low were quite poor and wind-blown.  But close to the clouds the upper-level convection was quite strong, often with climb rates of close to 10kts or even more.  Staying high isn’t only safer, it might actually also make you faster (even though you have to center more climbs.)
  • Never get caught in lee-side sink.  I should have been more mindful of the terrain when I was flying just below 14,000 feet N of Buffalo Peak.  I turned northeast in direct pursuit of the next cloud and got into sink on the back side of Windy Peak. This could have been easily avoided had I followed the ridge line before turning towards the cloud. (It probably would have saved about 1,000 feet, which could make all the difference when it comes to reaching a land-out field.)
  • When it comes to wind, always expect the unexpected, especially in the pattern.  The freakish gust after landing was completely and literally out of the blue. (I had monitored the conditions in Boulder on the radio off and on for the past 20 minutes and conditions had always been calm.)

Venturing South

One of the advantages of soaring in the American West is that frequent cloud bases of 18,000 feet or more allow for significant flight distances without ever leaving the glide radius of the home airport – especially if you’re lucky to have a high performance plane.

20 miles from Pikes Peak

E.g., let’s say you’re flying a ship with a glide ratio of 40:1 from Boulder, CO.  The airport elevation is 5,300 feet.  Pattern altitude is 6,300.   That means on good thermal days you often fly 11,000 feet or more above pattern altitude.  And 11,000 feet of altitude at 40:1 equates to a glide distance of more than 80 miles!  This means, at least in theory, you can start a final glide at 18,000 feet MSL above Pikes Peak to the west of Colorado Springs and reach the Boulder airport 700 ft above pattern altitude – provided that you’re flying at optimum glide speed and absent any sink or headwind.  Reality isn’t usually so kind, i.e., you’re not able to maintain optimum glide speed, there is at least some cross wind component to deal with, and you’re spending more time in sink than in lift. Therefore it’s prudent not to rely on a glide ratio that’s better than half the theoretical maximum.  But even that gets you home from 17,300 ft MSL above Mt. Evans, or from 17,300 feet above the northern-most corners of Rocky Mountains National Park.  In other words, on days with high-reaching thermals (not uncommon) you can fly more than a 200 km FAI triangle without ever leaving the glide radius around Boulder.


Above Mount Evans

In my first 9 months of soaring from Boulder I have largely stayed within this “glidable” area.  Having acquainted myself with the area I feel it’s time to start venturing beyond these confines.

One of the motivating factors are the remarkable differences in soaring conditions in Boulder’s vicinity based on the overall weather situation.  This past Wednesday provided a great example.

Boulder lies directly at the foot of the Front Range of the Rocky Mountains.  Usually a dry air mass dominates to the west of the mountains, whereas a more humid air mass lies above the flat prairie to the east.  These two air masses tend to have different characteristics in terms of moisture content, temperature, wind direction, atmospheric pressure, etc.  The western airmass is typically pushed eastwards by the prevailing westerly airflow whereas the low level winds over the plains tend to come from easterly directions.  As the morning sun heats up the east facing slopes over the foothills, this easterly flow is augmented by convective activity, pulling in yet more air from the plains.  Where these two air masses meet and mix – usually somewhere over the hills in the vicinity of Boulder – the soaring conditions can get very complex.

On Wednesday, a south-westerly mid- to upper-level flow pushed dry air up and across the Continental Divide.  The air above Boulder was more moist and the temperature gradient was very stable.  A typical morning inversion lay over the plains.  The skies were blue and the sunshine intense but at 11:30 AM any lift over the city of Boulder still topped out at less than 1,000 ft AGL, trapped by the inversion layer.  Some wisps had formed over the foothills but they seemed to be short-lived.  The only nice cus could be seen some 30-40 miles to the north, inaccessible by glider from  Boulder.

As so often, Skysight had projected the convergence between the western and eastern air masses to follow a line over the foothills, parallel to the Continental Divide.  In the morning this line was supposed to be approx. 10-15 miles east of the Divide and over the course of the day, it was projected to gradually shift westwards – I assume due to the fact that increasing thermal activity over the hills would strengthen the easterly airflow and push the convergence further back towards the mountains.

Thermals to the east of the convergence were projected to be weak, especially between Golden and Fort Collins.  By contrast, much stronger conditions were forecast to the south of Mount Evans with particularly strong thermals along the foothills between Denver and Colorado Springs, and along the southern end of South Park towards Salida.  I tend to think that this difference has to do with the fact that the Front Range is much higher to the north or Mt. Evans than to the south:  where the mountains are high they slow down the westerly flow and the air from the east can penetrate further westwards.  Where the Front Range is lower (to the south of Evans), the westerly flow is stronger and prevents the easterly air mass from penetrating westwards.

The following sketch illustrates what I think the situation looked like in the morning on April 16.  My limited experience suggests this situation is pretty common in Boulder so I think it’s worth capturing it for my own learning.

Approximate situation in the morning of April 16, 2018. This seems to be a relatively frequent scenario in Boulder.

– a strong morning inversion at first prevented, then capped early thermal developments above the eastern plains (shown in blue)

– first thermals could be expected to pop up over the foothills where the morning sun warms the east-facing slopes.  A prime spot for first thermals would be over the Flatirons thanks to their steep eastern face which get the most direct sun exposure in the morning.  However, before noon these thermals would still top off at a few thousand feet above the terrain. (shown in green)

– better (but not great) thermals could be expected to the west of the convergence line. However, because this area is in the lee of the Rocky Mountains, these thermals would likely be turbulent and there could also be significant sink in-between them as air would spill down the mountains (shown in orange).

– much better lift could be expected south of Mount Evans (in the red shaded area) because the south westerly airflow could bring the dry westerly airmass into this area.  Note how the convergence line shifts further east.  Due to the wind direction from the SW this area would also not be subject to lee-side turbulence and sink. The best lift would likely be along the lower, west-facing slopes on the east side of South Park, especially after noon when the sun would warm these slopes most directly, and the wind would help to break the thermals off from the slopes below.

Given these projections my plan was to try to get towards the south.  If I could make it past Mount Evans, I expected to find much better conditions there.  If all went well, I might even make it all the way to Pikes Peak, and maybe from there into South Park towards Salida.  Knowing that this plan could take me well beyond conservative glide ratio calculations, I had researched the airfields to the south of Denver as well as a few potential off-field landing spots between there and Boulder, as well as in the area around South Park.

If the intent was to go south, the question was how to get there. It was around 12:45PM when I took a tow.   I released near Crescent Mountain (just south of the Flatirons) where I only found a very weak climb that topped off at 11,000 ft.  But it was high enough to push to a promising-looking cloud SE of Thorodin mountain where I climbed in 3-4 kt thermal lift to 13,000 ft.  This lift was still on the east side of the convergence line: the wind drift in this thermal is clearly from east to west.

The additional altitude allowed me to move a few miles further to the SW where I noticed a cloud with a higher cloud base.  This time I aimed straight to its western edge.  And indeed: I had made it to the west side of the convergence line where I could expect higher thermals and better climb rates!  The direction of the wind drift was a sure sign that I had made it to the “right” side of the convergence.  The climb rate was also significantly better at approx. 8 kt.  In only 5 minutes I had gained another 4,000 feet and was now at cloud base around 17,000 feet, ready to push further south.

Notice the yellow arrows depicting the opposite wind drift when circling southeast of Thorodin Mountain and northwest of Ely Hill: The climb near Thorodin mountain is in thermals on the east side of the convergence line: the wind drift is from the east. The climb near Ely Hill, just a few miles further southwest is on the west side of the convergence line: the wind drift is from the west.  So the convergence line must have roughly followed a line from Thorodin Mountain to Ely Hill.

