Dr. Matt Nolan

Institute of Northern Engineering
University of Alaska Fairbanks

 

Home

Curriculum Vitae

McCall Glacier

Photography

Older Projects

Personal

Contact

 

 

Matt Makes a Map

Matt Nolan
7 September 13


Almost exactly a year ago (5 September 12) I took my first flight lesson.  Since that time, I got my private pilot license, bought a plane, flew it from Texas to Fairbanks, got my commercial license, modified the plane for vertical photography, built a low-cost mapping system that was simple enough for me to operate while flying on my own, got waivers on existing ethics laws to allow me as a State employee to hire myself as a subcontractor to fly mapping missions, succeeded at acquiring data over some of the most remote locations in Alaska, and processed those data into topographic maps of unprecedented resolution and accuracy.  This blog tells that tale. (To read the slightly more professional version of this story that focuses on the scientific results, visit here).

Matt buys a camera

My interest in photography began in 2003, when Austin Post mailed me some negatives of photos he took on McCall Glacier in 1958.  These arrived just days before my first August field trip to the glacier, a project I have now led for 11 years.  I remember holding those negatives up to my window in the late hours of the night before we left, searching for any that we could possibly repeat on that trip.  I found one of the terminus that was particularly striking, and quickly scanned it and printed it out so that we could relocate the photo position and repeat it.  Late one afternoon while out there, Kristin and I were in the lower ablation area and the weather was good.  I knew it would make for a long night to find it, but Kristin encouraged me as this was likely our last opportunity of the trip to do it.  So we hiked down the rest of the way and I scrambled up a number of side valleys, not realizing at the time that though I was about the right distance from the current terminus that of course I needed to be the right distance from the 1958 terminus, which was substantially further back.  By now it was approaching midnight, but the arctic sun was still up, so we hiked further down valley.  I gave it one last shot, hiking up the large loose rocks towards what I thought might be the place. Thankfully it was, and that photo pair has become somewhat iconic and widely reproduced.  The next year I tracked down some photos of Okpilak Glacier from 1907, and repeated those, again with my tiny 4 megapixel point-and-shoot camera, as well as others in the McCall area.  The feeling of finding those photo sites and standing in the footsteps of these prior photographers was euphoric, and ultimately motivated my interest in high-end scientific photography of all sorts. 


McCall Glacier in 1958


McCall Glacier in 2003

The next piece of the puzzle began in 2005 and came to a head in 2008.  Until Turner was born, we had always accessed the glacier by helicopter.  We hardly gave it a second thought that Turner would come with us once he was born, but the powers that be had other ideas and in a somewhat tense scene the day before our May 2006 trip, when Turner was about 8 months old, we got shut down from taking him on the helicopter and it was simply too late in the game to organize something else.  Over the summer we organized a new plan to circumvent such powers by flying in to the tundra via a fixed wing charter that we could pay for ourselves and then hike to the glacier.  You can read about these adventures here.  Hiking in gave me a new found impression of what the tundra was all about and how glaciers relate to it, and again ultimately motivated my subsequent projects in ecology and glacier-ecology interactions there.  By 2007, we were old hands at hiking in and during that intervening time I had read incessantly on high-end photography and bought some high-end gear.  My primary interest at that time was high-resolution 360 spherical panoramas and gigapixel photography, as a means to document the landscape by creating baselines for future repeat photography, searching for simple but valuable ways to put cameras to use for science.  For our August 2007 hike, we were joined by an ecologist and we combined ecology and photography in a variety of ways.  That 2007 trip also marked one of the biggest problems we had with helicopters, which for the past 3 years we had been using to transport our gear while we hiked in. Given that Turner was rapidly approaching the stage where we could no longer carry him into the field, I saw the writing on the wall that a new logistical approach was needed.  This came to a head in 2008, when we spent the summer on the glacier, and we had no end to air taxis bailing out on us and it was clear to me that we need to take logistics under our own control by operating our own aircraft. 


Our field work in the eastern Brooks Range with Turner wasn't always a walk in the park like this, sometimes it actually got gnarly.


Click here to zoom into this spherical panorama, taken in May 2010 while doing some shallow ice coring. Find tons more like this here.


Click here to zoom in on this gigapixel panorama, taken on McCall Glacier on Mother's Day 2008 during our deep drilling operations.

So that winter we bought a plane.  Kristin had been working as a commercial pilot for the past 5 years, and the idea was that we would move to Kaktovik, about 60 miles from the glacier, and run a small air taxi using that plane that would also support our research when needed but cover the costs of the plane through other work.  This plan seemed good on paper, but the results were mixed.  Besides damaging the plane on our first trip, and not getting it back until fall of the following year (another huge saga), the complications associated with supporting other projects to fund our own led to scheduling conflicts.  We were however able to support all of our research, including not only logistical access to our arctic field sites but also creating an oblique glacier photo inventory of nearly all of the glaciers in the area and some preliminary work with vertical photography.  So in principle though the method of operating our own aircraft was sound, it was accomplished at high cost to ourselves both financially and emotionally.


If only he could have stayed a 1 year old forever! Or at least visit once in a while.


By age three, Turner was over 40 pounds and just too much for his old parents to carry. By age six, here, he was almost able to carry us.


So we bought a Helio Courier in California to help support our research, and save our backs. You can read the full story here.

Since I became a professor in 1999, I had always had a strong interest in topographic mapping. I was the first in the state to make use of commercial airborne InSAR, a sensor that flies on a lear jet at 30,000 and can see through clouds. I mapped large areas of the State this way, including the entire Kuparuk and Putuliguyak River watersheds on the north slope and about a dozen mountain passes in Alaska that are useful to avaition, and I spent a lot of time advocating for new mapping missions in Alaska and the Arctic. Dealing with such large, high-resolution data sets and sharing them online became a challenge that resulted in EarthSLOT (note that I gave up the url a while ago and what you find there now has nothing to do with me), an online, global terrain visualization service much like Google Earth, only better, built on the engine from Skyline Software. I built it before Google even had the idea for its Earth and was able to serve my high resolution data in just the same way, in a 3D virtual globe that anyone could access from the web. About all that's left of it are some stills and movies I made using it that you can find here and maybe some movies I made of the Iditarod trail on their web site. When Google Earth came out, I spent a lot of time building layers for it for the International Polar Year and as far as I know it was their most popular way for people to explore IPY. So by 2008 I was very much into airborne mapping and using high resolution topographic datasets, and I had gotten major funding from NSF to make new lidar maps of nearly all of the glaciers in the eastern Brooks Range.

Those beginnings were what motivated my interests in scientific aerial photography.  By 2010, we really didn’t have enough work to finanically support the aircraft based just on our logistics and running an air taxi ended up being too counter productive, but by combining my by-now strong interests in photography with airplanes, it seems we had found a niche whereby we could combine our professional interests and stay afloat financially.  And while oblique photos from the air and 360 and gigapixels from the ground were useful and interesting, their scientific utility was primarily qualitative. Though a bunch of my photos ended up hanging on the walls of the White House, taking them is not particularly fundable essentially because they dont seem scientific enough to reviewers. But to be fair, what I and most arctic scientists really want are time-series of high-quality topographic maps that quantify the change, though personally I dont see that this has to be an either-or.  By 2011 I had been reading up quite a lot on photogrammetry and the commercial firm I hired to acquire lidar of our glaciers still had not produced final results 3 years later, and I saw no reason why I couldn’t do this myself and probably do it better and cheaper than other means.  So while the repairs were being made to the landing gear from the May 2009 accident, we added some large vertical camera ports to the plane that could accommodate large lidar or other mapping systems.  In fall of 2010 we were able to complete two vertical mapping projects, one in the Brooks Range for the Park Service and another on the Seward Peninsula as part of a geothermal study.  Though the end products were just high-quality photos and not products derived from them, with those proof-of-concept missions completed it seemed that we were now on track such that I could begin seeking funding for more serious airborne mapping projects.  That winter I secured such projects and it was time to get serious.


Kristin flies the plane, I gage the streams, and Turner swats mosquitos, along the Hulahula River in 2009.


Would you buy air photos from these two?

My initial interest in airborne mapping was primarily in lidar.  Like everything these days, digital just seemed to be the way to go, and photogrammetry had something of a stigma as ancient technology.  But lidar units are really expensive.  I worked hard on various schemes to buy or co-own units, but these efforts came to fucknaughting.  But both due to expense and personal preferences, I came to the realization that I would prefer to make topographic maps using photos – there was just something about photography that I found inherently compelling, and trying to look objectively at how I spent my time I realized that given a wide variety of options, I always seemed to find myself sitting at my computer processing photos.  So I came up with a compromise solution.  I bought a single-beam lidar intended for industrial use, adapted it for airborne use, attached it to a high-end IMU and GPS, sorted it all out into a functional airborne system including flight management, and combined this with a camera.  The idea was that if the photogrammetry didn’t work out, at least I’d have a laser track, but at best the laser track could aid in the photogrammetrical solution.  This system was novel in that my lidar unit had very high range so that we could fly above mountain terrain and still hit the glacier, something only the $500k-$1M lidar were capable of, and similarly photography had no range limitations.  Because it was all home-made, it was however a complicated system to keep functional and basically required my full-time effort while airborne.  This effort took up the bulk of my year, but by fall we had a few of my own projects done and in 4 day whirlwind we flew a couple of large missions in the Noatak and Kobuk valleys for the Park Service, all the while never really sure what the quality of the results would be.  We acquired 4 of the 5 large blocks we had committed to, but decided to hold off on the 5th, on the Seward Peninsula near Shishmaref, as the weather was changing, it was a big effort to get out there, and I wasn’t really sure yet whether the system was operating as well as I had hoped.  As it turned out, there was a glitch with the GPS system on some of those flights such that it was still working but the acquisition rate had reset to a lower rate, such that I was not able to do direct comparisons between my lidar and photos or position the photos as well as I had hoped.  But that winter I invested a lot of time with digital photogrammetric processing, trying out a lot of different software packages, learning the ropes, and getting to the point where it was clear that there was a lot of promise and that what we had acquired was good and useful.  That is, I was making beautiful DEMs and orthoimages that looked realistic, I just didn’t have anything to compare against to determine whether I was just making pretty pictures or actually creating a useful scientific product.


One of my first test maps -- a pingo in the Noatak River valley, from 2011.


An oblique view of the same pingo. The general idea is that by mapping them now at high resolution, we will be able to detect changes to them in the future.


Here permafrost is slumping into a lake, in the Noatak Valley in 2011. As ice buried under the surface begins to melt, the ground above settles and slides.


That same lake, with colorized height and contour lines.

While we had made a lot of pretty pictures and maps like this, at this point we still didnt know how accurate they were. By this time we had ventured into air taxis again as a means to stay afloat financially.  The new twist here was that rather than us supporting other projects, that we would lease the plane to an air taxi such that they could support other projects using our plane essentially as a backup, but when it came time for our projects we would again have first priority use of it.  This never really came to fruition in 2011 as there were a number of FAA issues associated with the air taxi we had partnered with, but it looked like it be on track for 2012. So 2012 seemed like the big year when everything would fall into place.  The plane was functional, the mapping system was functional, the air taxi approach seemed feasible and moments from finalizing, and we were excited about the possibilities.  After our May glacier trip, we loaded up the plane and headed north to begin that adventure in early June  The plan was Kristin would spend a week in ground school to get on the air taxi certificate and when the weather was good we would fly a bunch of missions we had planned to test out the system.  As it turned out, we spent the first week there flying 1000s of miles of missions over a variety of remote targets of interest.  When it seemed like the ground school wasn’t going to happen that week for a variety of reasons, we decided to head home for a while and regroup, so that I could process the data and begin planning our more serious missions the following week, including revisiting the Noatak and Kobuk sites, getting the Seward Peninsula site we had skipped the year before, and of course doing the arctic glaciers which were my personal motivation behind all of this in the first place.  Unfortunately, the trip home did not go as expected, as the plane got totaled with us inside it and threw our life and plans into chaos for the next several months.

\
It was a busy 10 days, June 2012.


And in the middle of it we spent a few days on the tundra stream gaging.


Unfortunately we made an unplanned trip past Denali on the way home...


... and had a one week vacation in a great hospital.

By early August, 2012, Turner and I were recovered well enough that we all went out to the glacier, got the bulk of our work done, and had a really nice time.  During this time, we had largely not discussed our future in terms of airborne mapping.  The accident also led to some conflicts with the air taxi, which left us screwed financially once again.  So by late summer, we were without a plane, without a plan, and with a lot of unfinished work we were on the hook to complete.

