ACTIVE VH-ARA Data Set Final Report Peter Isaac and Jörg Hacker 1 August 2007

Transcription

1 ACTIVE VH-ARA Data Set Final Report Peter Isaac and Jörg Hacker 1 August 2007 Airborne Research Australia / Finders University P.O. Box 335 Salisbury South, 5106 Australia ARA Technical Report No.??-2007

2 Table of Contents 1 INTRODUCTION OVERVIEW OF FLIGHTS GENERAL FLIGHTS STANDARD AIRCRAFT CONFIGURATION FERRY/LIDAR AIRCRAFT CONFIGURATION FLIGHT DETAILS /11/2005 : TE /11/2005 : TE /11/2005 : AE /11/2005 : AE /11/2005 : TE /11/2005 : AE /12/2005 : AE /12/2005 : AE /12/2005 : AE /12/2005 : AE /12/2005 : AE /12/2005 : AE /12/2005 : AE /12/2005 : SE /12/2005 : FE /01/2006 : FE /01/2006 : AE /01/2006 : AE /01/2006 : AE /01/2006 : AE /01/2006 : AE /01/2006 : LE /02/2006 : LE /02/2006 : SE /02/2006 : AE /02/2006 : AE /02/2006 : AE /02/2006 : AE /02/2006 : AE /02/2006 : AE /02/2006 : TE /02/2006 : FE DESCRIPTION OF DATA SET FILE FORMAT DATA QUALITY FLAG KNOWN PROBLEMS SYNTHESIS OF DYNAMIC PRESSURE USING ANGLE OF ATTACK SYNTHESIS OF DYNAMIC PRESSURE USING AIRCRAFT GROUND SPEED PROCEDURE FOR FILLING GAPS IN DATA NOISE IN STATIC PRESSURE AND TEMPERATURE ROSEMOUNT TEMPERATURE SENSOR DE-ICE HEATERS COMPARISON WITH OTHER DATA DARWIN GROUND-BASED DATA DARWIN RADIOSONDE DATA INTER-COMPARISON FLIGHTS WITH DCALM ACKNOWLEDGMENTS

3 1 Introduction The Aerosol and Chemical Transport in Tropical Convection (ACTIVE) experiment held field campaigns based in Darwin in November/December 2005 and January/February The Airborne Research Australia (ARA) high altitude research aircraft, the Grob G520T Egret (call sign VH-ARA), was used in both of these campaigns. This report and the companion Appendix have been written to act as a guide to the data set provided by ARA to the University of Manchester. The data set comprises all of the data recorded on the aircraft data loggers during the ACTIVE field campaigns. The Egret was configured to carry a wide range of aerosol, cloud particle and air chemistry instrumentation along with a suite of air state instruments for measuring pressure, air temperature and the three components of the ambient wind field. This report describes the instrumentation carried by the Egret, gives a brief description of each flight and describes the data set obtained from the instruments mounted on the Egret whose output was recorded by the aircraft data loggers. Section 2 presents an overview of the flights performed by VH-ARA and provides a general description of the instrumentation and aircraft data logger configuration for the flights. Section 3 gives a more detailed description of each flight. The basic flight details and any special circumstances are given, followed by a description of the instrumentation and data logger set up for each flight. This is followed by a detailed description of the flight pattern for each flight and a plot of the flight track overlaid on an outline map of the Darwin area. Each flight is dealt with in a separate subsection. Section 4 describes the data set provided to the University of Manchester. The names of the files, their contents and formats are given in separate subsections. To aid interpretation of the data, each data field has been assigned a data quality flag during processing and these are defined in the last subsection of Section 4. A number of problems with the data were identified during the field campaigns and during the subsequent processing of the final data set. Section 5 describes these problems and provides details of the methods used to identify and remove affected data. The techniques used to replace missing dynamic pressure data, caused by ice forming in the pitot tube or by leaks in the associated tubing, are given. The method used to fill gaps in the data due to logger failures or due to the reject of contaminated data is also described in detail. Section 6 presents the results from comparisons of the data from VH-ARA with data from the Darwin automatic weather station, the Darwin radiosondes and from the National Environmental Research Council (NERC) Dornier 228 (call sign D-CALM). The inter-comparison data is used to validate the air temperature and wind data from the VH-ARA data set. 3

4 2 Overview of Flights 2.1 General The data set from VH-ARA for the ACTIVE campaign consists of 32 flights, 15 in 2005 and 17 in The flights have been numbered consecutively from 01 to 32 and given a two letter prefix that designates the type of flight (A for science flight, F for ferry flight, S for survey flight and T for test flight) and the aircraft (E for Egret). Of the 32 flights, 4 were dedicated test flights. The first 2, TE01 and TE02 in 2005, were to test the general set up of the aircraft and the NOX and SP2 instruments. The third test flight, TE05, was flown following the propeller repair. Oil streaks were found on the aircraft canopy after flight AE04 on 16/11/2005 and these were traced to a failed start lock on the propeller. A replacement start lock was sourced from the aircraft manufacturer (Grob Aerospace AG) and installed on 26/11/2005. The fourth, TE31, was a calibration flight at the end of the 2006 campaign. During this flight, the aircraft performed a series of specific manoeuvres designed to yield correction factors and calibrations for the temperature (T a ), static pressure (p s ), dynamic pressure (q c ) and the angle of sideslip (β) and attack (α) measurements. The manoeuvres were performed at nominal altitudes of 10,000' and 30,000'. Three of the flights were ferry flights from Darwin to Adelaide (FE15 and FE32) or from Adelaide to Darwin (FE16). A gas chromatograph was installed on the aircraft for these ferry flights along with instruments for measuring pressure and temperature. The remaining 25 flights were science flights in the vicinity of Darwin. Twenty-three of these flights (AE03, AE04, AE06 - SE14, AE17 - AE21, SE24 - AE30) carried either an instrument for measuring NOX concentration (NOX u-bay) or black carbon aerosol (SP2 u-bay). Two of these 23 flights were survey flights (SE14 and SE24) which were performed to characterise background conditions in relatively clear air. For the remaining 2 of the 25 flights science flights (LE22 and LE23), the aircraft carried an ozone-detecting LIDAR in place of the NOX or SP2 u-bays. The aircraft configurations, in terms of instrumentation and data logging, used for the ACTIVE flights fall into two broad categories: standard and ferry/lidar. The standard configuration was used for all test and science flights excluding the LIDAR flights. The ferry/lidar configuration was used, naturally, for the ferry and LIDAR flights. The configurations are described in Sections? and?. 2.2 Flights Table 2.1 and Table 2.2 summarise the VH-ARA flights for the 2005 and 2006 campaigns respectively. 4