I passed Mount Evans, following a row of clouds along the eastern rim of South Park, headed towards Pikes Peak.   Unfortunately the clouds ended and the sky turned blue when Pikes Peak was still 30 miles away.   This was clearly too much for my comfort zone and so I decided to turn and backtrack towards the NW.  20 minutes later I was back on the south side of Mt Evans.  Now the clouds looked better towards Pikes Peak and I decided to make another attempt.  This time I got within about 20 miles when, once again, the clouds came to an end.  Once again I decided not to take the risk.

Circling southwest of Cheeseman Lake, above pretty but quite inhospitable-looking terrain.

From there I headed north intent on seeing if I could fly the convergence line past Boulder to the north towards Wyoming, despite the much poorer thermal forecast for this area.  This turned out to be impossible.  Once I was north of Mt Evans, the conditions deteriorated quickly.  For a while I could still make out the position of the convergence line by looking at the positions of some interspersed clouds with different cloud bases but once I got to Ward thermals were few and far between, climb rates had deteriorated, and there was no viable way forward in sight.

Instead, I decided to head west towards an emerging cloud that was just west of the Continental Divide.  I pushed through heavy sink in the lee of the mountains and crossed the Divide at just below 15,000 feet MSL.  There was some ridge lift from the westerly winds along the divide, but 2,000-3,000 feet above the ridge it was just sufficient to maintain altitude.  The cloud was another few miles further west.

Exploring ridge lift over the top of the Continental Divide.

I had to decide between heading back towards Boulder, i.e. flying back through the lee-side sink with very poor prospects of connecting back up to cloud-base, or taking the risk of pushing for the cloud, firmly on the west side of the divide.  Given that Granby Airport was clearly within reach I decided to take the plunge and continued towards the west.  I got to the cloud – the only one around at that time – but found only very weak lift.  It took me 25 minutes to gain 2,000 feet back.  Had this been a race, here is where I would have lost it.

In the meantime, new convection had formed several miles further south, still on the western side of the divide.  Gingerly I continued my way south trying to stay at altitudes that would have allowed me to cross back to the eastern side.  A few poor climbs later I was able to fly directly over Grays Peak and Torreys Peak, two Colorado 14ers that I had not passed over before.

Circling in weak lift on the west side of the Continental Divide. Keystone ski resort is below.

Southwest of Mt. Evans I spotted what looked like a powerful new convergence line that stretched all across South Park.  It was already late in the day but the line looked extremely compelling so I decided to go for it.  I was not disappointed.  In 14kt (!) lift I quickly climbed back up to just under 18,000 feet and from there I had to shift my main focus to staying at legal altitudes.  The lift was so strong along this line that I had to open the spoilers at 100 kts IAS to avoid climbing above 18,000 ft.  Within 13 minutes I covered 30 miles without losing any altitude – that’s an average ground speed of 220 km per hour.  I don’t think I’ve ever flown a glider this fast!

View of South Park from the east rim, much earlier in the day. I have no pics of the flight under the convergence line as I had my hands full keeping the ship below 18k feet while flying at 100 kt IAS.

Given the late hour in the day I didn’t go further south than Jefferson before returning back toward the north.   I crossed the Divide NW of Mount Evans and continued in the weakening convergence line along the east side of the divide.  I flew all the way past Mount Audubon without a single circle (50 miles in just over 30 minutes), took a brief climb to fly over the top of Longs Peak, from where I started my final glide via the east side of Estes Park and Longmont to the Valmont Reservoir and finally back to the Boulder Airport.

Final climb of the day before flying over Longs Peak, straight ahead, just below the yaw string.

It was a long (five hours) and satisfying flight in very varied conditions where I experienced everything from weak thermals, and weak ridge lift to the strongest convergence zone lift I have ever encountered.  It was definitely my greatest triangle flight with a triangle distance of 276 km, and possibly my greatest distance flight overall with 472 km.

A link to the flight track is here.

Lessons Learned

  • If there is a convergence line parallel to the Front Range, getting to the west-side of it is key to climbing high enough to push to the south, past Mt. Evans.
  • Increasing thermal activity through the day may cause the convergence line along the Front Range to move further west as more air is pulled in from the plains, thereby strengthening the lower level easterly airflow.
  • Timing can be an important factor when it comes to reaching the convergence line.  Launch too early and the first thermals may be too weak and too low.  Launch too late and the convergence line may have moved so far west that you can’t reach it.  (Several pilots who launched after me struggled all day in middling lift around Boulder and were never able to make it to the higher cloud bases on the west of the convergence.)
  • The wind drift when thermaling can be a sure sign whether you are on the east or the west side of the convergence line.  This is particularly helpful when there are no clear indicators (e.g. in blue conditions, no different cloud bases, no curtain clouds).
  • Lift along a convergence line can be incredibly powerful.  I never before experienced lift as strong as 14 kts average.
  • A glide ratio of half the ship’s maximum seems to be a good rule of thumb when it comes to estimating a safe final glide distance.  (Note that this will not hold up in extreme conditions such as heavy lee-side sink, wave / rotor flying, thunderstorm activity, strong headwinds, etc.).
  • There are several things that I did not understand on this flight:  Why was the lift to the west of the Northern Front Range so weak?  What caused the late-day development of the convergence line parallel to the mountains on the NW-side of South Park and why was the lift there so strong?  Are conditions south of Mt. Evans typically stronger when the wind is from the SW?  Why did the convergence line to the north of Mt. Evans support climbs to 18,000 until Ward but not beyond?

The Smoking Tire – A Relict for the Wall of Shame

I have put off telling this story for several weeks.  It’s a lot more fun to talk about cool flights than to write about my own stupid mistakes. And this is definitely about a stupid mistake.

This is what happens to a tire when pulling on a break-lever too hard and too soon after landing. “Flat-spotted” is too nice a way to put it. This tire is destroyed.

So here’s the precursor to what happened.  On April 28 I had taken our club’s Discus to the west side of a north-south convergence line.  Cloud bases on the west side were close to 17,000 feet whereas to the east of the convergence conditions were much more difficult.  A persistent inversion lay over the plains supporting only weak thermals topping out at less than 8,000 feet.  With the help of a deep tow I was able to push into the western airmass and get all the way to 17,000 feet.  There were plenty of snow showers and virga above the mountains and my cautious self told me to stay within glide range of the Boulder airport.  Still, I had a good flight, scoring over 300 points on OLC-plus by covering more than 270 km including a 177 km triangle in just under 4 hours.   The flight track is here.

However, this story is not about the flight but about my landing.  And, as so often in aviation, stupid mistakes begin with chain reactions.

In this case the beginning was when I noticed on prior flights that the wheel break on the Discus was basically ineffective.  However hard I would pull on the break lever, there was no noticeable deceleration at all.  This is not an issue on the long runway in Boulder but it could be an issue when landing in a short farmer’s field. So I worked with others in the club to adjust the break and thought I would test it again on my next flight.

While I was flying another unrelated thing occurred: after about two hours the battery supplying the main electric system including the 2-way radio ran out of power.  This had not happened to me before. Unfortunately I didn’t remember that flipping a small switch on the instrument panel would have shifted the power supply to the ship’s second battery.  I made a mental note to get the battery replaced after the flight, but otherwise, having no power didn’t bother me too much at first:  the plane’s transponder, which broadcast my position to air traffic control was powered by a separate battery and still working fine. The airspeed indicator and the altimeter don’t require battery power and the ship is equipped with a second mechanical variometer that was also still working.  In addition, I had my own flight computer as well, which also has its own independent power supply. So the only thing I didn’t have was the radio, the optional Flarm system, the acoustic vario, and the ship’s built-in flight computer, which I didn’t need anyway.