Matt makes a new plan

Though I’ve spent countless hours in a small aircraft going to remote locations throughout the world, I never had much interest in being a pilot.  I think a big part of that was that I was afraid I might like it too much.  For whatever reasons, I tend to gravitate towards the frontier, pushing the envelope of whatever interest I might have.  For the most part, there is rarely any possibility of personal harm in these endeavors, the risk is largely frying some piece of expensive electronics, or wasting a few months of my life on a dead-end, or similar.  With flying, it was clear the case was different in terms of potential outcomes.  However, from my experience working with a lot of different pilots, I came to believe that the essential ingredients of a good bush pilot are quite similar to those of a good scientist, it takes a mastery of understanding the envelope to know where to push and where not, it’s about asking the right questions, it’s about validating your own answers, it’s about having backup plans so there is no doubt about the success of the outcome, and similar.  I don’t think those things were really on my mind or motivating me, I think having been a passenger for so long mostly I felt the need to understand what Kristin and other pilots were going through in a more visceral way that only comes with being at the controls.  So in any case, with no particularly strong motivation or need for success, I decided to take a flight lesson in early September On September 5th, I walked into one of the local flight schools and asked to take their introductory flight lesson for $89.  It lasted about half an hour and I got to more of less take off, fly around, and land again, or at least pretend to.  It was a nice day and it went well enough.  So I signed up for a real lesson for the next day.  That day it air was noticeably rougher and they called to postpone the lesson, I think because they didn’t want me to get scared off by the bumps so early.  I replied that bumps were exactly what I wanted, as I wanted to feel what it was like at its worst.  I think other than the reasons I described above, my biggest concern about being a pilot is that I’m prone to air sickness.  But after an hour of getting bounced around while behind the controls, the conventional wisdom I had often heard from pilots seemed true – that you don’t get air sick if you are behind the controls.  So I kept at it and signed up for more lessons.  Initially I didn’t have any strong professional motivation, I think part of it was the realization that in several years of aircraft ownership, I don’t think we went flying for fun more than an hour or two, it was always a fairly intense and stressful under-the-gun feeling of accomplishing something, and here was a chance to fly more or less just for fun and try to understand why pilots are so addicted to flying. 


My first solo landing. So far so good.


I think it was more stressful for others than for me.

After a few weeks of steady effort I began to realize that flying and airborne geophysics was something I could potentially do professionally, even if just as a backup to Kristin, and a plan to do so began to emerge.  One of the biggest sources of stress with our first attempt at airplane ownership was that we owned only one.  The issue is that there are an unfathomable number of things that can go wrong with an airplane, and unlike a car you really cant defer maintenance on most of these things, any one of which can take from days to weeks to months to fix.  Combine this with short weather windows, remote operations, funding deadlines, etc, and it’s a recipe for ulcers and bankruptcy.  So I had decided if we were to do this again, we would need two planes, such that it was more likely that at any given time at least one of them would be operational, and preferably planes just like the ones everyone else owned so that they would be that much easier to fix,  That thought meshed with the idea that perhaps I would buy a smaller, cheaper plane that I could build time and experience in and that would largely be the backup plane, and that we would have a larger more capable plane for operational work.  In the meantime it had also become clear that partnering with air taxis was simply more trouble than it was worth, both economically and politically, and that if Attempt #3 was going to succeed we would need to be in direct control of our destiny.  Until now, we had been flying under the radar of State and Federal ethics laws by not actually owning the plane to avoid conflicts of financial interest, but this puts us and our friends in awkward positions that both air taxis we worked with found convenient to try to exploit by ripping us off for enormous sums of money.  So it seemed time to try to go in through the front door and get approval for us to own the plane and company that I would subcontract to as a professor.  And breathing down my neck was my biggest constraint – funding for the NPS mapping project that we had started in 2011 was going to run out in July of next year, so I had less than a year to make all this happen, and I still didn’t even know if my mapping system was up to the task or not, or if I could simplify it enough to run myself given my inexperience with flying.  So I began the process of plane hunting, legal waivers, and designing new technology, while continuing to train for my private license. And oh yeah, work a full time job.

Matt buys a plane and learns to fly it

The most common advice I got from experienced bush pilots was to buy an older Cessna 172 as a starter plane, or if I was serious about bush flying in the future then maybe a 170.  I had thought the combination of a 206 and cub to be the best for our needs, but I gradually started thinking that 3 planes might be the solution – each of which could back up the others in nearly every type of operations, just not as well as the primary plane, and hopefully 2 of the 3 would always be working so we could travel in pairs.  So I eventually added a 170 to the list, and this seemed to be the optimal combination of simplicity, performance, and cost for my needs as a nascent pilot.  But I promised myself that I would not buy the plane until I got my license and that I would buy it in Alaska to avoid the problems we had last time.

So I continued my lessons at Proflite while I kept tabs on planes for sale.  Lessons went well, and before long I had soloed, gone on cross-country trips, flown at night, etc.  It looked like by early November I would be ready to take the flight test with an examiner, and we scheduled the date for it.  In the meantime, I loosened my thoughts on buying in state and had my eye on a 170 in Arkansas that seemed to have all of the features I was looking for.  Unfortunately there was an issue with the plane I was training in and we had to postpone my flight test.  This dragged on for a while, and then we had problems rescheduling as 1 of the examiners (these are private people, not FAA types) was leaving town for the winter, another didn’t want to fly in the cold because there are some engine-out maneuvers in the test that might not lead to the engine coming back on, and the other was busy.  Unfortunately by this point I had already bought the plane in Arkansas. So, I scrambled to shift gears. 

My original plan when buying the plane was that the three of us would go down and fly it up together, and I figured even if I didn’t have my license at that point that Kristin could officially fly it.  But we had already missed one Christmas ferrying aircraft and Turner was committed to having Christmas here. It was now into December, and my next thought was to hire a ferry pilot that was also an instructor, such that we could fly it up together and I could log more instruction time towards an instrument rating and then be ready to take both the private and instrument test when I got back to Fairbanks.

I called maybe half a dozen ferry pilots advertising in the pilot magazines, but it turned out I went with the first one I called.  I described my plans and constraints to her, and though she was an instructor she told me about her mom and dad who ran a small flight school in west Texas.  Her dad had been a crop duster for a long time and was now semi-retired, spending his days doing flight instruction for the fun of it and occasionally some commercial work for a private owner, and was an expert in tailwheel aircraft like the 170.  Her mom was a designated examiner so could give me the test when I was ready.  I liked the idea, but was hesitant about it because you really cant predict what a cluster such a thing could turn into.  So I made plans to hire her as the ferry pilot, and in the meantime I found a large flight school in California which assured me that if I were motivated I could fly as much as I wanted in preparation for taking the test there.  So the plan was to take the test in California, fly commercially to Arkansas, and then fly the 170 to Fairbanks with the ferry pilot, using her parents’ operation as a backup for my test, all in the 2 weeks before Christmas.

Though I learned a few things at the flight school in California and enjoyed working with the instructor I had there, it was not the experience I was hoping for and it quickly became clear to me that I was not going to take the test there.  There are all kinds of different styles of instruction and training, and this place was at the far end of spectrum of ‘by the book’.  Maybe they need to be living in such a crowded and uptight area, but it was quite a shock coming from Fairbanks.  In the end I was only able to get 1-2 hours of flying a day, and this combined with their rigorous and highly structured standards for getting signed off to take the test made me realize that Plan B was a necessity after only a few days there.
Terry points to the nosewheel fairing where I, um, decapitated a Canadian Goose on landing.


They offered to give it to me, but I had just eaten.

Plan B turned out to be probably the best experience of my aviation training career.  Lawrence and Mary’s operations is based out of a small county airport where they have been for 30 or 40 years and where they rule the roost.  The atmosphere and philosophy here were also much more my style.  We had decided it best to practice for the test in a Piper Warrior as that’s what I had done all of my prior work in, rather than confuse the issue by learning fly a tailwheel plane, as mine had already been ferried to here.  I’m not sure if it was just the change of scenery of what, but I think it was only a few hours of practice with them over the next two days and I took the test and passed.  But now that the test was done, the fun really began. After I finished the test, Lawrence sent me out to taxi around in the 170.  It was humbling.  After having just passed this test and thinking I knew a thing or two, I could not keep the plane on the taxiways to save my life.  As I came to learn, you can essentially fly a warrior without ever touching the pedals, but a 170 requires continual, strong pedal adjustments to keep it tracking straight, on the ground or in the air.  And as I came to learn, Lawrence will fly all day long and all night long with anyone that shows up there, and he will work you for as long as you can stand it.  And that’s just what we did.


The new plane had already arrived in Texas

Lawrence and I got along really well.  His first plane was a 170 so there was clearly some nostalgia there.  I think also that they rarely get someone that wants to train in a tailwheel plane to land on glaciers and other odd places in Alaska.  But bush flying in Alaska and crop dusting have a lot in common, both in technicalities and in philosophies.  Lawrence and his cohort overlapped long enough with the real old-time cropdusters just like many of the older bush pilots here overlapped with the originals.  This somehow changes a person’s perspective, especially in regards to the continual addition of new misguided rules that are largely intended to keep idiots from acting like idiots, or more accurately to keep the idiots making the rules from getting grief from their boss.  It’s more than just scoffing at the rules, it has more to do with 40 years ago the FAA gave pilots much more freedom to make their own decisions and live or die by them, a freedom that once had can never be fully extracted. Now there is a pervasive attitude throughout our culture that anyone should be able to do anything regardless of skill or talent, and because that’s not accurate it necessitates the legal bar being lowered in fear of lawsuits or job-loss and perhaps worse that it inevitably evolves into the blind leading the blind.  It takes 1000s of hours and many years to master any skill, and flying is no exception.  But all it takes is 250 hours of flight experience and passing a few tests to become a flight instructor, and this is a primary pathway for would-be commercial pilots to build time.  Could anyone seriously expect to get the same quality of instruction from a 300 hour pilot compared to a 30,000 hour pilot?  But the fact is there are a lot more instructors with 300 hours, and this means there are a lot more private pilots who learned (or not) from such an instructor, and while this may boost the number of pilots it can only lead to a downward spiral of average quality, and the fatal blow is that as the blind rise to positions of power they assume everyone is as inept as they are and base their decisions accordingly.


This is spray-plane country!


That just about sums up my experience with Lawrence.

Lawrence doesn’t follow a textbook for his training.  As I came to learn, he has basically only one thing to teach, and that’s never to give up.  He never said that directly to me, but after thinking about the things we did and seeing so many other students get out of the plane in tears, that seems to be where he is coming from and after a year of flying I have to agree that there is nothing more important than never giving up.  I think that’s the primary reason we got along so well.  My first day in the 170 we flew probably 6-8 hours and in not one of the dozens of landings I made could I keep the plane on the runway myself.  We would be up on one wheel going sideways with the wind and momentum about to flip us over, and he’d be laughing at me because he knew exactly what I was doing wrong and I’d be laughing because I couldn’t figure it out.  But just before we would have bent some metal, he’d get the plane tracking straight and then fold up his arms, and tell me to try it again. And I would, again using more width than length of the runway.  I guess its part personality and part having been in so many desperate situations, but my mind just works better when there is a problem that requires my immediate and full attention, and any fear I feel about it usually comes only after the problem is solved. So it never really bothered me, and I seem to have the capacity to fly all day without much fatigue so I was able to learn a lot about what not to in a short amount of calendar time.

The second day started much like the first, with a lot of wandering between runway lights.  After a morning of this, he sighed and had me fly about 10 minutes away before telling me to land.  There was no airport or strip around us as far as I could tell, but he pointed a bit more and sure enough there was what looked like a driveway with a windsock and house at the end of it.  So I circled around and made the approach, lining up with the asphalt that could not have been more than 2 feet wider than my wheelbase, with ditches on either side of it.  After demonstrating so clearly my difficulties staying on a 200’ wide runway, this struck me as being an odd choice for practicing, but I lined up, let it down, and tracked it straight as an arrow to a full stop, long before running into the house.  He made a comment something like “I could have kept telling you step harder on them pedals but it wasn’t gonna do no good -- until you needed them you weren’t going to use them.  Now let’s go back to the airport and get some lunch.”  While it still took the rest of the day to convince myself that my 200’ wide runway was only 20’ wide, that lesson made an impact and over the next few days we landed on just about anything that would stand still long enough in just about any condition you might consider flying weather.

