“(the reviewers)… are still unconvinced by these controversial claims.” A science story.

Alternate titles, choose one or all:

 1) The story of APIPs (Aircraft-Produced Ice Particles)

 2) They said it couldn’t be done, but they did it anyway

 3) ‘An embarrassment for the airborne research community’–Dr. John Hallett, 2008

OK, “baby I’m bored” with the lack of clouds and precip,  and so I thought I would share my boredom with this long tome on aircraft effects on clouds.  Why not bore other people if you’re bored?  I’ve thrown in some alternate titles above to peak and pique your interest.  Speaking of “thrown”,  Mr. Cloud-maven person was also thrown off his big (I mean huge1),  young horse lately; “JohnT”, as he is named, doesn’t like people to sit on top of him sometimes.  Not easy to sit at a computer these days, hence the lack of “acitvity.”

OK, on to the story of APIPs.  The title quote was written in 1982 by the Chief Editor of the American Meteorological Society’s, J. of Climate and Applied Meteorology (JCAM) summing up the opinions of the three reviewers at the bottom of a second rejection notice of a manuscript, one that had been fluffed up with more evidence of APIPs.   However, the Editor allowed us (me and Peter Hobbs, director of the Cloud and Aerosol Research Group at the University of Washington) another crack at it, and by the THIRD submission (requiring a bit of chutzpah),  a colleague and me had found photographic evidence of aircraft having produced icy canals in supercooled clouds, and that visual evidence really pushed our third manuscript, now as big as JohnT, over the top in getting accepted and  published.

The phenomenon came up again last summer in a Wall Street Journal article, one in which Mr. Cloud-maven person was asked his opinion.  This phenomenon (APIPs) is attracting more attention these days, so I thought I would pass this background story along.  I hope will encourage authors with rejected manuscripts, which I myself have quite a few.   You might have something really good.

Yes, that’s right, lucky Mr. Cloud-maven person was involved in this interesting chapter of science that happened way back in the early 1980s when he was part of the flight crew in the University of Washington’s Cloud and Aerosol Research Group (CARG).  Occasionally, and mostly in studies of ice development in Cumulus clouds, I got to direct the University of Washington’s first research aircraft, a 1939 manufactured, Douglas B-23, into Cumulus and small Cumulonimbus clouds.  It was heaven for me, a storm chaser type person, having done that here in AZ way back in the mid-1960s chasing summer thunderstorms all over the State with my camera and rain gauge.

We had a viewing dome on the top of the fuselage of that B-23 and I sat in a swivel chair, head protruding into the “bubble.”   I was kiddingly referred to as the “bubblehead.”  I think they were kidding, anyway…  Those who know me will understand that title.   Sitting there with head in the bubble, allowed me to see EXACTLY where we exited a cloud and could direct the pilot to EXACTLY that same cloud blob we had just exited.  The pilot was fond of turning the plane sidewise for this return so that one wing was pointed straight down in the turns and we often got back in within 90 s to two minutes.  It was an exciting as well as sickening experience.

We did that because we wanted to see how that element of the cloud had changed with time.  Did ice form?  Did the drops get bigger or smaller?

This viewing dome gave us a huge advantage over other research aircraft doing this kind of research.  Below, that B-23 aircraft sitting on the tarmac at Boeing Field, Seattle2.  The second photo is a view from the “bubble” located toward the rear of the fuselage.  Nice!  I was so lucky!

One day, while looking over our Brush strip charts from the flights, I noticed some odd spikes in the ice crystal detector we had. Also, since we were one of the first groups to get a probe that produced shadows of the particles in the clouds as we flew in them, I was able to see that the particles producing those spikes were oddly similar sized, as though they had formed simultaneously, something not seen so much in natural clouds. Pretty soon it became apparent that these spikes and odd particles ONLY appeared after we had gone through the same cloud for the second or third time.

I remember walking into Professor Peter Hobbs grand office with a strip chart with those ice spikes and saying, “I think our aircraft did this.”

He was unfazed; did not have a particular reaction.  Peter Hobbs was always open to new thoughts, and that helped allow me to go forward with a further investigation even if it meant some of our past data and publications might conceivably be compromised, ones however, I was not involved with as a fairly new (5-years in) employee at the U of WA.  No vested interests here!