As I came back to land, however, the lack of a working radio was back on my mind. There didn’t seem to be much traffic around the airport but without a radio I wasn’t able to announce my position and intentions and I also could not receive other pilots’ announcements.  So I concentrated on watching for traffic and on making my own position and intentions as clearly visible and predictable as possible.  I watched another plane land on runway 8 despite a slight westerly wind on the ground.  I remember thinking that I would land on Runway 26 (against the wind) if I could announce my intentions but with other traffic using runway 8 (which is the default runway in Boulder for calm conditions), I decided that I would also land on runway 8 despite the slight tailwind.

Somewhat preoccupied by these considerations I did not think at all about the break and the adjustments we had made to it.  So after touching down, somewhat faster than usual given the tailwind, I instinctively pulled on the break lever just like I had done on prior flights.  I noticed some deceleration and I remember thinking, “oh, the break is working now”.  I did not notice, or even consider, that I might have pulled the break lever too hard.

Just before the plane came to a stop the back pressure on the stick was no longer sufficient to keep the tail wheel on the ground, and the plane veered slightly off the tarmac.  Reactively, I must have pulled on the break again and thus the plane briefly dipped forward with the nose touching the gravel just before it came to a halt.

Other pilots watching my landing had noticed a smoke trail from my tire and came to tell me that I had been breaking way too hard.  Still, I had no appreciation for what “breaking too hard” could mean for the tire.  I basically destroyed it – there is no better way to say it. (See the picture of the actual tire above).

Dipping the nose into the gravel also caused some scratches in the gel coat on the underside of the fuselage.  Fortunately these are minor and only cosmetic in nature, and apparently relatively easy to repair.  Had the plane dipped forward on the tarmac and/or at a higher speed, the damage could have been much greater.

I’m obviously not proud of this incident.  With 1000s of feet of remaining runway in front there was no reason at all to break hard, or to use the break at all.  Drum brakes are not the most effective brakes and they are best used sparingly and only when really needed.  I decided to share this story because I hope other pilots may learn from it before “gaining” a similar experience themselves.  (Since my incident I have witnessed two other pilots damaging their tires as well).

I will also say that I learned a lot from the process of replacing the tire.  It is not a quick and easy thing to do and definitely a lot more involved than pushing the plane an extra 100 yards to its parking position. 😉

Lessons Learned

  1. Remember that battery switch, stupid!  If the plane you’re flying has more than one main battery it stands to reason that there is also a switch to toggle between these power sources.  If a battery runs of power, find that switch and use the good battery!
  2. Land against the wind whenever possible.  Even a slight tailwind can cause or exacerbate issues.  In this case it resulted in a higher ground speed at touch down, a longer ground roll, and it contributed to the plane veering off the runway at the end of the ground roll (because the back-pressure on the stick was no longer sufficient to keep the tail on the ground, and the rudder was no longer effective in steering while the plane was still moving). (However, I still think that my decision to land on runway 8 was acceptable considering other traffic, the fact that the wind was only light, and my inability to announce an approach to runway 26.)
  3. Breaking at speed kills the tire. Do not engage the wheel break at all when the plane is still moving fast unless you absolutely have to (i.e. there is a danger of hitting an obstacle on the ground).  The faster the plane moves, the more lift the wings still produce; therefore: the less weight is on the wheel and the easier the wheel locks up.
  4. You may not notice when the wheel locks up. If you use the break, use it gently and only when the plane has already slowed down.  You probably will not notice the wheel locking up when you engage the break, especially when the plane is still moving fast.   (The deceleration of a locked-up wheel is only small.)
  5. Only use the break when you must.  Safe the brakes in a glider for situations when you have to use them (e.g. a short off-field landing).  Don’t use them for the convenience of not having to push the plane back for a few hundred feet.   I believe for some planes with drum brakes this is even explicitly mentioned in the manual (e.g. for the LS4).
  6. Glider tires are soft. Tires of gliders are much more prone to flat spotting than car tires or bicycle tires.  Gliders also don’t have ABS systems 🙂
  7. Replacing tires on a glider is a complex and error-prone process.  Make sure to lubricate all moving parts (except the inside of the break itself!) and do not over-tighten the nuts on the axle for this may lock up the gear-retract mechanism.  After working on the wheel or tire make sure to cycle the gear retraction mechanism several times to ensure that it works smoothly and without requiring excess force.

My Skills Getting Tested – Challenging Start of the OLC Season

Beautiful view of the Continental Divide near Ward. The Peak to Peak Highway is directly under the wing.

This past weekend marked the start of the 2018 OLC season for the Northern Hemisphere.  For those who don’t know, OLC stands for Online Contest and is an informal worldwide soaring competition. Pilots anywhere can upload their flight tracks to a website hosted by a group of soaring aficionados in Germany. The tracks are automatically analyzed and classified according to the rules of various leagues.

My club, the Soaring Society of Boulder (SSB), is very active in the “OLC Speed League”.  The Speed League runs on 19 consecutive weekends starting on the third weekend in April.  Club rankings are determined based on the three fastest flights per club on any particular weekend during a 2 1/2 hour soaring window.  You can read the full rules here.  In 2017, SSB pilots won first place in the US Gold League and came in eight place worldwide (out of 1,162 participating soaring clubs).

Participating in this friendly competition seems to be a good way to track the progression of my own skill level over time when measured relative to the skills of much more experienced pilots.  At the same time, I am acutely aware of the potential risks that participating in any kind of soaring competition could entail.  I have written before about the risks of soaring, especially in a competitive setting, and ultimately it is up to the pilot to stay disciplined and put his or her safety firmly ahead of any competitive ambitions. This is the only way to stay safe.

Both Topmeteo and Skysight predicted weak to moderate thermals up to about 10-11k feet.  A snowstorm had just dumped a few inches over the foothills the day before and – unsurprisingly – the forecast looked best for thermals over the plains that were sure to be free of snow cover.

A strong inversion lay over Boulder as I drove to the airport and the air was still on the ground. Speculating that conditions would likely improve later in the day I delayed my launch until 1 pm and watched other pilots take to the skies before me.  The fact that most seemed to be able to stay up was encouraging.

Just before 1 PM I took off into the glider box just south of the airfield thinking that I would try to stay over the plains as the forecast suggested.  However, as the tow plane climbed through 6000, 7000, and 8000 feet the air did not stir one bit.  That’s when I got on the radio and asked the tow pilot to take me over the foothills where I saw that some white wisps had already started to form.

View from the foothills towards the east. You can clearly see the thick inversion layer above the plains.

I kept my hand on the release ready to let go when I would notice a tangible updraft but the air remained still for a long time.  We had climbed to almost 11,000 feet above Big Horn Mountain when I decided it was time to set myself free even though we had still not crossed a single patch of rising air.

I pointed the nose straight towards a tiny cloud that was forming above Gold Lake.   Just as I hoped, I found the air stirring just enough to slowly gain some altitude back.  I could also see a convergence line with a much higher cloud base several miles further west but getting there seemed very difficult or even impossible to me given the lack of usable landing spots over the foothills.  With no other reliable options for lift in sight I decided to hold my ground for a while and wait for conditions to further improve.

After flying holding circles for almost 45 minutes a few additional small clouds had popped up here and there and I decided it was now or never if I wanted to get some cruising miles in.

The OLC rules require that the start of any flight has to be within 15 kilometers of the take-off airport.  Gold Lake is 20 km away from Boulder and I could not remember whether I had released from tow before or after leaving the 15 km radius around the Boulder airport.  So I decided to first head back and fly through the start cylinder near Lee Hill.