During that time, Lawrence showed no hesitation in demonstrating his mastery of flying 170s in any phase of flight, whether this was chasing down wild hogs close enough to chop their tails, or making 180 degree cropduster turns a few feet from the ground, or landing in 30 knot direct crosswinds. This wasn’t simply showing off, it was showing me the performance envelope of the aircraft, albeit in interesting ways.  It would have taken me years, if at all, to work up the courage or stupidity to try any of those maneuvers without having seen it done, but having seen it done and done it myself, I now know what is possible even if I never use it, and I think that is an essential ingredient in mastery of anything.

A good example relates to ‘wire strike avoidance’, as the FAA phrases it.  Their guidance on not flying into powerlines is simply to say that 1) power lines exist, 2) it is dangerous to fly through them, 3) they tend be located near the ground, and 4) so don’t fly near the ground.  Cropdusters don’t have this luxury, and of the many I met not one didn’t have stories to share about dragging cable home with them.  I have no desire to go out of my way to find powerlines to fly under or over or around just for kicks, but my course in wire strike avoidance could be summed up by the sentiment “You best drop down another 3 inches unless you want to lose some paint off your tail”.  There is a method to the madness, there are ways to tell if you have enough room to come out from the other side of them or not, and its not careless or reckless to train to do so – it’s really the opposite.  There is a lot of emphasis in FAA-based training on avoiding the canonical “fifty foot obstacle” on take-offs and landings.  They don’t specify what that obstacle is, but powerlines tend to be 50’ off the ground.  What do you do when you realize too-late that you are too low?  In an emergency how do you land in someone's backyard that is marked by a powerline at the threshold and a house 500 feet later?  Unless you practice it with someone who does know, the result seems far from assured.  And that’s exactly what the textbook method, as in the California flight school, is all about -- as long you color between the lines you will be fine, so never color outside the lines. What to do when you are coloring in turbulence is simply not discussed.


We landed in Monty's backyard a lot, practicing landing over 50' obstacles.


We landed here a lot too to check progress on a hangar/house being built by one of Lawrence's friends.


Anyone that was around the terminal at lunch or dinner time was invited out, and you had to be really quick to keep Lawrence from paying.

After a week of this, it was nearing time to head back to Fairbanks, it being mid-December.  At that point I made the decision to leave the plane in Texas so that I could spend the time that it would have otherwise took flying up here with more training.  I felt like I was going to learn a lot more by flying there than on the way up, and I’m glad I did.  I did go on one overnight trip to El Paso with Tamara, but the rest of the time I was largely practicing all day on my own with an hour or two instruction per day.  I made what seemed like hundreds of landings on grass and fields, dozens on small strips, and just flew around and had fun seeing sites from low level.  The value here to me was that I was in a sense doing the type of flying that I wanted to do in remote Alaska, but I was doing it in close proximity to civilization which eliminated much of the hassle, risk and expense of doing it remotely.  For example, over several days on my next trip I managed to get 4 flat tires (one of them twice…) landing on fields that I didn’t realize had mesquite in them and their associated thorns.  Getting a tire changed on the Jago River could be a week-long project costing thousands of dollars, but here it was a few hours and a few hundred bucks. In any case, Christmas approached and I headed back to Fairbanks commercially to begin working on the next phases of the project, leaving the plane behind to get brought up at some later date.



Our illbooten gotty.

Matt writes his own laws

Technically I am a State employee, because my salary comes from the State University.  Practically speaking, I am an independent contractor working through the University – my employment contracts last only 9 months and 3 months at a time (despite my grants lasting 3-5 years), they are not automatically recurring, I have no recourse if they are not renewed, and I have to bring in all of the money that pays for my salary, supplies, students, field work, etc from external grants.  Of the money that I bring in for this, the University takes about a 33% cut straight off the top for allowing me to work here.  About half of that goes towards a very good purpose – running the Institute that I am a part of, paying for the staff that helps me keep track of my accounts, purchases, compliance, etc.  The rest disappears and, from what I can see, largely works against my interests.  But I digress…

Being a State employee there are a number of laws and agency policies that attempt to prevent conflicts of interest, nepotism, embezzlement, etc., and nominally these laws are supposed to apply to me too.  Some far sighted individual, however, left caveats in most of them that allow, in this case, the President of the University to review individual cases and grant waivers to those laws and policies should he deem doing so in the best interest of the University.  And so began the process for me, as a University employee, to hire myself as a vendor of air charters. 

Back in 2008 when I first had the idea to do buy a plane, I had tried to get the University to own it  (using money that I brought in externally) but after multiple attempts at various angles it seemed like too much of an uphill battle to be a productive use of my time, there were simply too many people in the chain that were too willing to say No.  Actually none of them said No, they just kept generating lists of obstacles that I had the impression would never end no matter how many I overcame.  So in that process I decided it would be easier to go through a side door by owning and operating the plane privately, but avoiding the possible conflicts of interest inherent with that by not actually owning the plane ourselves.  But as I previously described, this created a lot of stress and complication for us and our friends, and opened the door for unscrupulous partners to attempt to rip us off.  So I felt it was time to try to go through the front door again, this time owning the plane and business ourselves but developing a transparent plan to mitigate the potential conflicts of interest.

This process had precedent both through the exemption-possibilities that already existed and by the nationwide precedent whereby business partners-with or gets spun-off from Universities in some sort of technology transfer arrangement, making use of University research commercially for the public good.  So that was the general angle we took.  A newly formed unit on campus was designed to facilitate just such endeavors and I turned out to be one of their guinea pigs in getting arrangements like this through the system.  I began the process in September and by Christmas it hadn’t really moved forward much, as there seemed to be a lot of research to be done and a fair amount of caution on the part of this unit in not wanting to move too quickly in running this up the chain or even letting the chain know what was coming.  So that spring there were a lot of drafts, modifications, course corrections, etc, before getting it up to the level where real approvals were decided.  Personally I would have preferred a more direct approach where we just got everyone in the chain in the same room, locked the door, and hashed out all of the issues, but on the other hand I had tried that 4 years earlier and got no where with it, so I was trying to stay patient and let the system work in it own way.  The time pressure was mounting, however, as my project expired in July and I needed to get the work done in May, or the latest June, but there was still a huge amount of risk here – not only did I not have the legal waivers to hire myself for the work, but I still didn’t have the plane here, a hole in the belly of it for taking photos, a commercial license so that I could legally fly the missions, or a system simple enough for me to operate the equipment while flying the plane.  So I needed to have backup plans.

One of these backup plans was to put out a request for bids to see if I was missing anything in terms of there being existing vendors out there that could handle all of logistics and charters required for my field work.  So I compiled all of my needs and submitted that paperwork in February, hoping that was enough time for that process to occur before my field season began in late-April. Another backup was to request a no-cost extension to my mapping project to extend the deadline through the summer to give me more breathing room in case any of the many unknowns didnt come together in time for June. Another was to find a supply of commercial pilots that could operate my aircraft.

By late February, it seemed like my legal paperwork was on track.  Contact had been made with all of people in the chain and it seemed like a plan was solidifying with no major objections.  The basic idea was that I made the pitch that University research, including mine and others, was hamstrung by the lack of availability of suitable aircraft to conduct airborne geophysics.  At UAF we do great work with satellites and great work on the ground, but that middle spatial-scale that airplanes need to cover is quite weak.  While Fairbanks is a large hub of small plane commercial activity, most of these aircraft are tied up in routine village traffic or in hunting/tourism traffic. In summer, it can be quite hard to line up a charter, especially in August (the peak of hunting season and the peak of field work season), because most vendors are booked many months or even a year in advance.  Given the nature of our airborne research, it is also impossible to schedule reliably because we need perfect weather.  So because there is so little of our type of charter work, because it is weather dependent, and because it requires a lot of ground time, aircraft modification, testing, etc, there is simply not enough profit motive for a commercial operator to have aircraft available for our purposes.  So in short, my company, Jago River Air, is designed to fit that need by providing and operating aircraft suitable for these kinds of research activities.  The document required a long list of signatures and essentially details the plan whereby all of the potential conflicts of interest related to this are mitigated by various processes designed to provide transparency, additional oversight and scrutiny, and keeping me as PI at arms length from decision making in the procurement process.  That is, the approvals I eventually received placed no guarantees that my company would be hired or get preferential treatment, it just allows that company to compete within the normal procurement processes.  It also allowed me to run that company for other commercial gains outside the University, provided the time I spend on such activities does not interfere with my job duties at the University.  But I am jumping the gun a bit here – by early March it was still far from clear whether this would get final approvals or not.

Matt lands in Fairbanks

Trying to keep all of the balls in the air simultaneously, the next task was to get the plane back to Fairbanks.  So I combined a business trip to the Texas area with spring break and made a plan to fly back up with Kristin and Turner.  They were going to spend a few days with my parents being tourists in the Texas area, where there are a lot of dinosaur museums and digs as it turns out, while I took care of business and got myself and plane ready to fly up.


I was surprised out how much dinosaur research and display there was in Texas. This display actually had an interview with a UAF faculty.

It had been a few months at this point since I had flown it and I knew I would be a bit rusty.  So before launching out on my own, I did a few touch and goes with Tamara the ferry pilot I had initially contacted several months earlier.  It was a good thing too, as I had definitely gotten rusty.  Though it was several months of calendar time, practically speaking it was like I had just gotten my license a week ago since I had only flown for a week after getting it.  But once it seemed like I had the hang of it again, we fueled it up and I headed off for Vernon to work with Lawrence again to hone my skills and absorb as much as I could from him.  That first take-off headed there was perhaps the sloppiest, bounciest, skippiest landing I had done on my own in that plane, but fortunately I got it off the ground before running off the side of the runway or hitting any hangars, and landed a few hours later in Vernon. Within a few days of being there I had once again made what felt like hundreds of landings and was once again feeling like I had the hang of flying.  Besides the instructional aspect of it, Vernon is a great place to fly because it is always windy there and there are many runways to choose from to get the level of crosswind practice you like.  Within a few days more I was feeling confident about understanding where the limits of my ability lay and ready to make the trip north.


The excess width of this strip would be considered a waste of money by most cropdusters. Most seem to be about half this wide.


The first of my four blown tires.


The second (same side, different rubber).


The third.


The culprit -- dozens of mesquite needles.


A blurry mesquite branch, showing the inch-long needles. Most branches are laying flat against the ground and cant be seen on landing.

For a variety of reasons our plans to fly back together didnt work out, so I decided to fly back alone.

The next day the weather was a bit unsettled with winds stronger than I would have liked, plus I was still feeling a little unsettled myself, so I spent the day repacking and sleeping, and that night we went out for some night flying.  As we flew around in the pattern, we tried different configuration of lights and techniques.  At one point Lawrence suggested turning the lights off completely, making the point that sometime that’s going to happen whether you want it or not, so you may as well practice it.  So we did, but just as we were about to flare for landing, the runway lights went out too.  Runway lights at most unattended airports are turned on by clicking the airplane raido’s microphone while on a certain frequency.  The lights then stay on for some amount of time, like 15 minutes or so, then turn off to save power and to not be a nuisance to others.  Whatever that timer was set at, we had been flying around there about that amount of time, so I got to experience what landing without any lighting aids was all about.  One of the nice things about this type of airplane is that it’s aerodynamics are designed to allow it fly slowly, unlike many aircraft that are designed to fly quickly by slicing through the air.  Flying slowly on landing has the advantage that by the time the wheels touch the ground, there is not much kinetic energy or momentum left to be disappated, so landing distances are short, like going from 40 mph to zero in a car, but with a lot less directional control.  All that just to say, landing in the dark in a fast, heavy, high-performance plane is likely a much different experience than doing it in a light, slow plane, and in my case turned out to be fairly benign.

The weather by now had improved, so I got an early start on my trip north.  When I took off I had about 137 hours of flight time, more than half of which was concentrated in two one-week sessions in Texas after getting my license.  Whether or not that was enough by aviation standards, I felt good about it and ready for the challenge.  In many senses, flying 5-8 hours in a straight line isn’t much different than flying that same amount of time around a local airport.  If anything, it’s mostly easier.  There is of course navigation, weather, new traffic patterns, terrain, etc, to factor in, but largely these were things that I had a lot of experience with from chartering aircraft for my field work over the past 20 years, so none of it seemed particularly intimidating or onerous.  My initial plan was to fly 1000’ above the ground the whole way, so that I could really see the country and better enjoy the flying experience.  But by early afternoon I was already climbing up, both to avoid the bumps caused by the heating ground and to go faster in thinner air.  I passed countless farms, cattle ranches, stockyards, and small towns on the way up.  I was really surprised at just how much of this part of the country is dominated by agricultural.  It’s a part of the economy and lifestyle that I have very little experience with, but I can see why there is so much reference to it within society.  In any case, by that night at I had successfully flown about 700 miles north and spent the night in Gillette Wyoming. 


