feb_temp.jpg

Just UT date in Feb 2007 vs ambient temp (weather station 1) in
degrees C.

----------------------------------------------------------


feb22_temp_t.jpg, feb23_temp_t.jpg, feb25_temp_t.jpg 

How this works, from top left across and down:

- Have model from Haslam, data from antenna (plots 1,2)

- fit each in a "good" range the normal way (polynomials plotted
  there, 3, 4)

- get G_0 and T_sys as before, see the bad fit (plots 5, 6)

- now read in the temperature data (plots 7, 8; filename gives
 antenna, polarization, day)

- Use T_sys and G_0 as before for initial guesses to curvefit, fit
 f(t) = (G_0 + G_1 * T(t)) * (Tsys + T_Has(t)).  this fit over entire
 LST range

- Apply fitted G_0, G_1, and Tsys to data and compare to original
model (plot 9)

For feb22, get G_0,G_1,Tsys= 18242.128, -348.35042, 12.520694
For feb23, get G_0,G_1,Tsys= 18684.362, -241.87101, -45.318575
For feb25, get G_0,G_1,Tsys= 15742.668, 99.564523, 20.942902
  but note temperature sensor problem?