After awhile, after aircraft plots showed that the spikes were within tens to a couple of hundred yards (meters) of where we had been before in a Cumulus cloud, a very short paper was written up on it and submitted to JCAM in late 1981.  It was quickly rejected.

Ours was a highly controversial finding due to both the high concentrations of ice that we found (hundreds  to over a 1,000 per liter) but most of all due to the temperatures at which we were reporting this effect, -8 to -12 C.   Our plane was,  in essence,  seeding these clouds with ice crystals, changing their structure.  Since the volume affected was initially quite small, it was likely that only having the viewing dome allowed us to find them on the second and third penetrations of the clouds.

This inadvertent aircraft effect had even been looked for by our aircraft group leader, Dr. Prof. Lawrence F. Radke before I had arrived and after the University of Washington acquired the B-23.  He didn’t find’em though.  Larry was also aware that an aircraft COULD do this in those early days with the B-23.

So, when I found them and a paper began taking shape, the skeptical Larry Radke called them,  “Art-PIPs.”  It was so funny.

Later, with the skeptical Larry at the helm, we got some money from the NSF to try to produce them in various clouds, and sure enough, we did.  It was amazing finding those crystals in those test flights since even I couldn’t be absolutely positive sure that this was real.  Why hadn’t this phenomenon been reported decades ago?   That, too, was part of our problem:  why you, why now?  And why hadn’t I seen the holes and canals of ice produced by aircraft as a cloud photographer for decades even by then?

Some ground observers had seen trails and holes in “supercooled” clouds like Altocumulus.   Those holes and canals were occasionally reported over the decades (!), but not in the technical journals.  A couple of really lucky observers had even seen the type of aircraft that had caused them.  But the airborone research community, ignored or did not know about these reports, ones that appeared in non-technical weather magazines like Weatherwise, Weather, and Meteorological Magazine (the latter two in England).

Furthermore temperature data were nearly always absent in these visual reports.  So, it could be reasoned they had occurred at very low temperatures, below -25 C or -30 C.  Clouds that cold, but still consisting of only or mostly of liquid droplets do occur, the ones in which an aircraft could leave an “ice canal” or a “hole” with ice in the center, falling slowly out.

If we had been reporting our finding at cloud temperatures of -25 to -30 C, maybe we’d have got into the journal on the first try and reviewers would have yawned.  But at -8 to -10 C cloud temperatures?  No way!


Research aircraft had been going back and sampling the same cloud, usually a Cumulus one,  for a couple of decades by the time of our report.   Furthermore, those aircraft re-penetrations were almost always in the same temperature domain that we were reporting this effect, to about -5 t0 about -15C.  And one of the main findings in those early days of aircraft sampling was that nature was producing far more ice in clouds than could be accounted for in measurements of ice nuclei, particles on which ice can form.  Concentrations of ice nuclei were largely determined from small cloud chamber measurements made on the ground.

These early cases of high ice concentrations in clouds with tops that were not very far below freezing (greater than -15 C) were called cases of “ice multiplication” or “ice enhancement.”   No one understood how such ice developed and many theories were put forward initially in the 1960s.  The issue was largely explained by the “Hallett-Mossop riming and splintering mechanism”, a mechanism discovered in the mid-1970s and today is still believed to be the primary reason for high concentrations of ice crystals in clouds with tops warmer than about -15 C.  Oh, yeah, ice multiplication is real and NOT due to aircraft penetrations!

But our paper on APIPs, if true and published,  would cause researchers to have to go back and look at their research data (even us!) and investigate whether their own aircraft had contaminated their published studies with artifact ice crystals.  An entire body of airborne literature would come under question.  This was not a pleasant thought for anyone who had  conducted such studies.

Why would you go back and sample the same cloud?

To see how it changed with time.   How many ice crystals formed as time went by?  Where, and when?  These were techniques used in trying to get to the bottom of the “ice multiplication” phenomenon.  In fact, the Chief Editor of JCAM himself was involved with numerous aircraft that sampled clouds in a huge summer Cumulus cloud study program in Montana in those days (called “CCOPE”-Cooperative Convective Precipitation Experiment)  That study, like so many other airborne studies, was to determine how ice onsets in clouds, how high the concentrations of natural crystals were in clouds with various cloud top temperatures, and the potential of cloud seeding to increase rain.

