Mifflin Airport Review – Take II

Today (and quite possibly tomorrow as well) is a rain day here at the Mifflin Region 2 Soaring Contest, so Karl Striedieck was gracious enough to sit with me and go over the many airports/airstrips in the current (2014) database to cull out the bad and verify the good ones.

First, we defined the boundaries of the Mifflin contest soaring area, as shown in the following Google Earth (GE) image.  As you can see, the boundary covers everything (and a bit more) that we normally reach even on the best contest days, and excludes a lot of airports/airstrips that are more normally associated with the MASA (Fairfield, Pa) contest area.

Mifflin Soaring Area, Take II

Mifflin Soaring Area, Take II

Next, Karl and I went through every airport/airstrip inside the new boundary that I had previously assessed as ‘problematic’ in one way or another (missing entirely, too short/narrow, wrong coordinates etc) to see whether or not it should indeed be removed from the database.  Of course if the airstrip under consideration was outside the newly defined soaring area it was eliminated without further consideration, but those inside the area had to be checked out.  Of those, there was local knowledge that caused 10 to be retained as ‘good’ airstrips (more about these below).

The next step was to integrate my initial GE review and local knowledge input into a database that John Leibacher can use to produce an updated soaring database. Following the procedure developed during my Perry airport review (see https://soaringcafe.com/2014/03/perry-wagener-sc-airstrip-review-part-ii/), I started by importing the 2014 ‘mifflin4.csv’ file (Comma-separated-variable) into Excel.  I prefer using an Excel file because Leibacher can handle it directly, and because there is a converter (https://www.earthpoint.us/ExcelToKml.aspx) that can convert a properly organized Excel file to KML, which can then be viewed in Google Earth (GE) for round-trip verification purposes.

The Excel file started out with about 170 waypoints, 45 of which are the official contest control points, leaving about 125 ‘auxiliary’ airports/airstrips. Of those 125, Karl and I were able to eliminate 30 or so ‘inside’ airstrips as visibly unsuitable.

Next I re-alphabetized the ‘auxiliary’ portion of the list, as the original was pretty badly organized.  I suspect there were at least three different lists that were concatenated over time to form the current database, and nobody took the time to integrate them properly.  After sorting just the aux waypoints, I then used some Excel magic to sequentially number them starting at 100 (to make the transition from control points to aux points obvious), and saved the result as ‘mifflin4_gfp3.xls’.

At Karl’s suggestion, I added a new waypoint called ‘Raystown LF North’.  This is the emergency landing field developed by Karl and John Good (rumor has it that this ‘development’ involved a midnight rowboat ride and a chainsaw) about halfway between the south entry point to the Raystown Dam ridge and Raystown Dam itself.  Itis very short (about 800′ long), but it’s a lot better than the alternative of a water landing if you fall off the ridge there!  I also renamed the original ‘Raystown LF’ to ‘Raystown LF South’ to be consistent.

Both Raystown Dam Landing fields

Both Raystown Dam Landing fields

Raystown Dam LF South - plenty wide and long enough

Raystown Dam LF South – plenty wide and long enough

Raystown Dam LF North.  VERY short, but beats a water landing every time!

Raystown Dam LF North. VERY short, but beats a water landing every time!

 

Next I corrected a number of problems where the FAA ‘official’ coordinates for an airport are incorrect – notably Shulls, Homan, Snook, and Tallman West. This can be a real problem, because Leibacher’s processing assumes the FAA coordinates are correct, and in the real world they sometimes aren’t.  In the case of the Homan airport (about 7 miles south of State College and slightly north of Rt 45), there was obviously nothing but trees at the ‘official’ coordinates, but Karl swore the airport really was there.  We searched and searched in GE, but couldn’t find it, even though Karl had seen it from the air and even landed there before.  Finally in desperation we called the phone number shown in the AirNav listing for Homan, and got Mr. Elwood Homan himself.  After chatting with him for a while (turns out Doris Knauff is his sister-in-law!), he was kind enough to talk us through finding his place, well over a mile from the waypoint coordinates.  So, the secret to finding Homan (other than using the new coordinates) is to find the Agri-Center, then find the small town of ‘The Meadows’ just to its north, and then find the large barn and Quonset building on the north side of the small two-lane highway (rt 3024) leading out of The Meadows to the west.  It’s the first set of buildings west of town, less than 1/2 mile from town center.  The result of this step was saved as ‘mifflin4_gfp4.xls’ (in case you are wondering, the reason I do incremental saves is to make recovery from my inevitable screw-ups easier).

Next I went through again and eliminated all the airports/airstrips outside the defined area, and re-numbered the survivors.  After this, I was down to about 60 auxiliary airstrips that are located within the confines of the KS-defined Mifflin soaring area and are wide enough for an 18m ship (as far as I can tell from GE imagery). This result was saved as (mifflin4_gfp5.xls).

The next trick was to round-trip verify the contents of mifflin4_gfp5.xls to make sure I haven’t moved a waypoint into the Atlantic or Pacific oceans, or failed to remove an airport that’s not in the soaring area.  The following screenshot shows the results of the mifflin4_gfp5.xls-to-kml conversion, and it is clear that I didn’t remove one out-of-area waypoint (Shreveport N).  A spot-check of other waypoints showed them all at the correct coordinates, so it looks like the updated waypoint list (with Shreveport N removed) is ready to go to Leibachers.

Results of round-trip verify step.  Note Shreveport N not removed.

Results of round-trip verify step. Note Shreveport N not removed.

 

The last thing to do before putting this into John Leibacher’s capable hands is to correct the coding for 09 Mill Run and 02 Orbisonia.  These waypoints were inadvertently coded as airports (I think this is due to a buglet in John’s processing algorithm) when in fact they aren’t landable at all.  So, I have recoded them from ‘TPA’ to just ‘T’.

OK, that’s it – off to Leibacher, and knowing him there will be a new Mifflin contest database available in the next day or two.  Stay tuned!

Frank (TA)