I passed tons of these cattle 'farms'.


If you look close, you'll see a spray-plane at work, just left of center.


If you look closer, you'll see he is hemmed in by powerlines.


As if powerlines werent enough...


A bit too much Agent Orange?


There are tons of small towns like this.


Even more like this.


Hog country!


Colorado is spray-plane country too!


And oil country!


It's been dry here a long time, only irrigated fields are green.


Even in cowboy country, the FBOs treats tiny planes like mine well.

A major side excursion of my trip home was to spend a few days in Polson Montana to have some modifications made to the plane.  There is an outfit here that changes the shape of the wing to give it the ability to fly even slower while at the same time improve control.  The kit can be installed anywhere, but these guys could do it much cheaper than I could have it done in Fairbanks, even including what I’d have to spend on hotels etc, and presumably they would do a better job since they were the experts.

Probably the tensest moments of the trip for me where trying to get into Polson, as it is surrounded by 9000’ peaks, near Glacier National Park.  The winds were strong from the west that day, and I was on the east side of the Rockies, meaning that I would have to enter the mountains on the downwind side, which generally speaking is the more dangerous side due to the downdrafts often associated with it.  As I continued northwards, the density of airports decreased and their characteristics became more challenging (like higher elevation, rougher condition, etc) and the larger ones became more highly controlled.  So I had to watch my fuel, route and winds more carefully to ensure that I had an alternate airport to land at that had conditions that I was comfortable with.  Willie, that runs the business, had previously given me his advice on a route through the mountains, so I was following that.  In case I couldn’t make it in, winds at the nearest major airport about 50 miles east were now over 25 knots and seemed to be increasing, so I was getting concerned about that.  But it was a clear, sunny day and winds were largely calm for me, so I kept heading north on the east side of the mountains. 

As I approached the pass I was aiming for, I could see lenticulars formed above it, an ominous sign.  The pass was at about 5000’ I think, but given the clear weather I kept climbing until I was above the peaks at 9000’.  All the while I was waiting for the turbulence that these clouds implied to hit me, and I was ready at a moments notice to dive down and head somewhere else, staying at an angle to the mountains to make that even easier.  By now I was almost over the foothills, and so close to the lenticular clouds I felt like I could almost touch them.  At each new burble I was white knuckled and braced for a 2000 foot per minute downdraft, but it never came, just a few occasional bumps to keep me awake and some minor pushes up and down within the waves.  Given the 30-40 knot headwinds, I wasn’t going much faster than the car traffic below me, so it was probably 30 minutes that seemed a lifetime to get past the peaks and into the valley.  But once there, winds were nearly calm and valley itself was huge.  I was quite relieved by this too, since the Polson airport is a small one and doesn’t report the weather, so I wasn’t really sure what I was getting myself into.  The airport sits right at the edge of a large lake and is nestled into some hills, such that the approach is flown below terrain and over water.  It took me a few laps to get my act together, but soon enough I was on the ground and taxied up to the hangar that would be the plane’s home for the next few days.


Translation: "Roswell is 250 miles S-SW"


Apparently this lake used to be a bit bigger.


I passed lots of mines like this too.


I thought Lewistown was a nice place, though I only spent a few hours there.


Nothing like flying under lenticulars when crossing the Rockies into a 40 knot wind...


The valley that Polson is in was wide and pretty, though the airport sits inside some hills right at the edge of the lake.


If not for the roads, I could pass this off for Arctic tundra.


Even though there was no fancy FBO here, they still gave me a car to use for the week.

Working on the plane in Polson was quite a pleasure.  The staff there were very friendly and I was able to do a lot of the grunt work to speed the process along.  Basically it consists of glueing some blocks to the leading edge of the wing, then attaching a piece of sheet metal over those blocks that essentially makes the leading edge of the wing fatter and wider.  Though work proceeded steadily, it inherently takes a couple of days because the glue needs to dry and many of the pieces need to be custom fit to each plane.  I was also there over the weekend, so we had Sunday off and I took the opportunity to go skiing at one of the local resorts, which was a small, family type of place much like what we have in Fairbanks.  The weather the entire time I was there sucked, so it was a good thing I got in the day that I did.  I was a little stressed about flying out beneath the clouds and terrain, but work went well and finished early on the last day.  I had meant to spend a day locally getting the feel of the new wing, but the weather was perfect on that day so I paid my bill, said my goodbyes, threw everything inside, and headed for Canada on my first take off.


Setting the foam blocks that form the shape of the sheet metal that will soon be bent over it.


Willie does the fussy work, like making sure my stall horn will still work with the new cuff in front of it.


Here you can see the old wing's leading edge inside of the new one. The difference in aerodynamics was profound.


The lake at Polson. Seemed like a nice place to spend a vacation.

Entering Canada is such a pleasure.  I’ve done it a few times by air and by car, and am always left with that impression.  Flying in requires calling customs in advance to let them know where you will be landing and when, as well as some details about your plane, passengers, and cargo.  I remember the woman I talked to asking me whether I had any alcohol and I said I had a box of wine.  She asked if that was 6 bottles or 12, and I replied that it was neither, I couldn’t afford wine in bottles, and she not only got a laugh from that but waived the customs fees since technically that box held the equivalent of more than the 2 bottles of duty free wine that is normally allowed.  I wasn’t sure whether I’d be able to fly over the top of the mountains or not, so I gave her an estimated arrival time that accounted for flying through the passes.  But as it turned out, I was able to fly over the top and saved an hour or more of flight time.  I killed some time by flying along the border for a while, but eventually turned towards Lethbridge where I landed about 45 minutes early.  From the plane, I called customs again to tell them I had arrived, connecting with the same woman from earlier who remembered me.  She asked why I was early and I told her, then she gave me an arrival number and said Welcome to Canada. No drama, no hassles, no interrogation or strip search, just Welcome to Canada. I called Hester somewhat out of the blue to tell her I just found myself in Lethbridge, so she dropped her plans for the evening and we went out to dinner and did a little shopping.  She was renovating the upper floor of her house and she was staying with a neighbor in the meantime, as the lower floor was nearly impassible with the contents from the upper floor.  But I found a couch to crawl onto and began planning my route through Canada.


I arrived in Canada a bit early, but they didnt seemed to mind much.


I went out to dinner with Hester in this water-tower-turned-restaurant.


I flew over Drumheller, searching for Dinosaurs.


And found one! Compare this to the 360 panorama I took on our ice-core journey in 2009.

I considered staying an extra day there both for the fun of it and because there was a storm brewing in the northwest, but left early the next morning to get as far north as I could before the storm hit.  I stayed ahead of it, shooting a line between Calgary and Edmunton, and hoping to get as far north as Ft St John.  Winds were picking up as a flew along and the main game here was figuring out where to get fuel, as each stop ends up taking an hour so you don’t want to do it too often but given how far apart the northern airports are you don’t want to run out either.  Being a little chicken, I stopped at Grand Prarie to fill up, and while checking weather there for the next hop, only another 50 miles or so, I psyched myself out and decided to spend the night.  By the time I found a hotel and got there, the snow had started, and by the time I woke up there had been significant accumulation.  But fortunately, there worst of it occurred overnight so by morning I was set to continue onwards. 


The FBO here was after bigger business than mine, but still did the job. This was about the last real FBO on the trip north.


The blue dot is where I was, in the eye of the storm, and trying to stay that way... The number of barbs on the arrows indicate wind speed, with many in the 30-50knot range.

As I continued north, I left the main sphere of influence of the pacific storm belt and entered interior continental weather.  Here it was still winter again, with winter snow packs and temperatures.  The weather was beautiful and I continued pressing onwards, staying right above the road and eventually crossing the rockies once again to get to Watson Lake.  Last time we flew through in 2009 we were concerned that we might get stuck there due to clouds and snow, so I was eager to make it to Whitehorse.  But the way things worked out, it was clear the best choice was to spend the night there as I would not get to Whitehorse before dark.  I had called ahead to arrange a place to tie down and plug in the engine heater, as no longer was I in the land of FBOs with courtesy cars and snacks, this was the North.  Watson Lake airport is a bit of a strange one, as it is huge on the one hand with a control tower and everything, but hardly anyone lands there.  It is largely the alternate airport for Whitehorse, so it has to accommodate the big planes upon occasion.  There is fortunately a Pilot Lounge there, right next to the control room.  In the two times I’ve been there, I’ve met probably 5 or 6 of the controllers, with a wide range of personalities and pleasantness, but all sharing in common that this job is for them a night guard position, where they expect that 99% of their time is going to be spent doing something of interest to them since there is nothing actually to do for work.  In any case, when I walked into the control room the first time, the guy and his girlfriend were in the middle of an argument and looked up at me like I had two heads, but told him I had just landed and was going to use the pilot lounge for a while, he kind of nodded and got back to his argument.  So I closed the door on the lounge and slept there for the night.


For the most part my travels were in clear skies, with just a few occasional clouds.


Climbing up to get over.


Hardly any wind, which was nice.


Crossing the Rockies one last time. But here it was still winter.


It didnt get any warmer on the ground.


I had to keep my soda in the sun to keep some of it liquid.


Watson Lake, frozen. The airport is the straight line to the right of center.



I liked the Watson Lake area. Reminds me of Siberia a bit.


Weather was once again fine in the morning, so I blasted off first thing after submitting my online custom form for getting back into the US.  No friendly phone call here, one can only use this online system and deal with all that it entails.  It’s not a horrible system, but its all kind of a pain, especially with a flaky internet connection.  In any case, I got it sorted and eventually found Whitehorse where I needed to get fuel, as there was nothing between there and Northway Alaska.  I found Whitehorse a friendly place both times I’ve flown through there.  While paying for the gas, I felt my phone find a connection and tell me I had messages.  I checked the voicemail and got a completely unintelligible message except for a part about it being from US Customs and they wanted me to call them.  When I did, they told me that they did not have the manpower to leave the road border crossing and send someone to Northway for me to clear customs there.  So they wanted me to continue on to Tok, where I should get fuel, and then continue on to Fairbanks and clear customs there (Yes, this means I could get out of the plane in Tok, drop off my nuclear warheads and illegal drugs or whatever else customs is concerned about, fuel the plane, and then going through customs in Fairbanks clean).  He told me I also had to modify my online custom form, and he didn’t seem to care that I had no internet connection.  So I found a connection and modified the form.  But then I had the bright idea to call Fairbanks customs and make sure I wasn’t going to be met by F16s.  They said that yes, this plan was OK, but that I had to arrive by 4PM.  I told them that I had just filled out the form with an arrival time of 4:30PM and it had been approved, as I didn’t think I could get there earlier.  They said to just be there as soon as I could, and we left it at that.


Not very hospitable terrain to land a plane.


Fortunately someone had a paved a runway under my entire route.


There were a lot of cool terrain features left over from glaciers.


The trip to Fairbanks went mostly according to the new plan.  The biggest issue of the trip was that I was flying into a 20 knot headwind the whole way once out of Canada, which slowed me down.  I was also a bit concerned about fuel, as making it to Tok left me with little reserve.  But soon enough I was in sight of the runway.  I called ahead to get a weather check, but no one responded, and before long I could see the wind sock aligned with the runway.  I made my approach as normal, but as I was on short final it seemed like something was odd when I just happened to glance at my groundspeed on the GPS – I was going much faster than my indicated airspeed.  I looked up at the windsock and then saw that while it was indeed aligned with the runway, it was pointing 180 degree to winds I had just been flying in.  That is, I had just flown through a 40-50 knot windshear about 200 feet above the ground and was about to land with a 20-25 knot tailwind, if I could get the plane down in time.  I opted to go around, which meant climbing back out through the shear zone, making a 180, and coming back down through it to land with the wind on the nose.  It all went reasonably smoothly and calmly, but in retrospect it was a lot to throw at a pretty low time pilot, and as I describe in the next section, hidden flaws in the airplane could have conspired to cause a much different result.

