The streaking is due to buildings, actually. I’m in the city, so it turns out 100 story buildings downtown, as well as 15-50 story residential towers, can block out good parts of the skyline. Those long sunburts roughly correspond to “holes” in the surrounding cityscape, such as where I can see clear to Lake Michigan.
I put the scripts up here now, as they have been cleaned up and tweaked a bit. It now uses a better map projection, and should be more generic as far as what data file it can take in. It still expects SBS input format, but should be easy enough to make it also accept a 3 column set of Lat, Lon, and Altitude. Pull requests definitely welcome.
30303 according to BartJr’s post earlier, using nc to capture the traffic. Script could also be modified to take in two input files and plot them in different colors, for purposes of doing ADS-B + MLAT, or seeing how two sites overlap in coverage, etc.
Oops, the default piaware dump1090 beast input port is actually 30104, I shouldn’t have written 30004. If you set it to 30004 you might not see mlat planes in dump1090.
To answer your question, yeah, that command just adds another output on mlat-client, in basestation format, on port 30303 (you can of course pick any port)
Could that be why I’m not seeing mlat planes anymore on the web view (view live data link)? How do I undo this? Do I just execute the same command with 30104 instead? (and restart, of course)
import copy
import matplotlib as mpl
import matplotlib.pyplot as plt
from math import pi, tan, log
d2r = pi/180
lat = ]
lon = ]
lon0, lon1, lat0, lat1 = (-1.75,-1.6,0.47,0.6) #map extent in radians
imgsize = 1000
for l in open('foo.latlon'):
a = l.strip().split(',')
lon.append(float(a[1])*d2r)
lat.append(log(tan(pi/4 + float(a[0])*d2r/2)))
#remove axis and border
fig = plt.figure(frameon=False, figsize=(1,(lat1-lat0)/(lon1-lon0)))
ax = fig.add_axes([0,0,1,1])
ax.axis('off')
#with a Log histogram, zero-value pixels are unpainted (white). So, copy the color map and set bad pixels to black.
cm = copy.copy(mpl.cm.get_cmap('cubehelix'))
cm.set_bad((0,0,0))
plt.hist2d(lon, lat, range=[lon0,lon1],[lat0,lat1]], bins=imgsize, norm=mpl.colors.LogNorm(), cmap=cm)
plt.savefig("foo.png",dpi=imgsize)
#plt.show()
If you look for MLAT flights in that area you can usually spot them during weekdays. They all have US military ICAO addresses. The few I’ve googled usually come up as Beechcraft T-6 Texan II trainers, presumably out of Randolph or Lackland AFB.
I never really noticed before now that each aircraft appears to get its own little practice area. Usually when I catch them on the dump1090 live map, they’re flying low enough and maneuvering so much that the MLAT paths just look like a scrambled mess. But as a scatterplot it makes for a neat pattern!
I wonder if we can also pick up the sky divers north of here. I think they’re somewhere between Austin and Waco. Probably multiple operations, but I always notice the one right off I-35 when I go to Dallas.
