Towertalk
[Top] [All Lists]

[TowerTalk] Where is that Perfect HF DX Hilltop

To: "towertalk@contesting.com" <towertalk@contesting.com>
Subject: [TowerTalk] Where is that Perfect HF DX Hilltop
From: Shawn Donley <n3ae@comcast.net>
Date: Thu, 10 Sep 2020 14:08:37 -0400 (EDT)
List-post: <mailto:towertalk@contesting.com>
With the sunspots down, work slow and COVID keeping me home, I had some time to 
start a project that I’ve been thinking about for a while.  I've always been 
interested in tools to discover locations for an exceptional HF DX QTH, one 
with terrain profiles that put a smile on your face when you run HFTA.  So I 
wrote some code to help find such locations.  The code basically extracts data 
from two VERY large databases created by Andrew Kirmse and the late Edward 
Earl, accomplished mountaineers, mathematicians and software developers.  
Andrew and Earl did the heavy lifting, writing sophisticated code to extract 
key topological measures from the Shuttle Radar Topology Mission's digital 
elevation data, with voids filled in by other means. The key characteristics 
contained in these databases are the Topological Prominence and Topological 
Isolation of a peak.  Rather than try to explain these topographic functionals 
here, look them up on Wikipedia if you're interested.   Larger Prominence is 
better.  Larger Isolation is better.  A combination of both can be better yet.  
 Are these two topographic  functionals the most useful characteristics for 
finding an exceptional HF DX location?   They are a good start but something 
called Omnidirectional Relief and Steepness (ORS) would probably be better.  
But computing that functional is a project for another day.

My code (in Matlab) generates an Excel file containing hilltops (or mountain 
tops) that have both Prominence and Isolation values within a given search 
area, sorted in decreasing Prominence.   The Prominence database has all peaks 
in the world with a Prominence of 100 ft or more. The Isolation database is 
even larger.

So will this data immediately point to "The Perfect Spot?"   No, but it will 
pare down the search.  Discovered peaks may have excellent low takeoff angles 
in only some directions, provided you can get close to the start of the 
downslope. Then there's the aspects of soil conductivity, background noise 
floor, access, neighborhood HOA's, CC&R’s, etc.   As I made some trial runs, I 
discovered many of the better spots already occupied by commercial towers, 
which in retrospect should have been no surprise.

I can highly recommend using http://www.heywhatsthat.com/ for a more detailed 
examination of a particular site of interest.  On that web site, you can 
generate a panorama but more importantly see the viewshed of the peak and 
generate plots of the terrain profile at various azimuths around the peak.  
Take some time first to read the FAQ at http://www.heywhatsthat.com/faq.html   
The site has a lot of features.

My pass-fail test for the current beta version of the code was 'discovering' 
the location of the legendary DX'er W3CRA (SK) in Canonsburg, PA.  It does.  On 
Google Map’s satellite view, the remains of his Yagi can be seen 485 ft to the 
northwest of the peak’s position (Hanna’s Knob) on the downslope.  Frank was 
well known for his booming signal into Asia when no other USA stations could be 
heard.

An example of the program's output is the link below. The search area for the 
example was a 20 mile box around W3CRA.  Frank's location is in red font in the 
spreadsheet and shown as the black square with a data tip on the map sheet.

https://www.dropbox.com/s/c3osexf6mginsm7/20%20Mile%20Box%20Around%20W3CRA.xlsx?dl=0

If anyone would like me to generate a spreadsheet and email it, please contact 
me OFF-REFLECTOR.  I can generate data for any US State or a country, but 
that's likely too much data to be useful.  You can define a smaller search area 
by supplying the latitude and longitude for the vertices of a polygon 
surrounding the area of interest, like a triangle, square, rectangle, 
parallelogram, etc.  That data should be supplied in the format below, 
including the square brackets.  That will make it much easier for me to cut and 
paste into my input file.

Latitudes = [LAT1  LAT2  LAT3 ... LATn]

Longitudes = [LONG1  LONG2  LONG3 ...LONGn]

Where LAT1 & LONG1 are the latitude and longitude of one corner of the polygon 
search area, LAT2 & LONG2 the next corner, etc.   The lat/long pairs can be 
supplied in any order as long as the resulting polygon is convex (doesn't 
dimple in on itself). For example, the two vectors used for the rectangular box 
in the W3CRA example were:

Latitudes = [40.389332 40.389332 40.099761 40.099761]
Longitudes = [-80.369059 -79.988657 -79.988657 -80.369059]

Northern hemisphere latitudes are positive.  Southern hemisphere latitudes are 
negative. Longitudes west of the prime meridian are negative.  Longitudes east 
of the prime meridian are positive.

Perhaps this tool could prove useful to someone or maybe just a curiosity to 
spend a rainy day looking for  the “perfect spot.”

Have fun and good DX.
N3AE


_______________________________________________



_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
<Prev in Thread] Current Thread [Next in Thread>
  • [TowerTalk] Where is that Perfect HF DX Hilltop, Shawn Donley <=