In any case, I filled up with fuel and took off again, this time with a pretty strong crosswind on the runway, and headed for Fairbanks.  By this time I was pretty used to flying along roads and opted to continue to do this all the way in.  I was still fighting the headwind, but it looked like I would sneak in right at 4 oclock.  I asked to land on the big runway so that I could taxi to customs, but got diverted to the small one at the last minute because a couple of jets were coming in.  This meant I had to sit at the taxi way that crosses the big runway for about 15 minutes.  I finally got across and taxied up to Customs where I was greeted by a very unhappy man with a large gun who bitched me out for being late and telling me he could charge me a $1000 fine.  I pleaded stupidity and ignorance, citing the name of the officer I called earlier who said 4:30PM would be OK, then blabbered about being a new pilot, headwinds, taxi delays etc.  But after his perfunctory reaming, he smiled, took my license, scanned it on his machine, walked me out the door, and headed to his car to go home, all in less than 5 minutes.  I taxied back across to our tie down, where Kristin, Turner, and one of Turner’s friends were waiting for me.  I and the plane had made it home!


I think this was part of the Tired Iron snowmachine race, flying over Tanana on the Yukon River.


Made it!


Just in time for the Easter Egg hunt.

Matt buys another plane

A few days later I went over to see a mechanic I had met a while ago to see if he wanted to take over general maintenance on the plane, as well as ask his opinion on a few things about it.  Before even looking at the what I asked about, in 5 minutes of walking around and inspecting the plane he convinced me that it was not safe to fly.  We brought it into his shop to look further into it, and he convinced me that it was really not safe to fly.  By the end of the next day, he had a list 2 pages long of things that needed to be fixed before the plane was safe to fly.  This list included replacing nearly all of the landing gear hardware including the legs, replacing the carry-through spars in the wings, replacing a wing strut, and number of smaller but equally important other issues.  Having just flown this plane from Texas in winter, he was shocked and amazed that I lived to tell the tale.  I went over all of the issues he identified and got second opinions on most of them, and indeed it seemed I was lucky.

So there were two main issues here, aside from survival – time and money.  It was now late March, and my plan was to fly like crazy over the next few weeks so that I could qualify to take the commercial test, which is an FAA requirement for being able to charge money for aerial photography.  My goal was to get that completed before leaving for the glacier in late April, such that I would have May and June to complete the actual science work, assuming that all of the other paperwork was sorted out by then too.  In terms of that license, I also needed a certain amount of night flying, and night was rapidly vanishing as summer approached in our northern location.  I had also planned to install the camera hole now so that I could begin testing, but there was no way to do this and the repairs at the same time. The second issue was money.  The cost of these repairs was approaching what I paid for the plane.  Our initial idea was to buy two planes  -- this one (the cheap one) and a bigger one that would ultimately be the main photography plane.  However, we had already taken some large hits by getting screwed by our former air taxi partners and hadn’t had time yet to sue them, and with this new maintenance would soak up all that was left and more.  So at this point, I felt like I was stuck in a sinking ship with no choice but to repair it, as even if I could find another plane in the price range in the time I had left there’s no guarantee we wouldn’t find something wrong with it too late too.  So I essentially bought the plane twice.  I sent this list of discrepancies and some photos to the folks I had bought it from, but they did not respond.  You can find all the gory details about the repairs required here. Of course spring is a busy time for aviation mechanics as everyone gears up for summer, so we couldn’t just dive into this project.  It did seem, however, that it could get finished by the time I returned from the glacier in mid-May.  This meant that in order to do the actual work, I would have to fly another 60 hours in last two weeks of May, take the test by about June 1, so that I would have the month of June available to do the work before the project expired in July.  As that seemed like it might be pushing the envelope of possibility even for me, I decided it was time to get serious about finding a commercial pilot with nothing better to do than help me; Kristin had by this point taken a job flying for a real company again. The work itself would only take a few days, but it can take a month of waiting on weather to get those few days.  And of course I also needed to get the paperwork completed so that I could do the work, as well as rebuild my sensor system so that I could drive and shoot at the same time.  Things were starting to stack up!



Fortunately none of this occurred in flight...


Seriously?


This was a pretty invasive repair.


This was perhaps one of the scariest issues -- near center coming out of the wing you can see a fuel line that is goes up and bends down to the left. The part going up effectively reduces my useable fuel capacity by 5 gallons or more since fuel is gravity drained from the wings. When I did that U-turn over the Tok runway, I only had about that much fuel left and could easily have run out of gas even though there was plenty in the tank.

 

Matt gets legal

In the downtime of getting the aircraft functional again, and in between preparing for my glacier field work and managing my other projects, I was still not yet able to hire myself to do this work.  By late April it seemed we were close to having a final plan in place to seek approval from the President, but it was still moving slowly.  Complicating matters was my upcoming field work on McCall Glacier which would take me out of action for 2 weeks while in the field.  Up to this time, I had thought that my backup plans were moving along behind the scenes, namely that the bids were being sought by Procurement to see if any one could provide all of the logistical services I needed, including either making the maps or providing an airplane to me that I could do it in.  As it turns out, this paperwork had been lost in a shuffle for two months.  So, the day before I left for the field, we had a meeting with Procurement to get things back on track.  The plan we devised was to extract the airborne geophysics and seek bids just on this because this work had the earliest deadline, and then after I returned we would begin work on supporting the other projects later in summer.  So, the morning I left for the field, I submitted a revised scope of work that sought bids on these two options – either delivering a final product (the map itself) or delivering the air support I needed for me to use my equipment in the plane (to make the maps).  The plan was that work on this would begin that day and by the time I returned we would have an answer as to whether anyone was available for either of these two options, and at a price that I could afford.  The reason for the two options was to cast as broad a net as possible.  It was my strong suspicion, having hired commercial contractors in the past to do this sort of work for me, that anyone bidding on the data deliverable would be orders of magnitude more expensive than my projects could afford (thus my reasons for wanting to do this myself), but I figured you don’t really know until you ask and this option would be best for the project if it was affordable since it would be much less work (and therefore cost) for me.


It was a stormy spring on the glacier, but not as stormy as campus on my return...

When I returned from the field I discovered that no progress had been made.  None.  Needless to say I was distressed about this, and my expressions of distress got things rolling.  In the meantime, my conflict of interest management plan was nearing final approval, and within a few days of my return it was officially approved.  So, now I was able to submit my own bid for this work through my company, which I did.  That is when the shit really hit the fan.

Apparently Procurement was quite distressed that they were not in the loop regarding my waiver approval.  They felt that there were issues on their end that were not addressed, and perhaps they were justified about wanting to have a say.  What was worse, however, and much worse, in my opinion, was that they had lost the paperwork I submitted the day of my trip and were now soliciting bids using text I had written up in February, which was wildly different in scope.  I only discovered this after my waiver approval when I asked them to send the request for bids to my company.  I pointed out to them that that these requirements were substantially different than what I had submitted to them as PI and indicated to them that at this point that I, as PI, had no need for the aircraft deliverable option, as my strong preference was to have a data deliverable.  Part of my conflict of interest management plans stipulates that I should ‘remain at arms length’ from the Procurement process, though of course I retain the right to control the scope of work since I am the only person that understands what work has to be done.  Regardless, they told me to shut up and eat my porridge and decided to restart their botched solicitation for both scopes, delaying things further.  It was by now late May, and my scope of work stated that work would begin June 1 after a week of testing in town (which was the next week), so that I as PI would have enough time for suitable weather windows to complete the work by July.  In any case, I kept my cool, barely, and let them do their thing, which they claimed would be completed within the week. 

That deadline came and went with no information.  Frequent voicemails and emails over the next few days resulted in no information about the results of their solicitation, which at this point had also turned into a requirement for my COIMP.  After a week (!!!) past this deadline, I lost my cool and blasted them with both barrels.  I sent an email, copying all of the lawyers, supervisors, and personnel involved with the entire process, essentially telling them that if I had hired Procurement to do my procurement services (such as if I hired a company to provide me with mapping products) that I would refuse to pay them and that if they did not release the results of their solicitation to me in the next day that I would use the State’s freedom of information laws to wrestle it from them, along with the past 10 years of related searches, with the implication that their violation of their own rules left them exposed to lawsuits.  Perhaps not surprisingly, this had more affect on prompting action than all of my previous emails politely suggesting we get together and sort out whatever issues were involved.

Fully expecting to get fired, I took a day off to cool down and ignore the world.  In the meantime, apparently a large meeting of lawyers, managers, and bean counters took place to try to understand and resolve the issues.  Procurement would still not release any information about the results (!!!), other than to say that my company’s bid was the lowest, at about $36k and the next lowest for data deliverables was $750k.  I wasn’t there so I don’t really know, but I suspect this went a long way towards further validating the claims I made to justify my COIMP, that the work I needed to do simply could not be done by anyone but me.  Still, Procurement claimed that I had “tainted” the process by “changing” my scope of work in “mid-stream” after my COIMP had been approved, and without me there to point out that I had simply discovered their screwup and used my prerogative as PI to cancel one of the scopes (leaving the other as-is) to save asking other vendors to put effort into creating a second bid for a scope that I no longer had use for.  Be that as it may, it seemed that there was an overall willingness to ignore my threats and find a constructive solution to solve my short and long term science needs.  The upshot was to get me in the air immediately they agreed to waive another policy that would allow me to fly my own plane right now despite not having the 1000 hours experience normally required for approval to do so as a University pilot, as well as let me charge any expenses required to modify my aircraft to support the work, such as installing camera ports, etc., while they in the meantime sorted out how to actually implement my COIMP in a way that Procurement could live with, which they thought could happen fairly soon.

In a sense, that was great.  In others, it was quite distressing again.  While I always had on my mind to be the pilot, they didnt realize that to do this sort work I was required to have a commercial license by the FAA and given my setbacks on aircraft repairs I had been planning to hire a pilot.  Further, it was now June, and those required modifications were going to take some time.  Worse, though it was certainly better than nothing, covering my current expenses did not help with all of the substantial expenses in buying the plane and making it airworthy.  So now I was legal (well, almost, it act, took another month to implement the short term plan of waiving the 1000 hour requirement) but not in the way I wanted and I wasn’t exactly ready for it.

Matt gets really legal

To get my commercial license, I had to have at least 250 hours of flight time.  I only had about 190 at this point.  The day after I got the plane back, May 23rd, I began flying 6-10 hours a day on long cross-country trips, making double duty out of building hours and scoping out all of my future staging sites for the research.  I flew over 50 hours in 8 days.  I then went back to ProFlite to use their ‘complex’ plane, of which I needed 10 hours to meet the commercial requirement.  A complex plane is simply one that has retractable landing gear and a constant speed propeller.  I knocked this out in short order and soon had only one hurdle remaining – taking the commercial checkride.


I flew about 50 hours in 8 days to catch up on the requirements for my commercial license.

I had avoided scheduling this checkride for several reasons.  First was that I had a lot of time to build and there was no way of predicting whether my plan to fly like crazy would actually work or not.  Second was that it takes an instructor to sign off that I was ready, and I was more or less on independent study.  One of my reasons for going to ProFlite again was to get the instructor sign-off, while building time in the complex airplane.  Unfortunately, just at the time I was finishing that up, that instructor got a new job (working with Kristin) and disappeared for training.  Like out of the state disappeared, with no warning.  And that was the only instructor qualified to fly with me in my plane, which was a tailwheel aircraft.  So now I had the hours, but no instructor sign off, and I really had to scramble.

As fate would have it, my mechanic Karl was also an instructor, though he hadn’t actually taken on students in some years.  After chasing down all other leads for instructors in perhaps the busiest week of aviation in Alaska, when the lake ice melts and summer truly starts, I was able to convince him to fly with me and determine whether I was ready for the test.  This went reasonably well, and soon enough I was ready to set a date for my test.  This is when the scrambling really began.

You can either get your checkride from someone at the FAA or you can hire a Designated Pilot Examiner (who doesn’t work for the FAA).  Nearly everyone uses a DPE, for a variety of reasons.  There are three of these folks in Fairbanks.  The first I contacted said he was too busy for the next several weeks, the next said she was going out of town for an extended time, and the last said he could do it in two weeks.  While two weeks was a long time for me, I could still be working on all of my prep work in the meantime, since at this point I still hadn’t built, tested or applied my sensor system at all!

About a week later, that DPE called me back to cancel my checkride.  He wouldn’t give me any reasons, except that it had something to do with me personally.  As far as I knew, I had never met this guy, but no amount of probing yielded any details.  He told me, though, it should not be a problem because he had called the local FAA and they were willing to give me the checkride, and that really it was to my advantage because it was free with them.  I never learned what the real deal was, but this was an enormous red flag to me that perhaps my prior dealing with the FAA were about to bite me.  