From there I returned to the area of lift near Gold Lake.  Now I had to decide whether to go north or south.  I remember flying several circles unable to decide.  There was a promising looking cloud with a slightly higher base about 6 to 8 miles to the south but it was uncomfortably far away.  If it did not work when I got there I would have no option but to bail towards the airport.  There were a few smaller clouds to the north.  Although they looked less compelling and had somewhat lower cloud bases they seemed to offer a more promising path forward overall.   My indecision with respect to my course direction clearly was not a good tactical move: I had already crossed the start line, the clock was already ticking, and I was just staying in place…

I have found once again that the best thermals are often near bodies of water. The temperature contrast between the cool air above the lake and the much warmer air above the surrounding forest and grassland can serve as trigger. The additional humidity that gets sucked into the thermal from the surface of the lake enhances the thermal.

If took me almost 15 minutes to make up my mind but eventually I decided to take the route to the north.  Once made, the decision felt liberating for now I had a direction and a plan I was going to pursue.  Why could I not decide quicker?

As I headed north towards the Twin Sisters I spotted some newly forming wisps that were about 2,000 feet higher than the cloud I had just left behind.  It seemed like a long shot but I wanted to give it a try.  I thought that if I could gain another 2,000 feet I might be able to get under the convergence line.

However, as I crossed the area below these wisps there was nothing but sink. There went my hope for reaching the convergence.  Fortunately I had already worked out a plan B and a plan C.  Plan B was sufficient.  A small cloud above some rocky outcroppings north of Cabin Creek allowed me to get back to cloud base at 10,800 feet and to continue my track to the north.

I headed toward another small cloud just across US36 between Lyons and Estes Park.  My Oudie indicated that my altitude was now barely enough to make it back to Boulder.  The cloud worked again and provided the strongest lift of the day with an average of almost 4kt.

This allowed me to keep going a little further north until I reached a point between Estes Park and the north side of Carter Lake where my Oudie indicated an arrival altitude above Boulder just below pattern altitude.  Considering the generally weak conditions it felt the right time to turn and head back south.

I followed a similar route on my return leg tracing along the most promising little clouds.  I made sure to maintain a reasonably comfortable altitude above the undulating terrain, rarely dropping below 1,500 feet AGL and never below 1,000 feet AGL.  I also always kept an escape path towards the plains and generally was within glide range of the Boulder airport (albeit sometimes with little margin).

View of the Convergence line along the Continental Divide. My glider in the foreground is to the east of the convergence line. The yellow arrows point at the relatively low hanging curtain clouds that separate the two air masses along the convergence. The cloud base east of the convergence was around 11,000 feet. The dotted red lines mark the much higher cloud bases to the west of the convergence line. Cloud bases there were around 14,000 feet and thermals there were likely considerably stronger.

Little by little, cloud by cloud, I made it to the town of Nederland, 30 miles from my turnaround point.  It was already 4PM MDT and the clouds began to dissolve around me.  So I decided that it was a good time to for a scenic cruise back to Boulder.I took advantage of the Discus’ 42:1 glide ratio and detoured via Gross Reservoir to Eldorado Canyon.  From there I followed the ridge line of the Flatirons  where I provided some entertainment for the hikers atop of Bear Peak.  The easterly flow was unfortunately insufficient to maintain altitude when soaring along the ridge.  (The windward side of the Flatirons was already in the sun shadow so I suspect any lift from the wind might have been negated by cool air descending the face of the mountains.)

Overall, this was a challenging but satisfying start to the 2018 OLC season.  Looking at the score board of the OLC Speed League, my flight was just fast enough to qualify to be scored for the Speed League as the third of the three Boulder flights that count this weekend.   (The Boulder pilot who flew the greatest distance yesterday did not fly through the start cylinder and consequently his flight doesn’t count for the speed league.)

A link to my flight track is here.

Lessons Learnt

  • Safety First, Always.  Not a new lesson but worth keeping in mind, especially when flying with a competitive streak.  There is nothing to be gained in soaring competitions; however, many lives have been lost when competitors didn’t always put safety first.  Only fools risk life and limb for no gain.  So don’t be one.
  • Always Keep a (Safe) Escape Path.  The terrain over the foothills is tricky.  Your computer may tell you that you are within glide range of the airport but it might not account for terrain that’s in the way.  Be especially careful south of Nederland where there is higher terrain to clear to the east if you want (or need) to get back to the plains.
  • You Cannot Fly In the Foothills, You Have to Fly Above Them.  When flying in the Alps you often fly very close to terrain and most of the time the steep valleys provide escape routes into wider valleys with land-out options.  There are no land-out options in the foothills and the canyons don’t provide safe escape routes into the plain.
  • 1,500 Foot Ground Clearance Above the Foothills Feels Ok.  1,000 Foot Feels Low.  That’s for a high performance ship such as the Discus.  For lower performance ships, maintain more ground clearance.  If the thermals don’t support it, get out of there while you can.
  • Don’t (Blindly) Trust the Weather Forecast. Again.  Both Topmeteo and Skysight had predicted the best thermals over the plains.  In reality the plains – where the inversion was very persistent – provided only very weak lift up to 8,500 to 9,000 foot while the better lift was clearly over the foothills and mountains. (Much of the snow over the foothills was gone by early afternoon and the south facing rocks heated up nicely.)  Skysight missed the convergence line along the divide.  (Topmeteo does not predict convergence.)  I will keep reading the forecast but always consider that reality is likely to be different.
  • Good lift can often be found next to lakes.  Today the first lift I found was next to Gold Lake.  Two of my other thermals were next to lakes too.  Most textbooks tell you to stay away from lakes but my (limited) empirical evidence suggests that the best thermals are often next to bodies of water.  There’s also a great German soaring textbook called Meteorologie für Segelflieger by Henry Blum (Meteorology for Soaring Pilots) that convincingly argues that humidity enhances thermals and the best ones are often found next to lakes, rivers, or next to moist forests, mainly because moist air is lighter than dry air.
  • Indecision Costs a Lot of Time.  If I want to improve my performance for the speed league, I need to get better at making decisions based on the information available at the time instead of flying holding circles until I have made up my mind.

Surprise: Fastest Flight of the Day

Wednesday, April 4.  I didn’t try but my two-plus-hour-easy-cruising soaring flight ended up being the fastest flight in a glider on that day.  Fastest as in: highest average speed. Worldwide. That came as a big surprise to me because I hadn’t even thought about it.  I had just been flying along – as it turns out at an average speed of almost 152 kilometers per hour (82 kts).  Equally surprising is the fact that the 325 kilometer flight was also the 10th longest flight that day. Here are the stats from the Online Contest:

List of soaring flights on April 4, 2018 sorted by average speed in kilometers per hour.
List of the longest soaring flights on April 4 sorted by flight distance in kilometers.

So how did that happen?  The answer is easy: strong winds from the west, increasing with altitude, blowing across the Front Range of the Colorado Rocky Mountains. In other words: mountain wave.

Looking at the sky in the morning it wasn’t all that obvious that a strong wave day was afoot.  Here’s a short time lapse of the sunrise from our home in the foothills looking east.  The wind seems to be coming more from the left, i.e. from the north.  The fuzzy edge of the cloud does not look like a lenticularis and you have to look very carefully to spot any indications of rotor activity.

The forecast from Skysight on the other hand looked very confident:

Forecast of vertical velocity at 13,000 feet for 13:00 MST (Skysight). The dark orange band indicates wave lift of approx. 5 m/s (10 kts) over the Colorado foothills along the lee side of the Rocky Mountains. Winds are westerly with a slight northerly component. 