While academic scientists did not particularly welcome these reports and were dubious and largely ignored them (did not change their aircraft sampling strategies), or when they looked could hardly find any APIPs, it was soon evident that purveyors of cloud seeding services were elated!   Our finding suggested to THEM that all that natural ice formation reported in re-penetrated clouds  in research articles over the years might be wrong, and rather due to ice produce by the aircraft!  Maybe those clouds that had been reported with a lot of natural ice, which made them unsuitable for seeding, was because the researcher’s aircraft had produced it, not nature.  Purveyors of seeding would like clouds that are below freezing, about -5 C and colder, with no ice in them.  If the concentrations of natural ice crystals forming in clouds ice get to 10s to 100s per liter,  those clouds are deemed unsuitable for seeding to add more ice.  The crystals might be too small if you add more in those cases, and not fall out.  If surveys of clouds in a region find that they have lots of ice in them, its “no paycheck” for commercial cloud seeding interests. (Usually, cloud surveys aren’t done before commercial programs begin.)

Thus, those who had interests in cloud seeding actually saw our result as a way to discredit findings of high natural ice concentrations in clouds, findings that made them appear unsuitable for seeding.  It was a bogus argument since numerous FIRST penetrations of clouds had encountered high ice particle concentrations, still, they had SOMETHING to hang a hat on.

This was indeed an ironic twist, being supported by the cloud seeding community!

Me, usually with Peter Hobbs as a co-author, had been discrediting various published cloud seeding results in the literature via reanalyses and journal commentaries for several years (e.g., here) when our APIPs finding finally hit the “streets” in 1983.

Given these a a priori possible biases between academia and in the commercial cloud seeding world in detecting APIPs you can imagine where the major “confirmatory” studies of this phenomenon came from. Yep, those associated with cloud seeding programs!  It took 8 years (1991) for our finding to be independently confirmed (the best way) using several types of aircraft in marginally supercooled clouds.   Then pretty much the same workers amplified their findings with another paper in 2003 WOODLEY et al. 2003.  For those of you who don’t know the cloud seeding literature, Woodley and Rosenfeld and Peter and I have had a major clash in the cloud seeding literature (i. e. and big i. e., and bigger still)

We loved it!  They loved it!  Even the great John Hallett got involved and found in lab experiments that the mechanism was the extraoardinary cooling at the prop tips, momentarily down to -40 C, a temperature at which ice forms spontaneously in high concentrations (here).  It had also been suggested that prop aircraft could do this by the late Bernard Vonnegut back in the late 1940s in a less widely distributed report from a General Electric research lab and in the J. Applied Physics.

Today this phenomenon is taken pretty much for granted, and has been more widely detected from time to time in satellite imagery in thin clouds as here.  In thicker clouds, the effects of aircraft go largely undetected.  Recently, in a widely distributed news release that accompanied their formal publication, Heymsfield et al reported a case in Colorado in which aircraft-produced ice effected a snow shower on the ground instead of just being a hole or canal in some thin clouds as we normally see.  They opined that aircraft could actually help delay flights from the airports that they were taking off from or landing at in special conditions.  (That’s what the Wall Street Journal article was about.)

Why was it an “embarrassment” to the airborne research community, as John Hallett (of Hallett-Mossop) asserted?  Because they should have found out about APIPs right from the get go, especially in view of the occasional lay publications that had photographs of ice canals in supercooled clouds even in the 1940s, ones  that could only have been produced by aircraft.  It turned out to be a major oversight.

Below, a cartoon I did before the paper was accepted making fun of how a researcher, thinking that natural ice multiplication processes were taking place (i.e., the Hallett-Mossop riming splintering mechanism) might overlook all those ice crystals streaming off, in this case, the Husky 1 aircraft.

Below, some photographic evidence of what aircraft can do to supercooled clouds, the last one taken about two weeks ago over the Cat Mountains.

Finally!  The End.










1The 6-year old horse in question is about 15 hands, 1200 lbs, not really a Clydesdale.  I have overemphasized our horse’s size for personal reasons.   You don’t want to be injured getting bucked off a Shetland pony, but rather something HUGE!  It just sounds better.

2That B-23 aircraft, wherever it went, brought a crowd out to see this antique “tail-dragger.”