I’m not one to put any stock in divine intervention, but if I were a religious person I’m sure it would be crystal clear to me that after four years of continual roadblocks like these that fate did not want me to succeed in my endeavor to make these maps.  But, not being religious, I sucked it up and secretly implemented Plan D2.1 and made plans to return to Texas.  I told no one about this, certain in the knowledge that the enormous conspiracy obviously working against me would try to thwart that plan before I could succeed in it if they found out about it.

So a few days later I flew in to DFW on the red eye, drove 3 hours to Vernon (my 3rd trip in 7 months) to get there after missing lunch, and was promptly told to preflight the plane be ready to fly in 10 minutes.  This was my kind of place.

Texas in June is hot.  Really hot.  So hot you can barely stand on the tarmac let alone sit on it.  Nevertheless, knowing how things work here, I decided to do a thorough preflight.  A preflight inspection is required by the FAA, but there are no rules on really what is done in such an inspection other than to say “If there’s something broken, it’s your responsibility to find out before you take off”.  But of course most pilots aren’t mechanics, so mostly we just check the fuel and oil levels and make sure all of the big pieces of the plane are still attached firmly.  In preparation for my checkride, however, Karl the mechanic had drilled into me the mechanic’s view of how to do a preflight, and on top of everything I learned about what to look for when buying a plane, I was actually pretty savvy.  One of Karl’s rules of thumb is to look make sure every bolt that you see has a nut attached to it and that every nut has a safety lock on it, which could be a lock washer, nylock, or cotter pin.  As I examined the nosewheel of this 172, I realized that the main bolt that holds the scissors (that allow the shocks to compress while still steering the wheel) in place had no nut.  Further inspection, while lying on the broiling hot tarmac, revealed that the nut wasn’t missing, but rather the bolt head had been sheared off and only a minor tap caused it to fall out completely.  It’s debatable whether the loss of steering or a blown nosewheel on landing would lead to an accident, but it is certainly conceivable that a low time pilot unfamiliar with that airplane, that airport, that heat, having flown through the night, and in a rush to get a commercial certificate could have run off the runway, bent the prop, and toasted the engine.  Fortunately we didn’t find out, but now their only plane was grounded, and this is the sort of event that can lead to driving 3 hours back to Dallas to take the next redeye home and updating my resume.

Fortunately, Lawrence had been a repo and salvage man in addition to being a cropduster and mechanic and was not one to be above robbing Peter to pay Paul, and soon enough a different plane was grounded and we were airborne.  By the end of the day, I had my instructor sign off to take the test and it was just a matter of waiting a few days for scheduling with Mary, who had a full schedule testing the students that were already there and also ready to finish up.  This left me plenty of time to get all of my night flying requirement completed, since I had lost the opportunity to do that in Fairbanks, it now being well after equinox.

In the end, my checkride went well enough to pass, though I was a bit embarrassed to have to use my “go-around demonstration” when I was supposed to be demonstrating my emergency engine-out landing.  My practice until then was always in the morning or evening, just by happenstance, and the checkride was in the hottest part of the day.  My approach in this maneuver was fine, I was set up perfectly over the threshold.  But once over the tarmac, there was so much hot air rising up from it that I floated nearly a thousand feet before giving up and going around.  Fortunately the next time went better and a few hours later I was on my way home, commercial license in hand.  I was now legal!  But was I ready?


Rayna and Cindy, now legal to fly airplanes on their own, getting their logbooks signed by Mary.


Me too, but now I can get paid for it.

The next day Rayna, a helicopter instructor, taught me to fly helicopters. Though I'm not a fan of them, flying a helicopter is simply 100x funner than an airplane.

Matt makes a hole

The dilemma I faced at the end of May was this…  I had hoped that by April I would be approved to use my own aircraft to support the work.  But I didn’t want to invest the many thousands of dollars it would take to modify the plane to support that work if I wasn’t able to use it.  I knew my COIMP would explicitly state that I get no guarantees or favoritism, only that I’m able to compete for projects, and I was willing to take that risk, but not the risk of not getting the COIMP itself approved.  This didn’t happen until mid-May.  After I dropped the bomb to get things rolling with Procurement, I was then stuck being the pilot.  But I couldn’t be the pilot until I got my license.  I couldn’t get my license until I built enough hours.  And I couldn’t build those hours or take the checkride if the plane was in the shop getting modified.  So I made the decision to keep the plane flying until I got my license, which meant that I would face that downtime for modifications afterwards.  Worse, it’s not like Karl and every other mechanic didn’t have many other projects and commitments of their own.  So added to all of the stress of getting myself legal and wondering whether my photogrammetric system would even work at all, was knowing that there was an excellent chance that my plans to modify the plane could take weeks to months, or even fail completely.

Unfortunately the good old days are gone and you cant just cut holes in your plane wherever you like to install cameras.  There are several different ways to go in doing it legally, but the easiest is to find plans that were approved a long time ago before anyone at the FAA really cared.  Such plans usually come in the form of factory drawings (made by the manufacturer) or Supplemental Type Certificates (made by 3rd party inventors and approved by the FAA).  Fortunately Cessna had factory drawings and one time sold a kit for installing a large camera hole in 170s.  Even more fortunately, I met a guy on a 170 forum who had installed this kit and even had some of the pieces of it to use for templates.  Even more fortunately, Karl and Roger were the right men for the job.


The hole from the outside.


From the inside, along with sensor mount and camera. There is a piece of glass sealing the hole.


Ultimately I stripped everything out of the plane so that I could carry more fuel inside to extend my range. The real trick was getting this fuel into the wings while flying...

If we were just installing the hole, likely the work would have got done while I was in Texas.  However, I also decided that I wanted to have more fuel on board, so I had them also install an STC’d option for additional fuel tanks, which required ripping apart the wings and putting them back together.  The good news or bad news about this was the amount of corrosion and other weirdness discovered while doing that.  The good news was the plane was made much safer than it had been, but at the expense of money and time.  I got back from Texas on July 3rd, and it was another two weeks before I could fly again.  On the bright side, the day I got the plane back my company was finally awarded the subcontract to do the work.  So after all that effort to get legal personally, I could now hire a pilot.  But now that I was legal, I had the contract, I had the deadline extension, and I had a plane suitable for the job, so I flew.  And flew.  And flew.

Matt learns to make maps

After all of this effort and heartache, now that I was ready to go it was still not clear that I had any idea of what I was doing scientifically.  For the past several years I had done quite a bit of vertical photography, but I was never able to turn the corner to produce the final product that I was after.  My previous system was quite complex, and required my full effort as a passenger to keep functional. It involved running a Windows laptop with several pieces of finicky independent software running, 3 different GPS, 2 monitors, an expensive IMU, a laser, and a camera.  It was a great idea on paper, but highly problematic to keep functional.  And it lacked a precise timing system to determine the camera position when the photos were acquired.  I was able to make pretty pictures out of it, but I never had any independent data to compare with to determine its accuracy.  And I knew there was no way that I could keep that running and fly the plane, at least not at my level of flying experience.  So I scrapped it all and took a much simpler approach.

The fundamental difference in approach related to determining the position of the camera when the shutter tripped.  My previous system used a flight management system running on the laptop which monitored the aircraft’s position with low resolution GPS (about 5 meters) to determine when it was over a predetermined photo location and then fire the camera.  This is the elegant way to do things, but complicated.  The new system used an intervalometer to fire the camera at a specified time interval.  This approach takes more photos than necessary, but eliminates the laptop completely.  More importantly, the intervalometer I was using got feedback from the camera’s flash port to determine the moment the shutter was tripped to better than a millisecond and simultaneously sent an event pulse to the high resolution GPS to yield a photo location within about 10 centimeters.  As I discovered in the many, many hours I spent working out the processing chain, this accuracy of timing is essential in creating an accurate map.  So the penalty of using more memory by taking more photos than necessary was more than offset by both the simplicity of the system and the accuracy of the final result.

I picked a variety of targets in the Fairbanks area for testing.  These included the Fairbanks International Airport (where the plane is based), UAF, the Tanana River flats, the Minto flats, the Ft Knox goldmine, and anywhere else I felt like. Those last two weeks of July were a blur of flying, processing, engineering, tinkering, and repeating the process until I had a workflow for planning, acquiring and processing data.  All of this was done while planning and preparing for another two week field trip on McCall Glacier, starting the first week of August.  By the time I left for that trip, I felt like I was finally on track and where I had hoped to be by last April.  Though I still had not yet compared my results to any other data or similar or higher accuracy, by the time I left, I felt like my plan was finally on the verge of turning the corner, had proved to myself that I was capable of doing the work, and that it shown promise for achieving the accuracy I had hoped for.  I returned from that field work to find the plane and equipment still functional, and did further testing while waiting for my weather window.  All of the pieces had fallen into place, and now it was showtime.


I did a lot of local testing. Lake Minchumina (lower left) is about as far away as I would consider local.


Turner practices at home...

... and at the museum...


...and at the fair...

... and more at the fair...


... so he can be ready in case papa falls asleep.


Or in case he gets too distracted taking pictures.


Lake Minchumina, with Denali in the background.


We caught a pike on our first cast!


The lakeshore is at the bottom, but the lake itselfcant be mapped with this technique. Here a large forest fire was raging weeks earlier.


We can map the burn scar pretty easily. Here an island of trees remained, with half of it almost going up in smoke.


A new buliding on campus, next to mine.


Before they go up, they have to dig down.


East Ramp, Fairbanks International.

Matt loves it when a plan comes together

The time had arrived.  I was monitoring the weather constantly and it seemed that in the last few days of August a several day window was opening up for me to grab what had become the holy grail of my existence – making the maps near Shishmaref, as well as repeating all of the ones we had done in 2011 but this time at the accuracy I was after.  My flying skills were at their peak, the plane was functional, the new fuel tanks gave me the range I needed, the photogrammetric system was stable and simple enough to use, and I was feeling ready.

The weather window arrived and I departed on August 30.  My initial goal was to fill the wings in Ambler using the 40 gallons of 5 gallon jugs I had inside the plane and then fly into the Noatak valley and hit those sites along the way to Kotzebue.  Unfortunately I got a bit of a late start, so I did most of the block in the Kobuk River valley near Ambler and went straight on to Kotz.  I had no hotel reservations and it took some time to get my phone to work there, but by 10PM I was in a room downloading the days photos, uploading the next day’s flight lines, and shoveling some food in me before getting a few hours sleep for an early start in the morning.

That morning the weather looked good for the Seward Peninsula, but there was an enormous low pressure system barreling down from the northwest that threatened to shut down the State by the next evening, so I knew it was now or maybe never.  I launched by about 9AM with fuel fuel, empty data cards and my fingers crossed.

My first real scare came about an hour into the trip, just before reaching the first photo block.  I have an electronic fuel flow meter in the plane that does an amazingly accurate job of indicating how much fuel I have burned, something I considered essential in long-range photo missions.  However, it doesn’t really know how much fuel I have left because there could be a leak.  In my excitement to get there, I hadn’t checked the actual fuel gages inside the tanks until I got there, and my heart sank as I realized that one tank was prematurely almost empty.  I checked the wings and did not see anything dripping, but I faced the dilemma of turning around to check it out before something went really wrong (and risk never getting back here) or continuing on.  A real fear here is that the leak is at the fuel selector valve on the floor, which meant the belly could be filling with fuel and fumes.  The other fear is that if I pumped fuel from my aux tanks into the empty tank, that it would also leak out and then I might not have enough fuel to make it back.  I decided to keep flying, but to do every other flight line as a means to get some of the work done (there is still sufficient overlap in every other line to make a seamless mosaic) and in the meantime draw from the good tank until it was near empty.  Then I would pump the aux tanks into both of the main tanks and see if the sketchy one went dry right away, knowing that I still had enough fuel in the other to make it a real airport. 

As things turned out, there was no leak and I had let my imagination get away from me.  What had happened, as often happens, is that the left tank was being drawn down first rather than both tanks equally and it had burned exactly what the electronic meter said had burned once I did the the math correctly.  So, once my aux tanks were pumped and I saw there was no leak, I then went back and did the every-other-line that I had skipped and completed the first block!

The weather was still holding out, but it was clear that a system was moving in.  I started with the southernmost block because the weather was coming from that direction, and moved into the bright sunshine of the northern block about 20 miles away.  There was a strong wind blowing about 25 mph in the direction of my lines, but it was not particularly choppy.  The lines were also long, so I had plenty of time to set them up and fly at my planned speed.  This required full power with the nose down when flying into the wind, and flying a very low power with the nose up when flying downwind.  It was an excellent example of how power and pitch get used to control flight, and I was glad that I had long since figured this out.  I kept flying until I was empty enough to fill the wings again using my internal jugs, and headed to Shishmaref about 20 miles away to do it.

