This web page summarizes efforts I've made to calibrate the various MTP instruments. Radisondes are used to provide the temperature standard, and MMS data is used only as a "transfer standard." Use of MMS temperatures data for calibrating "flight level" temperatures requires that the MMS temperature data be calibrated against RAOBs, and this has been done for several missions. MMS has been found to be "too warm" on three missions (typically by about 0.8 K). Some missions did not have MMS aboard, and for these the "navigation" static air temperature was sometimes used as a transfer standard.
Users who need high temporal resolution temperatures are advized to use the MMS data, provided they apply an appropriate correction to bring the MMS data into agreement with RAOBs. As far as I am aware, I am the only person who has compared MMS temperatures with RAOBs during missions (measurements were made under Ed Danielsen's direction many years ago, which is the only other comparison with RAOBs that I'm aware of). Although the MTP data have poor time resolution (compared to MMS), they at least have the advantage of many man-hours of calibration using RAOBs as the standard, and are thus the best source for accurate in situ temperatures for all missions beginning with AASE II. The MMS system has many man-hours of calibration, mostly for the wind data, and their philosophy is to have their calibration traceable to the NIST standards laboratory. However, the MMS temperatures are consistently warmer than RAOBs, so either something is missing in their calibration procedure or RAOBs in the United States are consistently cold.
The following sections represent all missions on which an MTP instrument have flown, prior to my retirement (with a couple post-retirement missions added). I will describe two aspects of MTP accuracy: 1) flight level temperatures (corresponding to in situ temperatures), and 2) altitude temperature profiles.
This will be a "work in progress" web page, and eventually I hope to describe the state of calibration for each mission.
STEP, ER-2 (1987)
I don't think I ever compared ER-2 T(z) profiles with RAOBs, so I don't know the calibration offset value. This could be done, but at this time I don't think anyone is working with the STEP data.
AAOE, ER-2 (1987)
I don't think I ever compared ER-2 T(z) profiles with RAOBs, so I don't know the calibration offset value. This could be done, but at this time I don't think anyone is working with the AAOE data.
AASE, ER-2 (1989)
I don't think I ever compared ER-2 T(z) profiles with RAOBs, or I don't know the calibration offset amount. This could be done, but at this time I don't think anyone is working with the AASE data.
AASE II, DC-8 (1991/92)
A detailed analysis was performed on three flights from this mission that were especially suitable for evaluating MTP performance. The conclusions can be summarized:
1) MTP T(z) accuracy, when compared with RAOBs, agrees
with pre-deployment predicted performance,
2) MTP T(z) accuracy is <1.0 K across an altitude region 9 to 16 km (for typical cruise flight altitudes),
3) MTP T(z) accuracy degrades beyond the above altitude region to ~2.0 K at 6 and 18 km.
In addition, it can be stated that the MTP/DC8 retrieved T(z) has a lower S.E. than the NMC T(z) for the altitude region 7.5 to 17 km when flying close to RAOB sites, and approximately 6 to 18 km when flying at more remote locations.
The MTP MP-files in the archive (both on the CD and on the ARC archive computer) incorporate all calibration insights that have been achieved to this date (and described in the above referenced web page).
AASE II, ER-2 (1991/92)
I don't think I ever compared ER-2 T(z) profiles with RAOBs, so I don't know the calibration offset value. This could be done, and maybe it should be, since some people may still be working with this data.
ASHOE/MAESA, ER-2 (1994)
I don't think I ever compared ER-2 T(z) profiles with RAOBs, or I don't know what the calibration offset could be. This could be done, and maybe it should be, since some people may still be working with this data.
STRAT ER-2 (1995/96)
I did compare MTP T(z) profiles for some STRAT flights with RAOBs, and I suspect that I have adjusted all STRAT MP-file data to be compatible with RAOBs. I could verify more of this if it would help a "user."
TOTE/VOTE, DC-8 (1995/96)
Another web page describes a detailed analysis of MTP performance based on 55 comparisons of MTP-retrieved T(z) with RAOBs from the following three missions: AASE II, TOTE/VOTE, SONEX. Figure 3 in that web page shows "altitude profiles for RMS accuracy of MTP's T(z)" for each of the missions separately, as well as a weighted average performance profile. Since there is no reason to believe that the MTP/DC8 performance characteristics changed between these missions, the weighted average performance profile should be an accurate appraisal of inherent performance:
1) RMS performance is approxiamtely 1.0 K from 8
to 13 km,
2) RMS performance degrades to 2.0 K at altitudes 6 and 17.5 km,
3) RMS performance degrades to 3.0 K at altitudes 4 and 20 km.
SUCCESS, DC-8 (1996)
POLARIS, ER-2 (1997)
Another web page describes the in situ temperature accuracies for MMS and NAV: "MMS & NAV OAT Corrections for POLARIS and SOLVE" I conclude that MMS temperatures are too warm by 0.84 K [for POLARIS], and they have been warm by this amount for several years, whereas navigation system temperatures are usually too warm, but because of a dependence on altitude and ambient temperature they are simply too variable for assigning a single-number correction. After using the suggested corrections the experimenter can assume that MMS temperatures will exhibit an accuracy of less than 0.40 K, whereas navigation system temperatures will exhibit an accuracy of approximately 1.0 K.
The MTP temperatures in the archive files have been corrected so as to agree with ROABs.
SONEX, DC-8 (1997)
The following graph shows a comparison of MTP-derived tropopause altitudes and RAOB-based tropopause altitudes.
The altitude scales are "with respect to the DC-8 altitude." For these 15 comparisons there were no cases of a tropopause being below the DC-8. I am prepared to assume that the RAOB-based tropopause altitudes are "true" for this comparison. The large error bars reflect the fact that sometimes the T(z) profile has two tropopauses, and occasionally there is an altitude region having a lapse rate so close to the critical value of -2 [K/km] that the entire altitude region is close to being assigned as the tropopause altitude. There is a weak pattern of tropopause solutions being worse the farther they are from the DC-8's altitude, which is to be expected. For the "relative altitude" region 0 to 5 km, the RMS uncertainty of the MTP-based tropopause altitude is 0.5 km. (For the ER-2, which only has 2-channels, a similar analysis yields SE_trop_alt = 0.64 km for -2 km < dZp < +2 km.)
ACCENT I, ER-2 (1999)
SOLVE, ER-2 (1999,2000)
On the web page MTP/ER2 MP-File T(z) vs RAOB T(z) for SOLVE I conclude that MMS temperatures are acceptably close to RAOB temepratures, and that subtracting an additional 0.3 K (as earlier studies indicated was needed) is not necessary.
ACCENT II, ER2 (2000)
This site opened: July 20, 2001. Last Update: March 12, 2003