Window Correction Table
for TexAQS

The following analysis documents my determination of the Window Correction Table, WCT, for the MTP/DC8 when it was used during TexAQS.

First I will present a comparison of Electra navigation temperatures with RAOBs.  Next, I present WCT graphs for each flight that is useable for this purpose.  In a final section I will search for patterns that might reveal whether or not low altitude WCTs can be used.

Calibration of Electra Navigation Outside Air Temperatures

The ferry flights are ideal for calibrating the navigation outside air temperatures, Tnav, using RAOBs.   The RAOBs are interpolated in time and altitude.

Figure 1.  Comparison of Tnav with Traob for 5 events.

As the graph shows, there is good agreement with RAOB temperatures.  The formal result is

    Tnav - Traob = +0.22 +/- 0.47 K

This result is close enough to zero that no correction should be used at this time.


This was the ferry flight from Denver to Salina, KS and on to Houston, TX (Ellington Field).  Three RAOB sites were useable:  DEN, OUN and FWD.  The altitude of the NCAR L-188C Electra, hereafter referred to as the Electra, was 6.41 km, 6.11 km, and 6.11 km for these 3 RAOB sites.  Since the ferry flight altitudes are higher than for the science flights, we may end up relying upon them more than the science flights.

Figure 2.   Channel 1 WCT based on TX000816.

Figure 3.  Channel 2 WCT based on TX000816.

Figure 4.  Channel 3 WCT based on TX000816.

In every case the shape of the WCTs is the same (for a given channel).  This implies that the WCT shapes are approximatley correct.


This is the ferry flight at the end of the mission, and goes from Houston to Denver.

Figure 5.  Channel #1 WCT for TX000913.

Figure 6.  Channel #2 WCT for TX000913.

Figure 7.  Channel #3 WCT for TX000913.

For some reason this flight seems "better behaved" - meaning that the WCT is closer to the vertical zero line for all channels.  It's difficult to achieve "close to zero" WCT profiles, since every shortcoming in the analysis prcess will introduce departures from the zero line.  However, unless a reason can be found for giving more weight to the second ferry flight in relation to the first, I would propose to merely average the 5 WCT profiles for the ferry flights.  Because all the science flight WCT estimates will be suspect, due to their low altitudes, it is worth presenting the best WCT based on the ferry flight data.  It will be useful to at least compare the science flight WCTs with the more reliable ferry flight WCT.  The best ferry flight WCT is presented in the next section.

Ferry Flight WCT

A straight average of the 5 ferry flight WCT profiles is shown in the next 3 graphs.

Figure 8.  Ferry-flight average WCT for Ch#1.  The 5 individual WCT profiles are also shown.

Figure 9.  Ferry-flight average WCT for Ch#2.

Figure 10.  Ferry-flight average WCT for Ch#3.

Visual inspection of these 3 graphs shows an outlier for Channels #2 and #3.  The same RAOB comparison produces both outliers.  In fact, this one RAOB comparison produces the left-most profile inthe Channel #1 WCT as well.  The suspected outlier RAOB is the very first one, the DEN RAOB at 62.7 ks, which happens to be shortly after takeoff - perhaps before the MTP has warmed up and stabilized.  The following 3 figures show what would happen if this outlier RAOB comparison were omitted from the analysis:

Figure 11.  Ferry-flight average WCT for Ch#1, without the suspect "outlier" RAOB data.

Figure 12.  Ferry-flight average WCT for Ch#2, without the suspect "outlier" RAOB data.

Figure 13.  Ferry-flight average WCT for Ch#3, without the suspect "outlier" RAOB data.

These WCT plots look much better!  If instead of "averaging" the profiles I had taken the median profile, a similar result would be achieved.  This is how I've handled similar situations in the past, but since there could be a warm-up problem with the first DEN comparison, I think it is justified to accept the second set.

The ferry flight recommended WCT is therefore:

          80    -1.08  -0.40 -0.30
          55    -0.40  -0.56 -0.05
          42    -0.54  -0.28 -0.09
          25    -0.20   0.01 -0.03
          12     0.05   0.08 -0.09
           0    -0.04   0.01 -0.00
         -12     0.04  -0.10  0.09
         -25     0.19  -0.38  0.08
         -42     0.77  -0.56 -0.16
         -80     0.57  -0.51 -0.22

The solution


This site opened:  May 28, 2001 Last Update:  May 31, 2001