One of my rules of thumb when flying in Alaska is never to land somewhere I’m not comfortable spending the night or leaving the plane unattended for a week or more.  Shishmaref didn’t really fit that bill as I had no contacts there, but there was little choice in this case.  I had been monitoring the weather there over the radio with their automated broadcasts, and I was increasingly concerned that every time I checked the wind speeds were increasing.  Fortunately it was more or less right down the runway so it didn’t cause any hassles on landing, but refueling was a different issue.  Filling the main tanks with jugs is not too hard, as I can stand on the main wheels to reach the fill caps.  The aux tanks fill at the wing tips and present a challenge.  I had worked out a system in Fairbanks whereby I could strap three of my best jugs together and stand on them while fueling.  This system is sketchy at best, but in a 15 knot wind the wings rocked with several feet of travel.  So here I am, standing on plastic fuel jugs, holding a full 30 pound jug over my head, watching the waves whip up on the Arctic Ocean, trying to get most of the fuel into a tiny hole without taking a tumble in one of the most remote places in the country.  But in any case I succeed without major incident and popped off the runway, hoping that the next few hours would bring a four year effort to a successful conclusion.

By the time I went to refuel, I had finished about a third of the large northern block.  There’s not much to say about finishing the rest of it except that it went well.  The winds kept increasing, but were stable, the camera was working fine, and soon enough I was done.  I remember flying that last line, counting down the seconds, in somewhat disbelief that it was actually happening, thinking about what it had taken to get here.  But there it was, I passed the end of the line and turned the camera off.  While it was not yet clear whether there was some as-yet hidden flaw in the images, but even if I had to do it again, at the least I had proven I can do the hard parts of the job.

The only remaining stress for the day was getting back to Kotz.  I had hoped to be back by 7PM as that’s when the fuel placed closed, else I’d have to pay a $175 callout charge, but at this point there was no way I was getting back by then unless I wanted to take the shortcut by making a 30 mile open water crossing.  I decided to take the scenic route home and fly a few lines on Cape Krusenstern where I knew there was a lot of scientific interest that could lead to future work, before skirting the bay in a big arc back to base.  I arrived back in Kotz about 9PM, after 12 hours of flying, refueled to be ready for an early start in the morning, and got my old room back to once again download data, upload tomorrow’s flight lines, and shove microwave burritos down my throat before passing out.


The orthoimage from the lower block superimposed in Google Earth along with the flight lines.


You can compare my image (here) to Google's above, using the flight lines as reference.


The numbers at right are the accuracy of camera positions in meters, indicating positional accuracy without ground control. Cant ask for better than that.


The image mosaic draped over terrain (above) and a shaded relief of just the terrain (below).

As predicted, the weather that day was worse and shown every sign of getting really bad.  I had hoped to fly up the Noatak valley to do the sites I had meant on the way down.  The weather there looked like it was going to be really good, but the weather on the south side of the Brooks Range showed every sign of getting bad, and my fear was that I would get stuck on the north side without enough fuel to get somewhere to land.  So I took the coward’s way out and flew up the Kobuk River to Bettles.  Along the way, I was continually squeezed further north against the mountains, as the large weather system brought clouds down the ground to my right, overcast just above me at about 4000’, and clear skies where I really wanted to be to the north.  I tried doing a small block over the smaller dune near Ambler, but the winds were blowing 35 knots and the short flight lines meant rapid and strong changes in pitch and power while trying to stay out of the clouds or get completely pinched but the northward moving fog.  Near the end of it the engine nearly quit while I was messing with the mixture and I decided I’d had enough fun and beelined it straight for Bettles. 

The Kobuk Valley is really beautiful and I took some time to look out the window on the way there.  One thing I found a bit distressing was how many trees there are, as in how few places there are to land if need be.  As I continued east, I was eventually pushed up against the mountains by the southern wall of clouds and started getting into rain and snow showers.  My plan was to cut the corner if possible and go straight to Fairbanks, but there was simply no possibility of that, and at this point I was worried about whether I’d even make it to Bettles.  As I was now out of room to the north and getting forced lower to the ground by the clouds above, I texted Kristin via satellite to find out current conditions in Bettles.  It was still good there, so I pressed onwards.  I remember coming through a squall and being face to face with a float plane going the opposite direction.  Fortunately he saw me at the same time and we chatted a bit, him getting the bad news and me getting the good.  But by this time we were getting bounced around quite a bit and things were changing rapidly enough that I wasn’t sure of anything.

About 15 miles northwest of Bettles I called in to say I was heading that way.  I had just come out of a squall and hadn’t heard any radio traffic for a while and wasn’t exactly sure I had the right frequency.  But a few seconds after I called in, three other planes called in to say that they were also 15 miles northwest and a fourth was just a bit further south.  We didn’t see each other, but stayed in contact well enough to avoid each other, and about 5 miles out of Bettles the weather broke to light winds and almost clear skies, and soon we got in line to land, which we did in rapid succession.  I refueled as quickly as could, checked the weather in Fairbanks, and headed off again.

I departed about 10 minutes after two of the cubs that landed with me did and a few minutes before another Cessna took off.  We were all heading to Fairbanks.  I made it to the road first, where ceilings had dropped substantially and we were getting a strong headwind.  The other Cessna eventually passed me, then passed the cubs, and was the first to make it to the Yukon River bridge.  He reported back that the clouds were down to the ground on the other side of the bridge and said he was going to try to follow the river under the clouds.  I wasn’t far behind him, having also passed the cubs by now, and started to follow him.  The Yukon River enters a canyon about here, where all of the river is enclosed between high walls.  The clouds were obscuring the tops of these walls, and after a minute or so of feeling like I was committing suicide I turned around and reported back to the cubs.  They were in no mood for that much fun either and by then had found a stretch of ground to set down on to consider options.  By then I had caught back up to them and watched them land.

One of my other rules of thumb up until now was not to land off-runway until I had finished this mapping project to avoid the risks associated with that.  Well, I had now finished that project, and they reported back that the surface was fine and the winds calm where they landed.  So I went around and set up an approach.  Everything went fine with the approach, but as I got closer I realized that I was way too high.  So I slipped down a bit, then slipped harder, then slipped even harder.  Eventually I was set up again and stopped slipping, but then I was too high again.  As I got close to the ground on short final, I knew that there was a good chance this would work out OK, but knew the smart thing to do was to go around and try again.  So I firewalled the throttle and prepared to shed flaps.  But I had been at idle so long that the engine did not respond immediately, so I pulled it back out, knowing at that point there was no time left to wait.

In that instant I knew that I had already made my first mistake, waiting too long to go around, but more importantly that now was not the time to make a second mistake.  Here is where those hours upon hours of going sideways on Texas runways really paid off.  There was no panic or fear, there was just focus on flying the plane.  Power off, no new flaps, yoke full back, stall horn screaming, tailwheel drags, mains come down with no bounce, stand on the brakes, keep it straight, keep the rubber side down, and take my lumps.  The tailwheel touched on the last hundred feet of the clearing and the plane stopped with just enough room to turn around before really getting into the weeds.  It was probably the smoothest, shortest, and best landing I had done in the plane, especially considering I was near gross weight with all that fuel on board.  By then the guys in the cubs were already out of their planes and standing only a few feet away, and I remember thinking that these guys are going to think I was a total putz or a total stud.  So of course I meant to land like that...

Of course what had happened was that we all landed downwind.  The cubs started much lower to the ground where the wind was weaker and being cubs they can do what they want.  I started much higher and it never occurred to me to consider whether I was landing downwind because I had just watched them land in that direction, and they had reported back that winds were calm, which they were at the surface.  I came away with two things from that experience (well three if you include never follow cubs…).  First is to never stop flying the plane.  It was something I always knew was true, but this was really the first time in the realworld that it was essential and where I fortunately it seemed to be my natural reaction.  Second was that by being forced to push the envelope I realized how far away from that envelope I had been, and in the past fews days have practiced such short landings extensively and have gotten comfortable with doing it reliably.  Both of these relate to what I think is one of the most important skills in flying, and that is having an accurate assessment of what your skills really are, such that you can avoid getting into situations that involve exceeding those limits.

In any case, we sat around a while hoping the clouds would lift and I checked weather with the sat phone.  Our fear was that if we waited too long that Bettles would go down and we would be stuck here for a week.  So we took off (into the wind…), and headed back.  The strong tailwind got us back in no time.  While I was checking weather in the lodge, I bumped into a few friends from NPS, the very ones that funded this project, who had just returned from a float trip on the Kobuk River.  So we spent the evening talking about science and future projects before getting the longest night sleep I’d had in a few days.

The next morning the weather in Bettles wasn’t nearly as bad as forecast and the weather in Fairbanks was even better.  The problem was still at the Yukon River bridge, were weather cams showed it still down to the ground.  By now there were four of us in the same boat, the two cub guys from yesterday, Phil and Walt from King Salmon, and Tom in a champ from Homer, all trying to get south.  We talked with all of the commercial flights landing there and the news was never good – there was simply a fog bank over the river that stretched for miles in all directions.  We swapped a lot of airplane stories and largely kept trying to convince each other that patience is the best virtue for pilots of tiny planes, but we all really wanted to leave.  Depending on the pilot, we got reports of the fog bank being 10-50 miles long or wide.  We discussed flying over the top, flying around, flying commercial, but by 5PM it was time to make a decision.  No one wanted to be the first to say let’s go in fear of being judged too eager to die, but I said I was going to go have a look as there was nothing to lose but some fuel in doing that.  So in a few minutes we were all airborne, heading southwest to get around the fog by way of Tanana.

We stayed in contact along the way, but we were all flying different speeds, with me the fastest.  So eventually we were split up, each trying a slightly different route.  The issue was that the clouds covered the hilltops and some of them were raining or snowing.  I opted to go the long way around the west side of the mountains and through the valley on the other side, and this proved the wettest way.  Visibility was poor, but looking out the side window was not bad.  I had four GPS running and they all said more or less the same thing, and I was visually following a river within a wide valley that I knew led someplace good .  So it was intimidating but reasonable. The others poked through the mountains at different places, and maybe even went to the east where there was blue sky showing.  In any case, we all eventually popped out the other side over Tanana, where the huge valley was nearly clear all the way to the Alaska Range.  Tom decided to head straight for there rather than go to Fairbanks and I last heard from the others as they were landing in Manley to splash in some fuel before continuing on to Fairbanks.  I went straight there, eyeballing a huge, dark, angry looking rain squall heading towards town.  I skirted it by cutting over the hills behind the Minto flats and soon enough landed and pulled into my tie down spot.  The winds hit a few minutes later, and by the time I was in my truck driving away it seemed I was in a small hurricane.  I never heard from the cubs again, but I suspect they gave that squall one look after departing Manley and followed Tom.

I got back home and Kristin gave me my mail.  As I was describing the trip, no sooner did I say “On this trip I became a commercial pilot” -- I had just completed my first commercial job and it required demonstrations of nearly all of the skills that commercial pilots need to have – than I opened a letter from the FAA and inside it was my official plastic commercial pilot certificate.  It was a fitting end to a year’s worth of enormous struggle and ultimate success.

I remember looking in the mirror after that and thinking that I looked 10 years younger.  I certainly felt 10 years younger.  I have achieved a lot of goals over the past 20 years as an Arctic scientist, and most of them took an enormous amount of hard work, but none of them were as intense or as satisfying as completing that job.


Those 3 days took a full year to prepare for...

Matt finds a new career
8 October 2013

But, especially with that success in hand, I still had more work I wanted to do. The data I acquired on that first trip was great and processed as well as I hoped. Now I wanted to refly some of the sites we had done in 2011using my new system, as the results would be better and they could be used to improve the 2011 data. I also wanted to map McCall Glacier and some of the glacier around there, as that was my underlying motivation for all of this work in the first place. The plane was still running great and my success on my first trip gave me renewed motivation for doing more. The biggest unknown at this point was that I still didnt know whether I was just making pretty pictures or whether my data was as accurate as I had hoped it should be, so I was on the lookout for sites with high-quality data that I could also map and compare to.

By now it was early September and the Aleutian lows began ramping up for their winter onslaughts. The big system that was predicted to shut down the state while I was to the west never really materialized, both others were already threatening. A week after my return, September 10th, a window opened up to the north that appeared like it was only going to last for a day, so I prepped the plane the night before and headed out early the next morning to conquer my personal quest to map McCall Glacier on my own. Not coincidentally, I had several years of reasonably high-quality lidar data from these glaciers so I would also have a great comparison to make to test the accuracy of my system.

