Challenges of Evaluating Atmospheric Moisture From Several TAMDAR Soundings
Prepared by Gene Brusky, SOO NWS Green Bay
A strong arctic front had swept through eastern Minnesota overnight with strong cold advection and rapidly falling surface temperatures occurring over the entire region at 13Z January 13, 2005. The 1345Z IR image below (Figure 1) indicated that the trailing edge of a post frontal clouds (yellow line on the image) extended from near DLH...to just east of MSP to near KFRM, with generally clear skies to the west. Figure 2 is a comparison of two TAMDAR descent soundings approximately 1 minute apart . The 1341Z sounding (red/blue) is from an aircraft descending from the northwest (green flight path in upper-left insert in Figure 2) and the 1342Z sounding (purple) is from an aircraft descending from the south (purple flight path in upper-left insert in Figure 2). Note that the temperature curves for these two soundings compare quite well below about 800 mb. What about the differences in the temperature profiles above 800 mb? Note that the two aircraft were approaching from opposite directions and at the 700 mb level, these two aircraft were probably still at least 100 nm apart. The general synoptic situation was characterized by a a north-south temperature gradient at mid-levels over the region thus these the differences in mid-level temperatures seem plausible. How what about the differences in the moisture? Note that the 1342Z sounding (aircraft descending from the south) appears to be considerably more moist than the 1341Z sounding (aircraft ascending from the northwest). At first glance, it again seems reasonable to assume, given the synoptic situation, that the two aircraft may have been sampling real variations in mid-level moisture as they were beginning their descent to MPX. However, how do we account for the moisture differences in the lowest portions of the soundings...particularly near the surface? Are we seeing "real" meteorological differences in low-level atmospheric moisture between two aircraft approaching from opposite directions and landing approximately one minute apart? Which sounding (if any) can we then believe? Figure 3 is a comparison of the 1341Z sounding with another aircraft descending from the northwest approximately 15 minutes later at 1357Z. These two aircraft were were both descending through mainly clear skies. Note how remarkably well these two soundings compare below 700 mb suggesting that the 1342Z sounding may be suspect. Figure 4 is a comparison of two ascent soundings at 1332Z (toward the southwest) and at 1346Z (toward the southeast). Although there is bad data in the 1346Z sounding around 850 mb, these soundings seem to compare "reasonably" well. The moist layer near 900 mb observed in both soundings is probably a result of ascent through the shallow stratus deck that persisted just to the south and east of the airport. Note that neither of these soundings indicate saturated conditions near the surface as in the 1342Z sounding in Figure 2, thus again suggesting that the 1342Z sounding may be suspect. Note that there are notable differences (2-3 C) in the Td below about 925 mb (Figure 4). Are these differences simply due to the fact that there was strong CAA/drying behind the front and the sensors were simply sampling changes in T/Td over an approximate 15 minute period? It turned out that aircraft #5601 (1342Z sounding) was experiencing problems with its humidity sensor and had a significant (~30%) moist bias.
This purpose of this case is to illustrates the challenge of evaluating high temporal frequency sounding information that is now available as part of the TAMDAR GLFE. The purpose of the GLFE is to evaluate the accuracy and utility of the new TAMDAR sensor that will be installed on 64 Mesaba aircraft flying to large and smaller regional airports. Forecasters within the experiment domain should expect to see some questionable data from time to time as the sensor continues to be tested during the course of the experiment. Bias information (daily, weekly and interactive statistics for aircraft data, validated against the RUC model) are available at http://acweb.fsl.noaa.gov/ruc_acars/ (This page is linked from the main FSL aircraft page at http://acweb.fsl.noaa.gov/.) In addition to being aware of occasional sensor errors and biases, forecasters within the experiment domain will also be exposed to high temporal and spatial sounding data...particularly at the large hub airports like Minneapolis, Memphis and Detroit. This will expose forecasters to the inherent heterogeneity in temperature and especially moisture, that has not been available in real time via sounding data. For example, in regions of strong temperature and moisture discontinuities, two aircraft taking off/landing in opposite directions at approximately the same time may very likely have markedly different sounding structure particularly as they descend/ascend through mid-levels of the atmosphere. Thus, in order for forecasters to obtain an accurate picture of the environment via aircraft soundings, it is critical that they are cognizant of the direction/flight path of the aircraft that generated that sounding. Flight path information is available on the FSL website at http://acweb.fsl.noaa.gov/ and in AWIPS D2D via the "sampling" function.
Thanks to Rich Naistat (SOO MPX) and Tom Hultquist (SOO MQT) for reviewing this document and for the helpful information provided by Bill Moninger (FSL) regarding aircraft data QC.
Figure 1 - 1345Z January 13, 2005 11-3.9 um image and 14Z surface plot
Figure 2 (above) - Comparison of 1341Z (red/blue) and 1342Z (Purple) TAMDAR descent soundings at MPX.
Figure 3 (above) - Comparison of 1357Z and 1341Z TAMDAR descent soundings near MPX
Figure 4 (above) - Comparison of 1332Z and 1346Z TAMDAR ascent soundings near MPX.