Around noon, however, the wind on the ground was blowing firmly from the east.  The sky was overcast due to a layer of high clouds and seemed deceptively calm.  The only real indication of wave aloft came from pilots flying into Jefferson County Airport a few miles south of Boulder, who reported moderate to severe turbulence a few thousand feet above ground:  just because you can’t see any rotors doesn’t mean that there aren’t any.

Just before 1pm I was ready to launch.  The air became turbulent at about 1,000 feet above ground and when I hit the first strong climb at 1,600 AGL I released without hesitation. That turned out to be a mistake. After climbing in very choppy lift to 2,300 AGL I bumped up against a strong wind shear layer that I was unable to get through.  After a few attempts in different locations I returned to the airport and decided to take another tow.

This time I asked the tow pilot to take me above the wind shear layer and over the foothills.  After a very bumpy ride on tow where I involuntarily practiced several slack line maneuvers I released at 10,400 MSL directly in rotor lift. I had no difficulty to climb to 12,500 MSL where I first encountered laminar airflow and the climb rate improved. Within a few minutes I had ascended above 16,000 feet and from there everything became very easy.

The lift was strong and consistent, the wave bar wide and forgiving.  The wind was blowing at about 40 kts but the Discus flies fast and so I could easily cope with the necessary crab angle and make rapid progress relative to the ground.  As the air got thinner and thinner the difference between Indicated Airspeed and True Air Speed (and therefore ground speed) increased.  The lift was so strong that I had to trim all the way forward and fly at 100-110 kt IAS to avoid climbing above 18,000 feet.  At several occasions the lift was so strong I even had to open the spoilers to stay below Class A airspace.

View of the Front Range from 17,500 feet MSL. The cap cloud above the mountains was poorly defined. You can see the Föhn Gap at the top of the picture with a thin cirrus layer high above.

The clouds were only moderately useful to gauge the location of the wave.  There was a fuzzy cap cloud along the front range with a Föhn gap between it and the next layer of clouds to the east.  However, a high cirrus shield often obscured the position of any lenticular clouds.  For most of my flight I navigated by looking at the position of the mountain ranges relative to the direction of the wind, which blew from northwesterly directions.

Cockpit view. The town of Estes Park is directly in front of the nose and about 10 miles ahead. Note the fuzzy cap cloud indicating air streaming down the mountains on the left. There is one very clearly defined lenticular cloud far in the distance to the right of the picture (the brightest cloud in this shot). The high cirrus shield on top obscured the location of other lenticular clouds.

The position of the primary wave was a few miles further west than Skysight had predicted, and the northerly component of the wind was a bit more pronounced, but other than that the forecast turned out to be pretty accurate.

Unlike prior wave flights, I had no difficulty passing Longs Peak and continuing beyond Estes Park to the north.  To the south I crossed I-70 until I had a great view of South Park behind Mount Evans to my right.  I briefly considered an attempt to fly past Mount Evans but decided to err on the side of caution not knowing where I would land in South Park if things didn’t work as well as I thought they might. (I later learned that Bob Faris, another Boulder pilot flying that day, ventured into the South Park area.  He reported that the wave turned violent around Mount Evans and that conditions were much more difficult further south.  He dropped below the laminar layer and worked rotor lift and thermals all the way to Fairplay and back.  Otherwise his average speed would have been much faster than mine.  His flight track is here.)

Nice view of Mount Meeker and Longs Peak partially shrouded by the cap cloud.

So I kept cruising back and forth along the Front Range.  There was very little effort involved and not much decision making.  I enjoyed the scenery and was happy that my transponder was broadcasting my location and altitude so that air traffic control could make sure that no jets would suddenly emerge out of the cap cloud and into my flight path. After about two hours I felt quite cold and decided to return to the airport.  There is no doubt that I could have kept yo-yoing along the mountains for several more hours. It was one of my easiest wave flights to date.

The return from laminar airflow into the rotor zone brought me back to the harsher realities of wave flying. I encountered severe turbulence as high as 16,000 feet in the secondary wave and cautiously descended with open dive brakes at a save speed of around 70 kts through some of the most violent wind shear turbulence I had so far experienced. Fortunately I had remembered my lessons from earlier flights and tightened my straps and removed all loose items before starting the descent.

As I got close to the airfield I was surprised that the wind on the ground was still blowing hard from the east.  I checked three times to make sure that my mind wasn’t playing tricks on me.  There was no turbulence in the pattern and the landing was smooth and easy.

Here’s a link to my flight track.

Post flight: the level of moisture increased later in the day and the position of the wave became much more visible.  Here’s a short time lapse of the wave at sunset:

Lessons Learned:

  • Wave may be hard to see even if there are clouds in the sky. For most of my flight, the moisture level was too low at the height of most of the wave activity so no clouds formed except for the fuzzy cap cloud over the Front Range and the high cirrus shield above (that probably had little to do with the wave itself.)
  • Being fast can be really easy if the conditions are right. I made no attempt to fly fast.  My high average speed was simply a function of flying straight in consistently strong lift, even requiring high speeds to avoid climbing above 18,000 feet into Class A airspace, coupled with flying at high altitudes where true air speed is 36% higher than indicated airspeed.  So my cruising speed of 100 kts IAS was really 136 kts TAS at 18,000 feet.  My average ground speed was “only” 82 kts.  A big factor explaining that difference is the crab angle necessary to compensate for the wind drift when flying along a wave bar (when the wind must logically always come from the side).
  • Ground speed varies dramatically in strong wind conditions depending on your heading relative to the wind.  This is rather obvious and not a “new” lesson.  But the stats make it very clear: my flight trace shows that my maximum ground speed was more than 160 kts (considerably faster than the maximum airspeed of the glider) even though I never flew above 110 kts IAS, while my minimum ground speed was below 20 kts (much slower than the minimum air speed of the plane).
  • Remember that flutter, and therefore Vne, is a function of TAS, not IAS.  Don’t trust the red line on the airspeed indicator to determine how fast you can fly safely at altitude.  Fortunately the Discus is built to go fast, even high. According to the flight manual,  Vne is 135 kts all the way to 13,000 feet and only then begins to drop off.  At 16,400 feet it is still 131 kts and at 19,600 feet it is 124 kts.

It Doesn’t Take Much Sun

Thermal flying under overcast skies

Yesterday was supposed to be a really good day for my first thermal flight of the year.  I now use several soaring forecasts to see which ones are most reliable in Colorado. However, this tends to be more pain than boon and can be quite confusing because these forecasts rarely agree with one another.

To my amazement, yesterday they all pretty much lined up, forecasting moderately strong thermals between 11:30AM and 5PM MDT, attractive attainable flight distances, no risk of overdevelopment, and an amazing 100+ mile convergence line all along the foothills.

Thermal height around 14,000 feet over the foothills.
Possible flight distance in a standard class glider (like the Discus) around 400-500 km.
Virtually no risk of overdevelopment.
Amazing convergence line all along the foothills of the front range – from south of Denver, CO all the way to north of Cheyenne, WI.

The thin cirrus shield above the plains soon dissolved after sunrise and the day started out under cloudless blue skies.

The thin cirrus clouds visible at sunrise soon disappeared and the day started out blue.

I planned on a takeoff around 12:30PM MDT.  When I arrived at the airport around 11 AM the sky looked just like the forecast.  A long line of cumulus clouds had already formed to the west of the Boulder airport, stretching from Golden, CO to the north as far as the eye could see.

I hadn’t rigged a plane in more than four months and things went slower than I had expected.  There was also a lot of activity at the airfield as many pilots had turned out for their spring checks and by the time I was ready to get in the air it was already past 1PM.

In the meantime the clouds had developed much faster than projected and when I finally took off around 1:20PM, the sky had become completely overcast.