The flight went well. I skirted some low fog over the Yukon Flats and refueled in Arctic Village. Here my timing was poor because a Wright's caravan was landing at about the same time, meaning that half the village was at the airport on their four wheelers. I pulled over the FWS tanks to poor my jugs in, and before long inevitably one of the locals drove over with a menacing look. I said hi as I was fueling, and he asked abruptly whether I worked for FWS. I said no, I work for the University. He asked what I was doing here and I told him that I came to take pictures of the glaciers. This apparently caught him off-guard and he laughed, totally changing his demeanor. By way of not-really-an-apology he said that they have a lot of thefts on the runway, and that white men steal. I told him by way of I'm-not-just-any-white-guy that all people steal, and he accepted that without further comment, and we had a pleasant chat about airplanes, glaciers, etc while I continued to fuel. He was an older guy and had flown 170s as well, and he gave me his phone number and an invitation to stay at his house if I ran into any trouble with the plane, which fortunately I didnt.

The Brooks Range to the east of my sites was in the clear, but my sites were just on the edge of the clouds. I was able to sneak around to get on top of the glacier, but it was clear my other sites were not a possibility. Even here the clouds surrounded the glacier and caused me to shorten my flight lines to avoid them and by the time I was finished with the north-south lines there were random clouds floating over the glacier. So I did some crossing lines in between them and headed over to Esetuk Glacier to see if it might have a hole over it too. It sort of did, but not really, so I eyeballed a few lines over it and scrammed back to the east and into the clear and back to Arctic Village for another splash of fuel, before heading home. It was a long day, but again after landing I had that same huge sense of relief that I just accomplished something that I had wanted to do for years and that sense of happiness and hope that this was something I could do again in the future.

The news got even better though. I processed the data the next day and all of the internal metrics of accuracy looked great, as did visual inspection of the resulting map. So I held my breath and subtracted it from my earlier lidar. When the results were done and the image appeared on my screen, my jaw dropped. The parts of the scene that should not have changed, that is the non-glaciated mountains, showed exactly no change and the changes on the glacier popped off the screen. I remember saying "Ho-lee-shit", and for the next hour being able to say nothing else. These results demonstrated that my $3000 camera produced results as accurate as the $1M lidar that produced the comparison data, and my data were 16 times higher resolution and came with a perfectly registered image mosaic to drape over the top! This comparison was the first time that I really could determine whether or not I was just making pretty pictures or creating something scientifically useful, and these results exceeded my hopes and expectations. And especially considering that I use no ground control to make it and it dropped into the right place in the realworld, my vision was vindicated -- no longer would we have to land to make ground measurements to align these maps, I could just fly wherever I wanted to start taking pictures and the result would likely be more better than anything ever made in Alaska! I cried.


A long day trip, but worth it.


I sent this image out to a few friends the night I got back. Several commented on what nice weather it was. I take that as a great compliment, since this is not an oblique photo take from the plane, but rather my orthophoto map draped over topography and a fake sky in the background.


The orthophoto mosaic, looking down.


The topographic map I created, with the red indicating high elevations and blue low. The glacier flows from right to left

.
Color here represents height.


The difference image with lidar, the glacier flows to the left and the terminus, showing the most change, is in blue. The deep red areas are rock, which indicate little to no change, as expected.


A vegetation plot in the Sheenjek River that I mapped on the way back from the glacier for a friend.


A shaded relief of the topography.


Here's a pingo with spruce trees on it.

With that success in hand, I began planning to re-aquire the maps in the Noatak valley that we made in 2011.

The idea here is that because my current methods are so accurate that I could use them as ground control to improve prior imagery and bring them up to the same standards. Again I waited for the weather and about a week later, September 16th, my window opened up again and I went for it. By now daylight was getting shorter and my concerns about flight duration were more than just my bladder capacity. I left as early as I could, waiting for the weather cams in the area to give me some indications, but it was a beautiful day across that whole part of the state without a cloud in the sky. I splashed some fuel in at Bettles from their tank, then a bit more at Dahl Creek from my jugs, and headed north into the Noatak valley. Here I found some low clouds that concerned me so I flew the lowest passes to get a feel for them in case the weather forced me to use them on the way back, but in the end they never got in my way. More of a concern was avoiding topography. My goal here was to make the highest resolution maps I could and this meant flying low. The target sites were all on flat lands, but the mountains rose abruptly at the edge of them, so I took a lot of care making my turns, mostly by not getting too distracted with dealing with the camera system and nav. The lines all went well and soon I had finished 4 of the thermokarst sites and one large block. By now it was about 4PM. I had enough fuel to get to the other large block and back to Dahl Creek, but I knew it would take at least 2 hours more to finish up and by then I would for sure be flying home in the dark or spending the night somewhere. Kristin was on call starting that evening so I opted for heading home and picking up anything of interest on the way back.


Almost all the way back to Kotzebue for this one.


My orthoimage draped onto Google Earth.


Lines up pretty well with Google's imagery. Actually mine turns out to be more in the right place than their's.


Here we found thaw slumps over ice rich ground. Those are bushes in the river bottom for scale.


The headwall here is several meters high. You wont get this kind of detail from any other maps available in Alaska, and certainly not updated regularly.

The valley was filled with caribou.

There must have been a thousand or more there, in groups of 50-100 or so. While in Bettles on my way back from Shishmaref, my friends at the Park Service mentioned wanting to find a new way to count caribou. I told them that I thought it should be possible to measure them topographically -- that is, make topographic maps at sufficient resolution to actually resolve their bodies, then it was just a matter of some programming to count them automatically. They thought it was a great idea and suggested putting together a pilot project. My experience with writing scientific proposals is that it is always best to have done nearly everything you said you are going to do to reduce risk, which kind of means doing the project before its funded. In any case, I had the opportunity to demonstrate whether this would work or not, and flew some lines by eye over a lazy looking group of 50 or so, completing our pilot project and giving me something to show off if it worked. And it did, so I did.


Yes, these are caribou.


Here the colors represent slope, and swirly blues and greens are polygonal ground with ice wedges. The reds are caribou.

The trip home was reasonably uneventful. I decided to skip Bettles as I had plenty of fuel and cut the corner towards Fairbanks. If there was no haul road, I would have flown straight back. But something about knowing there is a road to follow and not following it made me uneasy. This part of the landscape also was completely tree covered. The issue here is that while the plane was flying great, there was always the possibility that the engine would stop and not restart until much later, so always having a place to land in sight is a comforting thing. In the end, I didnt cut the corner that hard and given all of the stress I would have been better off flying my usual route towards Bettles where I knew I could land on most of it, as it would only have added another 20 minutes to the flight. But in any case, I hit the road before the sun set fully, climbed up to 10,000 feet, and flew south as fast as I could.

I would have stayed the night in Bettles or somewhere if it had not been a cloudless night and a full moon. This was the first night flight I had done since being in Texas, where there are tons of street lights to provide a ground reference. Here all there were was the occasional truck headlights and the faint outline of the road in the moonlight. But all went well, though at one point I was certain I was about to have a head-on collision with an approaching plane, which turned out to be truck headlights 30 miles away... Though I didnt map as much of the area as I wanted it was nonetheless a successful flight and gave me enough data to validate my idea of using my current data to improve my prior data.

After sharing a few of these results with local colleagues and friends that need high resolution DEMs and time-series of them just like I do, I started picking up new projects. One of these was to create a base map of a section of the Wind River near Big Ram Lake, about 30 miles west of Arctic Village. Again the Aleutian Lows dominated the State's weather, but a narrow window opened up and I went for it. It was a beautiful day to the North and almost the entire Brooks Range was in the clear. As I talked with pilots flying around the area, however, I heard that Arctic Village only had a 200' ceiling, which is way too low for me to land. Here I had planned to add fuel so that I could spend as much time as needed flying the mission. But I had about 2 hours extra fuel to play with, so I went directly to the site. Here that same low cloud layer was stretching over from Arctic Village, and only 5 miles further west it was clear. The ceiling here was about 4000' and I had planned the flight lines in the wide valley at 4000' as well, so I poked in there and gave it a try. The mountains surrounding me were obscurred above me, but there was plenty of room to maneuver and the major challenge was staying well enough below cloud height but high enough for the flight lines. Another issue was that there was a fresh snow layer here, unlike the Yukon River valley, and the thick overcast created pretty flat light. Flat light is an issue because it washes out the contrast in the snow and these photogrammetric methods rely on sufficient contrast to pick common points between photos. But I figured if it worked here, then it would work anywhere, so spent the next 2 hours flying back and forth and completed the lines that I had planned. I was a bit concerned that the even lower clouds around Arctic Village might move over and trap me in this valley, but every once in a while I poked my nose out of the valley to ensure I still had an escape route. Arctic Village remained down throughout, so when I was done I headed to Fort Yukon to refuel and then on to Fairbanks. It was over 12 hours of flying squeezed within the now shorter northern daylight, but despite the long day I stayed up late into the night excited to see the results of the processing. I was once again amazed at the results -- by using proper exposure and post-processing of exposure, the flat light did not interfere with the quality of the final product, and I now felt renewed confidence that if it succeeded here, it was likely to succeed anywhere.


This site near Big Ram Lake was about the worst possible conditions for photogrammetry -- fresh snow on flat ground under a thick overcast. But it worked!


The spiky things are black spruce.

Over the next few weeks I continued to improve my post-processing techniques and resumed flying local test sites to further test accuracy. The most important of these was the Ft. Knox goldmine. It was only about 20 minutes from the Fairbanks airport and provided all of the characteristics of an ideal test site because it had steep topography along its open pit, but this topography had benches wide enough to drive big trucks on. The basic idea is that I would map this several times and subtract them. If my results showed no change in topography in places they were not mining, I would know that my results are accurate. Though I did something similar on McCall Glacier, the steep mountain slopes had no benches in them, and these benches are useful because they are wide enough to image well and would show me any subtle offsets along the way down, and I also had the feeling that a comparison of my results would show even better accuracy than comparing my results to lidar. I flew the first one September 30th with Turner as my camera operator and then another on October 6th. The first processed great, as did the second. So once again I held my breath when I hit the button to subtract them -- this was the really big test and if this worked I would not only convince myself but could convince any one else interested in such data. The results floored me once again. There was essentially no change over most of the scene, but everywhere the earlier images showed unexploded blast fields, the new images showed excavators and enormous change. I was once again getting differences under 20cm for the areas of no change and the two maps aligned perfectly with no tinkering or use of ground control. I could not be happier or ask for more from such data. I dont know of any system or method that could produce higher accuracy or precision results from the air and I have a strong suspicion that my system is better than any other that exists currently for applications north of treeline. Even in the trees results were amazing -- in the week between acquisitions most leaves had fallen off the birch trees and I was actually measuring the difference in canopy height caused by this.


Ft Knox is an open pit gold mine. Here is the pit, it's pretty massive.


Here's part of the pit on September 30.


Same day, just without the image overlaid.


A week later. Download these two and flicker between them, it's pretty impressive.


Here's the difference image. Everything in green did not change, highlighting the shot field at top and the earthwork at bottom.

It doesn't get any better than this with any mapping technique, even at 100x my cost. So here I am a little over a year after I took my first flight lesson, looking back on everything it took to get here. Part of me wondering what the hell was I thinking, but part of me is still looking at my results with a clear answer. Of all the things I've done scientifically, this was by far the most effort I put into preparing for a few days of field work, and I guess as a result it was the most satisfying after having succeeded. In the process I learned a tremendous amount about flying, weather, photogrammetry, legalities, people, Alaska, and ultimately about the impacts of climate change on the landscape. I also learned a lot about myself. Among many other things, I learned that making maps like these from the air is perhaps the only thing I've ever done that I could envision doing for the rest of my career. In field science, one of the basic choices is whether to specialize in a technique (and then use that technique on a variety of locations) or specialize in an area. Most opt to specialize in a technique, in many ways it is a much safer approach in terms of funding, it adds diversity and interest, and eventually they become known as the guy (or girl) who does X. I chose to specialize in an area and tried to approach it from all angles, largely because I start feeling a bit claustrophobic when I do the same thing over and over, and I also seem to have some fundamental aversion to being labelled as being anything but me. Only the future will tell what will happen next, and maybe once others start doing it too I'll move on again, but for the moment I would feel happy to become known as the guy that makes awesome maps of the Arctic.


Not bad for a year's work, starting from scratch without a pilot's licence, plane, or mapping system. We'll have to wait and see what the future will bring.

(c) 2010 Matt Nolan.