Title: Water Vapor Sensing System (Winds, Temperatures, and Water Vapor Fields from Commercial Aircraft) Author: Dr. Rex J. Fleming UCAR Visiting Scientist 3300 Mitchell Lane Boulder, CO 80301 Telephone: (303) 497-8165 Fax: (303) 497-8158 Email: fleming@joss.ucar.edu 1.0 Data Set Archives The use of commercial aircraft for obtaining weather observations is not new. We are, however, in the very beginning of a revolution in the use of this convenient platform for a variety of environmental parameters. This data set represents one aspect of this evolution-the use of a Water Vapor Sensing System (WVSS) for commercial aircraft. The Lockheed Martin Corporation (LMC) served as the prime contractor for the first generation WVSS (hereafter referred to as WVSS-I). They had several subcontractors that had important roles: BF Goodrich Aerospace (formerly Rosemont Aerospace), who supplies 99% of the world's total air temperature (TAT) probes for commercial aircraft, actually manufactured the WVSS-I. United Parcel Service (UPS), a commercial cargo air carrier who assisted LMC in obtaining the FAA certification of the WVSS-I on all B-757 cargo aircraft. AlliedSignal (now part of Honeywell, Inc.), who installed the ascent, descent, and enroute software for the WVSS-I on their avionics boxes on the aircraft. These avionics boxes continue to evolve toward more powerful, user-friendly systems and are usually referred to as digital flight data acquisition units (DFDAUs). There have been five UPS aircraft flying the WVSS-I units since April/May of 1999, and one unit since September 1999. Over 200,000 quality-controlled reports of winds, temperature, and water vapor information (mixing ratios, relative humidity, and dewpoints) have been received from these six aircraft. The measurement technology of the WVSS-I is described below. The fundamental measurement is the Vaisala thin-film capacitor for relative humidity (RH); however, BF Goodrich has improved this technology on aircraft over that which is possible on radiosondes. The WVSS-I data is available over the continental United States and the Gulf of Mexico. Winds, temperatures, and water vapor fields of information are available on ascent, descent, and enroute portions of flights over these regions. Unfortunately, for these first six prototype aircraft, the enroute data was missing for some months. However, the primary interest of the FAA and UPS was on the high resolution ascent/descent data. All meteorological data (winds, temperatures, and water vapor information) from commercial aircraft are available in real time via the Aircraft Communication Addressing and Reporting System (ACARS). Details of this system can be found in Appendix B and C of the WVSS home page. This information is available now on the Internet and soon will be available at the National Center for Environmental Prediction (NCEP). Dewpoint information on the Internet has been corrected (see Appendix A), however, mixing ratio information on the Internet provided by NOAA's Forecast System Laboratory (FLS) has not been corrected and is, on average, about 5% too wet. This WVSS-I data is also distributed on the UNIDATA NETCDF local data manager (LDM). However, we caution the user about the quality of this data. Not only is mixing ratio information inconsistent with the dewpoint information (see above), there appears to be occurrences of missing data, data with no dewpoints, and inappropriate Mach number fields. Researchers are encouraged to use the quality-controlled WVSS-I data through UCAR's Joint Office for Science Support (JOSS). This data is available through CODIAC at http://www.joss.ucar.edu/codiac/. This data has consistent water vapor information (dewpoint, RH, and mixing ratio are consistent), and has the benefit of time continuity to check for changes in WVSS-I calibration. 2.0 IMPORTANT NOTE ON WVSS-I DATA The sensor technology of the first generation Water Vapor Sensing System (WVSS-I) is Vaisala's thin-film capacitor. Though a better sensor than found on radiosondes, it still loses calibration over time. The data from WVSS-I gathered from July 1, 1999 to December 31, 1999 has been evaluated and a scientific article is being prepared. This data is useful with data from four of the six aircraft considered quite good. One aircraft (#376) appears approximately 5% too wet and one aircraft (#441) appears approximately 5% too dry (very dry in November and December 1999). These sensors were on the aircraft for longer periods than we wanted (we originally thought that the sensors would be recalibrated after six months of operation). The length of service up until December 31, 1999 was: 10 months for #376, #378, and #441 8 months for #97 and #714 3 months for #375 The above six aircraft were the original aircraft with AlliedSignal avionics that downlinked mixing ratio information. During 2000, there was a switch of avionics equipment from AlliedSignal to Teledyne. This was a lengthy process and the WVSS-I units were left on the aircraft without recalibration. Thus, AlliedSignal-equipped aircraft in 2000 have data that is suspect! This did allow us to learn the failure mode (data becoming progressively drier) and to obtain an indication of the sensor lifetime. A summary of each aircraft follows. #378: 12 months of good data but was removed in January 2000 for Teledyne equipment. #376: 21 months of good data (5% bias wet), became progressively drier in 2000, very dry in October 2000, and was removed in November 2000. #441: 9 months of good data (5% bias dry), became very dry in November 1999, was removed in June 2000. #714: 11 months of good data, became very dry in March 2000, will be changed in early 2001. #97: 15 months of good data, became very dry in July 2000, will be replaced in early 2001. #375: 12 months of good data, started getting drier in September 2000. New Teledyne-equipped aircraft will be coming online in 2001. These will downlink relative humidity. We will announce when they are available in the archive. The second generation Water Vapor Sensing System (WVSS-II) has a diode laser as the sensing element. This data looks really good and will be certified on Delta Airlines in November 2001. Plans are progressing toward a national implementation of this second generation system. 3.0 Instrument Description The water vapor sensor on the WVSS-I is a modified Vaisala system with a thin-film capacitor as the fundamental sensor for relative humidity. A good description of accuracy, frequency, etc., can be found in the technical references (Appendix A and B listed on the WVSS homepage). The first six prototype aircraft have mixing ratio downlinked (even though relative humidity is measured). As long as an accurate temperature is available, the three water vapor measurement fields of dewpoint, relative humidity, and mixing ratio can be made consistent. All three fields are supplied with the CODIAC data set. Note that as the "quasi-operational" phase of the WVSS program begins (54 aircraft will be flying), the avionics vendor will be switched from Allied Signal to Teledyne. The instrumentation for sensing water vapor will be the same, but relative humidity will be the variable that is downlinked. This will be transparent to the user as all three fields will still be supplied in the CODIAC data set. 4.0 Data Collection and Processing All of the WVSS data are first organized by aircraft tail number within a given day, then chronologically with time over that day. There then follows the latitude, longitude, height (feet), pressure (hPa), temperature (Ts in K), dewpoint (Td in K), relative humidity (%), mixing ratio ( r in g/kg), and a quality control flag (Q). 5.0 Quality Control Flags The single digit quality control flag has evolved in meaning in a few cases. The values available in this data set are summarized as follows. Q-value Data Characteristic Comments 0 Data good 1 Data good 2 Data good 3 Data "saturated", probably RH=100% wet or other condition. Do Td = Ts not use mixing ratio (r). r = 1.0 4 Data "saturated" for greater RH = 100% than 120 seconds. Do not Td = 999 use data. r = 1.0 5 Data bad due to probe heater RH = 0% Do not use data. Td = 999 r = 1.0 Flags of 6 through 9 have not been seen in the data-except Flag =7, which implies that the WVSS unit was not installed on the aircraft, or a key parameter was missing. Flag 7 data are excluded from this data set. In summary, data with a Flag of 4 or greater should not be used and data with a Flag of 3 is suspect and the mixing ratio should not be used.