The cumulus clouds during the second half of my flight were fairly easy to read and provided good indications of lift. During the first hour of the flight however (sorry no pictures), the sky was an amorphous grey.

There wasn’t much lift as I followed the towplane into the foothills where I hoped to find rising air from the convergence.  The clouds were dense and grey and I could not discern where the best lift was likely to be.  With my hand on the release I followed the tuck for a long time and finally set myself free in weak lift over the mountain hamlet of Ward at an altitude of 11,000 feet.

I looked all over the sky around me and still detected very little movement in the clouds.  I was able to hold my altitude in the narrow, broken lift and was basically just buying time to see if the conditions would change.  I was also at the bottom of a wind shear layer and had to pay attention not to stall each time when I turned into the direction of the wind.

Some streeting along the conversion line over the foothills.

After more than 10 minutes of parking in the sky I saw than the sun had broken through the clouds  a few miles further to the east, directly warming a south facing slope.  I held my position for another five minutes to give the slope some time to warm the air before making my move.

By the time I got there the slope was already in the shade again and I was doubtful that five minutes of sunshine could have made much of a difference.  However, to my surprise the slope actually worked.  The lift wasn’t strong but I managed to climb about 1,000 feet in six minutes before the energy was exhausted.  In the absence of clear indications in the clouds, the same strategy helped me locate my next lift as well.

Pretty view of the Continental Divide from a position along the Peak-to-Peak Highway south of Ward. The Eldora ski resort is just atop the wing and James Peak is in the background.

Then the weather changed.  The clouds over the plains started to dissolve and once again I headed for the area that was in the sun.  I followed the first row of foothills where the canyons open up towards the plains.  There are several bowls into which an easterly wind from the plains gets funneled. Yesterday there was a light wind from the southeast at the lower levels.  Combined with the afternoon sun (shining from the southwest) I was hoping to find lift above one of these bowls.

Lift along the bottom of the foothills with winds from the southeast: the wind is funneled into the canyons and up along the south-east facing slopes that are warmed by the sun. When the air reaches the top of the slope it can no longer cling to the ground and instead rises above.

I crossed over these bowls flying from north to south (from right to left on the map above) and found my best climb of the day in the Seven Hills area (the second such indicated bowl from the left) allowing me to climb from 7,500 to 10,500 feet in about 12 minutes.

As I was climbing, dark clouds began to rapidly form once again.  Unlike earlier in the day they were much better organized and provided much clearer indications of the areas of lift.  I followed a cloud street a few miles north where I effortlessly climbed to cloud-base.  Then I pushed south-west under another row of dark clouds where I finally topped out at 12,000 feet, my highest altitude on this flight.

Once again, the sky had completely overdeveloped. Heavy snow showers had engulfed the peaks around Rocky Mountains National Park, about 25 miles north of my position.  With the sun completely shielded off, the thermals rapidly weakened once again.  Even under the completely closed ceiling there was still enough lift to stay in the air.  However, climbing was slow and after a while I lost interest in flying holding circles and decided to return to the airfield.

Here’s a link to the flight track.

Lessons Learned

  • Five minutes of sunshine can be enough for thermals to form.  If the temperature profile is right (i.e. the air is sufficiently unstable), it doesn’t take long for the sun to heat the ground for thermals to develop. The sun at the end of March is already quite powerful.
  • When the clouds don’t tell you much, the ground can. For the first 90 minutes of today’s flight I was unable to read the clouds for indications of lift. Observing areas where the sun broke through the clouds to warm the slopes, and overlaying my understanding of the direction of the wind helped me identify areas of lift.
  • Weak thermals can even persist in a completely overcast sky.  Thermals definitely weakened when the sky was completely grey and overcast but there was still enough lift to stay airborne.
  • Parking in weak lift can pay off. When the sky over-develops and the thermals dramatically weaken it may pay off to hold on to a spot with weak lift where you can hold your position and wait for conditions to change.  Had I not done this, I would have landed within 40 minutes of releasing. I just waited long enough for the sun to break through the clouds and warm certain areas long enough for new thermals to form.
  • Stalls can happen very quickly when thermaling in wind-shear conditions. Finding the best speed to fly when circling in narrow, broken lift can be quite tricky. At my first climb of the day staying in lift required steep circles flown just above minimum speed.  I was coming up to a wind shear layer and once in three circles or so I was hit by a gust from behind that was sufficiently strong for my flying speed to drop below stall.  Small, hard-fought altitude gains can quickly be lost in a stall and then the slow and steep thermaling technique becomes quite inefficient. At one point my inside wing dropped and I had to quickly correct with opposite rudder. Flying so slow is definitely a no-no when close to the ground (I was at least 2,000 feet AGL and there were no other gliders around so safety was not an issue.)
  • The weather forecast can be wrong even when all forecasts agree. I learned that I cannot rely on the forecast even when all forecasting tools say the same thing.  None of the forecasts for yesterday predicted any over-development. Forecasts are still far from perfect even when based on data collected just a few hours before the flight.


Hypoxia Simulation – Get To Know YOUR Symptoms

On August 14, 2005, 121 people died in the crash of Helios Airways Flight 522 after the aircrew became hypoxic due to the air pressurization system being incorrectly set to manual.

On April 1, 2011 a glider flight from Boulder, CO ended in a fatal accident after the pilot had spent 14 minutes above 22,000 feet. From there the sailplane spiraled to the ground. The accident report found hypoxia of the pilot to be the most likely cause.

These accidents were on my mind when I attended yesterday’s Hypoxia Simulation Training session, provided by AirCare Facts at Independence Aviation in Centennial, CO.

After an hour of classroom training covering the causes as well as the potential signs and symptoms of Hypoxia, I had the opportunity to participate in a simulation of low pressure conditions at up to 28,000 feet.  This was accomplished by breathing through a mask feeding reduced levels of oxygen into the respiratory system.

Hypoxia is an insidious killer because it is often very difficult to recognize any symptoms before it is too late.  The potential symptoms even include feelings of wellbeing and euphoria, which may make it even less likely that a pilot would take corrective action before passing out (and eventually dying – either due to oxygen deprivation or due to the plane crashing in uncontrolled flight).

The only good news is that the symptoms of hypoxia tend to be specific to each individual and relatively constant over time.  Hence, it is possible for everyone to experience and “get to know” their early indications that something may be amiss.  Recognizing these indications early is likely one’s best (and maybe only) chance to take the necessary actions.

At the earliest onset of hypoxia symptoms at altitude it is vitally important to act immediately (while still being “usefully conscious”). Normally this means beginning a rapid descent to lower altitudes where the air pressure is higher and normal oxygen saturation levels are restored relatively quickly (normally within a few minutes).

I took the following video during my own training session so that I would be able to see my own reaction and be able to remember my specific symptoms.

I can recommend to any pilot to participate in such a simulation. Knowing your individual symptoms may one day safe your life.

The Dangers of Sailplane Racing – What Condor Taught Me

I recently demonstrated that soaring is an objectively dangerous pastime.  On a per-activity-hour basis it is approx. 35 times as dangerous as driving, 70 times as dangerous as bicycling, and still about 3 times as dangerous as riding motorcycles.

One contributing factor has to do with the high number of fatalities during soaring competitions. (This article shows that during global soaring contests, the number of fatalities per number of flights has been more than 10 times higher than during flights outside of competitions.) Even though there is no (relevant) price money on the line, contests tend to tempt pilots into lowering or suspending their normal safety standards. To have a chance of winning or placing well, pilots are often inclined to take higher risks than they normally would accept – consciously or subconsciously. E.g., they will fly closer to terrain than they would on a normal cross-country day; they will fly in bigger gaggles, thermal closer to stalling speed, get closer to Vne – even in turbulent air, attempt safes lower to the ground, scrape across ridges or mountain passes, fly low over unlandable terrain, calculate their final glides with a narrower margin, etc.

It’s the same behavior I observe (and – to be honest – participate in myself) during races on the Condor competition soaring simulator.  Fortunately it’s a simulator so if you crash you still get to live another day. But I’ve found that the dynamics in human behavior are very similar to real life competitions. No one wants (or expects) to crash but at the same time, most everyone flies in ways that they would consider irresponsible outside a contest environment. Sure, taking great risks won’t guarantee a good placement, but a good placement almost inevitably means that the pilot assumed a high degree of risk.

Here’s an example: yesterday I flew a Condor race set in the foothills of the French Alps as part of a competition called Regatta Cup. It was a short 188 km Club-Class task along and across several low mountain ridges. Thermals were moderate but a steady 12-14 kt wind from west-north-west made for optimal ridge flying conditions along the steep slopes of the area. As the name Regatta Cup implies, the race was to start at a set time for everyone with all 26 gliders trying to cross the start line below 1,800m simultaneously and as close to Vne as possible. From there they would all fly along the same course, round eight tightly-spaced turn points, then dash for the finish line.

Task set in Provence, France. With wind from WNW, the entire race could be flown in ridge lift.

The start was set away from any of the ridges so big gaggles formed underneath one of the few cumuli west of the Romans Saint Pau airfield.  Cloud base was around 2,200 meters so everyone circled up to the base of the clouds and tried to stay there in order to maximize the potential energy when the start gate would open. 26 gliders were sharing two thermals, all flying within an altitude band of approx. 50 meters. I was not surprised when I witnessed two pairs of gliders colliding with one another. In fact, I had several close calls myself – and all of that before the race even got underway.

Trying to stay aware of everyone around me I began to dive about 20 seconds before the start of the race. Burning excess energy using the spoilers and watching the altimeter, the speedometer, the GPS, and the traffic around me all at the same time, I managed to cross the start line about 5 seconds after it opened, 20 meters below the ceiling, flying just below Vne with a ground speed of 272 kph.  I considered this a very good start even though about half of the competitors were already ahead of me and the other half only seconds (or fractions thereof) behind.  And, luckily, I was still alive.

With the wind at the tail and flying at a “conservative” air speed of 160-170 kph the altitude at the start was just sufficient to get the unballasted LS4 to TP1.  I was in the bottom third of the pack but the leaders were less than a minute ahead.  From there the ridge race began. The strongest lift tends to be near the top of the ridges and not more than one or two wingspans away from the terrain.  That means everyone will attempt to fly in that narrow zone between winning and dying, and with so many gliders packed into the same tight spot at the same time, surviving is not much more than a game of Russian roulette.

Added to this is the complexity of different climb rates based on the angle of the ridge line with respect to the wind.  You look ahead trying to anticipate where the best climbs are likely to be.  Just before you get there you pull up sharply in order to fly two or three seconds longer through the best climb zone, then you push the nose down again, even more so if you anticipate an area of sink. Everyone else tries to do the same thing: flying as close to the ridge as possible, pulling up right before the best climbs, pushing down right before any anticipated sink. In doing so the speed of the glider might vary anywhere between minimum sink speed and three times as fast. Each time the altitude fluctuates by 200-300 meters as speed gets converted into height, or height gets converted into speed.  Pulling up or pushing down too early or too late, or incorrectly judging the strength of the lift, costs precious seconds that add up and ultimately decide about your placement. Flying like this is more than risky enough if you are the only one around but doing so in the midst of a pack of more than 20 gliders is simply an enormous gamble.

After several close calls between TP1 and TP4 and being about two minutes behind the leaders at this time, I decided not to follow the gaggle in front of me on the direct route to TP 5 but to take a slightly longer route along the higher and steeper ridges further to the east, hoping not only for better climb rates along this route but for some stress relieve as well.

Soon I discovered that I faced another challenge: there was a mountain pass in front of me.  I didn’t want to waste any time turning so I hugged the mountainside and flew at minimum sink speed hoping that the lift would be strong enough to carry me over the pass before I got there.  I was aware that this was a hugely risky maneuver: if a wind gust would force the glider to stall I would not have enough altitude to recover before hitting the ground. And if a thermal would break off on the valley side and turn the glider towards the slope, I could easily get pushed into the trees. “There is no way I would fly like this in real life,” I thought.

The gamble paid off and I made it across the pass and dove for TP 5.  As I got there I noticed that I had caught up with the pack.  I could still see several gliders ahead of me but they seemed a few hundred meters lower.  With a valley to cross ahead, they would likely have to stop to climb somewhere while I could cruise along the top of the ridge at a much higher speed. I dove across the valley at over 250kph and still reached the ridge on the other side at a good altitude.  Again, I hugged the higher ridge to the east while the handful of gliders ahead of me were lower and further west.

There was another valley to cross between TP 6 and TP7.  Again I put the nose down only to realize shortly thereafter that I would arrive too low on the other side. I dialed the speed back to 140kph – my slowest cruising speed of the entire race – to conserve what altitude I had left.  As I headed straight towards TP 7 the trees on the slope ahead of me were getting closer and closer.  Would I be able to get to the turn cylinder before the trees would get me? I wasn’t sure but just as I was forced to initiate a turn right over the tree tops the GPS confirmed that I had rounded the TP.  If I had only been 20 meters lower I would have had to turn away from the slope and find a spot to climb.

Having turned TP7 I could now hug the ridge again and fly toward TP8, the final turn point.  I could only see one glider ahead of me.  My final glide calculator indicated that I would arrive a few hundred meters too low but I was confident that I could easily make that up by hugging the ridge between TP 8 and the finish line. So I decided to dive, flying a direct route towards TP8. I kept watching the other glider to my right and though that I might have a chance to beat him.  I reached the next ridge closer to the valley with only 5 kilometers to go to TP8 and 20 kilometers to the finish when … I suddenly died.

Another glider that I had not noticed must have been slightly above or below me.  I had not seen him at all and I must believe that he had not seen me either. It was quite a shock and a revelation.  Obviously, had this been in real life I would not be here to write about this experience.  Instead, my wife and children would stand by my graveside and wonder, with tears in their eyes, how this could have happened.

Yes, I know, Condor is just a game.  I must and do believe that I would not have taken many of the risks described, if I had been in a real glider. Also, some of the race settings described here are not realistic. For starters, contest directors are unlikely to plan a task that is as dangerous as this one.

However, there are many things that are not so different from real life. The pressure of the competition, the desire to win, the fear of embarrassment.  Also the fact that taking high risks does have the potential of giving you an advantage in the race: circling in gaggles under the cloud base to conserve energy for the start, flying close to ridges without adequate safety margin, hugging the tree tops, scraping over mountain passes, circling at minimum speed close to the ground, flying low over unlandable terrain, aggressively calculating the final glide – all these are risks that have killed many real life pilots. I believe that this is especially true in race settings when usual risk-mitigation strategies get too often ignored or even willfully suspended.

BTW – in case you’re wondering: I did analyze my flight track and those of the eventual race winners.  I believe that I would have come second or third in that race – it would have been my best result against some of the world’s best Condor pilots.  You can see the race results here. However, it was definitely not worth dying for.


Boulder Wave Routes

The best soaring routes almost always correspond in one way or another to the terrain below, no matter what lift you use.

E.g., you would expect thermal lift over terrain that is most exposed to the sun (e.g. slopes that are most directly warmed by the sun based on the time of day); you would expect convergence lift where terrain features redirect the wind such that air masses collide with one another and are forced upwards; and you would expect ridge lift along long and steep slopes that are more or less perpendicular to the direction of the wind.  (It’s no surprise that pilots love to fly along the top of ridge lines where thermal, ridge, and convergence lift often come together.)

It’s no different with wave lift.  Wave lift forms when the wind pushes (relatively stable) air downward along the lee slope of a mountain, thereby warming it at the dry adiabatic lapse rate (such that it becomes warmer than the surrounding air near the ground). It will then rise again because it became lighter than the surrounding air mass, thereby starting a wave motion that oscillates on the back side of the mountain.  (You can find more details about wave lift here.)

Wave lift will form only if the wind is relatively strong.  In most locations, such strong winds usually come from the same direction.  In Boulder that is from the west – especially in the wintertime when the jet stream blows at our latitudes. What makes Boulder a particularly great wave location is the fact that a tall, nearby, mountain range – the Colorado Front Range – is conveniently laid out in north-south direction (hence the prevailing wind has to cross it at a perpendicular angle) and the Boulder airport is just to the east in the lee of the mountains.

With all that said, it should be no surprise to see that wave flights from Boulder tend to follow the same routes: parallel to the mountains on the lee side. In fact, the following chart depicts 40 wave flights from Boulder from 2010 and 2017 that were longer than 2 hours in duration and extended above 17,000 feet.

Source: OLC.  Wave flights from Boulder. Depicted are 40 flights from 2010-17, each more than two hours long and with a maximum altitude of at least 17,000 feet. All depicted flights were in December, January, or February, and all were flown in wave and rotor lift. (Note that good wave flights can also be flown in other months, especially November, March, and April.)

If you study the flight logs a bit, you quickly notice that the traces tend to be parallel to the curving ridge line.  The distance of each trace to the mountains depends on two things:  (1) the wave length on the particular day (it can be longer or shorter depending on the strength of the wind and the stability profile of the air); and (2) in which wave bar the pilot was flying (e.g. the primary, secondary, or tertiary wave).  The primary wave is the one closest to the mountains; it usually (though not always) provides the strongest and highest lift. As the name implies, the secondary is the second wave bar behind the mountains, the tertiary the third, and so forth.

Take a look at the red trace that extends furthest to the west – it is the only one in this set that crosses the Continental Divide.  This flight was flown by Al Ossorio on Dec 29, 2010 in the club’s DG505 and reached more than 27,000 feet within the designated wave window (Arapaho Peaks Soaring Area).  However, note that the high point was not over the Continental Divide; it was several miles further east, just where the red trace blends with all the other traces – the typical location of the primary wave.

Also quite interesting are the two greenish traces that extend furthest to the north. Both were flown much more recently by Bob Faris on two subsequent days in December 2017 (Dec 1 and and Dec 2) in his DG800.  Both flights reached altitudes of just under 18,000 feet. During the more yellowish of the two, Bob got above 17,000 feet only on the outbound leg (following a fairly straight line parallel to the mountains).  He then lost the wave near the Wyoming border and had to fly the return leg at much lower altitudes between 9,000 and 12,000 feet mostly in thermal lift (a very warm day in December!). During the more greenish of the two traces, Bob stayed in wave above 16,000 feet almost the entire time and actually flew back and forth along the mountains three times, covering 617 kilometers at the remarkable average speed of 174 kph (108 mph).

Where Are the Ghost Gliders? or: How Dangerous is Soaring Really?

The tragic death of Tomas Reich during the last day of the most recent Sailplane Grand Prix final in Santiago de Chile and the ensuing debate about the safety in soaring competitions brought – once again – a key question to the forefront of my mind: How Dangerous is Soaring Really?

When I started soaring in 1983 at the age of 16, I often heard people say that “the most dangerous aspect of gliding is the drive to the airport”. Intuitively this never felt right to me and several people have since pointed out that it is indeed far from the truth. (See, e.g., the speech Safety Comes First, delivered by Bruno Gantenbrink).

But just how dangerous is it? To get a better sense we need a reference point. I believe the best way to think about the dangers of soaring is to compare it to the dangers of other relatively dangerous activities we might indulge in: e.g. we could go on a road trip, ride a bike, or ride a motorcycle.  And I think the best way to make such a comparison is on the basis of participation hours (rather than on the basis of miles traveled for example). E.g., when we have an afternoon to spend we may want to know: is it more dangerous to spend that time riding our bike or to go fly our glider? We have all seen the white-painted “ghost bikes” on the side of the road marking the spots where a cyclist was killed but we haven’t seen any “ghost gliders”.  However, we would be kidding ourselves if we thought that gliding was somehow less dangerous. (Spoiler Alert: the comparison is not even close.)

Unfortunately, good, global statistics about the dangers of soaring are hard to come by. In most countries, a comprehensive and reliable database of gliding accidents does not exist. Nor is there a reliable global record of the number of flights or hours flown that would provide a good reference point.

However, while the available data is not globally comprehensive, there is enough out there to draw these comparisons – at least directionally.

My analysis of gliding accidents is based on data from Germany: the German government keeps meticulous track of all flights and even separates out glider flights and flights in motor gliders. It also maintains a database of all flight accidents and reports on an annual basis the the number of fatalities, and the number of persons injured.  Now, one might think that using German data is rather limiting.  But that is not quite true because gliding is much more popular in Germany than elsewhere.  In fact, according to a report presented to the International Gliding Commission in 2010, Germany accounts for approx. one third of all glider flights worldwide.  If there is a limitation to using German data, it might be that it actually underestimates the dangers of soaring elsewhere simply because Germany has such a particularly well developed soaring and safety culture. But, since I can’t prove that, let’s assume the German stats do a fair job of representing the dangers of soaring in general.

So here is what I found.  The result is – unfortunately – rather sobering.

On average, soaring pilots have an accident every 10,000 flights (this is based on all flights in Germany from 2002 through 2016 – the exact number is 10,070). Fortunately some of these accidents only damage the glider or some other property. But once every 60,000 flights someone (usually the pilot and/or passenger) is seriously injured, and once in 83,000 flights the pilot and/or passenger dies.

If you consider that the average glider flight takes about 38 minutes (arguably my least generalizable assumptions since it is simply based on the flightlog of all club flights of members at the Soaring Club in Boulder between 2002 and 2017) this means that soaring pilots can expect to get seriously injured every 40,000 hours and die every 50,000 hours.

Wow! Fortunately we do also other things in life because these stats mean that we would die every 6 years if we did nothing else but fly gliders!

So how does this compare to other activities? Well, not favorably to say the least.  On a “per hour” basis, gliding is about 35x more dangerous than driving; 70x more dangerous than riding a bike, and still 3x more dangerous than riding a motorcycle.

Risk of dying per hours of engaging in a particular activity. Note that the comparison is directional because the data for the various activities are from governments in different parts of the world. (Gliding is based on German data, bicycling and motorcycling are based on UK data, and driving is based on data from the AAA in the United States).

Another way to look at this is to say that 1 hour of gliding is about as dangerous as going on a 35 hour road trip in a car, e.g. from Denver to San Francisco and back again. Or as dangerous as riding a bicycle from Denver all the way to Minneapolis (70 hours). Or as dangerous as riding a motorcycle from Boulder to Salida (3 hours).

Is this an acceptable risk to take? I think that is a question we all have to answer for ourselves. But the important thing is that we should all ask that question and think hard about what we can do to minimize the risk in our own flying decisions.  And no one should kid themselves into believing that those stats don’t apply to them because they are simply a better pilot.  (Instead, they should remind themselves that it’s often the best pilots, like Tomas Reich, who make up the sad statistic.)

With sincere condolences to the family and friends of Tomas Reich.