5 Date Flight Take Land Type FR RR p s T a q c UV TDL1 TDL2 TSI1 TSI2 Comments 11/11 TE01 06:44 09:45 Test P P P P N N N N NOX u-bay, loggers stopped repeatedly 13/11 TE02 05:51 07:18 Test P P P P N N N N SP2 u-bay, loggers stopped repeatedly 15/11 AE03 06:16 07:59 HA P P P P N N N N loggers stopped repeatedly 16/11 AE04 06:28 09:47 HC P P P P N N N N loggers stopped repeatedly, q c bad after 07:58 27/11 TE05 06:42 08:10 Test Y Y Y Y Y Y N N SP2 u-bay 30/11 AE06 04:14 08:11 HC P P P P P P N P loggers stopped repeatedly 01/12 AE07 05:01 08:34 HA P P P P P P N P loggers stopped repeatedly, q c bad after 06:17 03/12 AE08 06:17 09:45 HC1 1 1 Y Y P P Y Y Y Y q c bad after 07:13 04/12 AE09 06:51 09:42 HA1 1 1 Y Y Y Y Y Y Y Y 05/12 AE10 06:04 08:27 HC1 1 1 Y Y Y Y Y Y Y Y inter-comparison with DCALM 06/12 AE11 04:39 09:07 HA1 1 1 Y Y P P N Y Y Y q c bad after 05:40, no TANS until 06:56 08/12 AE12 03:51 08:03 HC2 1 1 Y Y Y Y Y Y Y Y 09/12 AE13 04:25 08:07 HA1 1 1 Y Y N Y Y Y Y Y q c bad for entire flight 10/12 SE14 01:53 04:47 HA2 1 1 Y Y Y Y Y Y Y Y 11/12 FE15 23:55 06:05 Ferry 1 1 Y Y N N N Y N N Darwin to Parafield Table 2.1: Summary of flights for the 2005 ACTIVE campaign. 1) HA flights had the SP2 u-bay fitted; HC flights had the NOX u-bay fitted. 2) HA1 and HC1 flights were through cirrus; HA2 and HC2 flights were survey flights in clear air. 3) FR is the Fuselage REMlet (CR-2, TE-49C O 3, TSIs); RR is the Right-wing REMlet (pressures, temperatures, winds), 20 means data is at 20 Hz, 1 means that data is at 1 Hz. 4) "Y" means data is available for entire flight; "N" means data is not available. 5) "P" indicates partial data is available, missing p s, T a data have been interpolated using the GPS altitude, missing q c data has been synthesised from the pressure difference across the angle of attack ports when available or from ground speed. 5

6 Date Flight Take Land Type FR RR p s T a q c UV TDL1 TDL2 TSI1 TSI2 Comments 17/01 FE16 23:13 06:47 Ferry 20 N Y Y N N N Y N N Parafield to Darwin 20/01 AE17 06:47 09:05 MA Y Y P Y Y Y A A q c bad after 07:21 22/01 AE18 07:05 10:05 MC Y Y P Y Y Y Y Y q c bad after 08:01 23/01 AE19 02:55 06:35 MC Y Y Y Y Y Y Y Y 25/01 AE20 05:20 09:35 MA Y Y Y Y Y Y Y Y 27/01 AE21 05:40 09:35 MA Y Y Y Y Y Y Y Y 31/01 LE22 10:51 13:23 LIDAR 20 N Y Y N N Y Y N N 01/02 LE23 10:55 14:12 LIDAR 20 N Y Y N N Y Y N N 03/02 SE24 05:12 09:10 MA Y Y Y Y Y Y Y Y 06/02 AE25 06:17 09:10 MA Y Y N Y Y Y Y Y q c data bad for whole flight 08/02 AE26 08:08 10:50 MC Y Y N P Y Y N N q c data bad for whole flight, gaps in TANS data, TSIs not plugged in 10/02 AE27 07:12 11:00 MC Y Y N Y Y Y Y Y q c data bad or whole flight 12/02 AE28 07:48 09:52 MA Y Y Y Y Y Y Y Y 13/02 AE29 05:55 10:12 MA Y Y N P Y Y Y Y q c data bad for whole flight, gaps in TANS data 14/02 AE30 02:44 05:22 MI Y Y Y Y Y Y Y Y inter-comparison with DCALM 15/02 TE31 02:48 06:32 Test Y Y Y Y Y Y Y Y calibration flight 18/02 FE32 23:19 06:17 Ferry 20 N Y Y N N Y Y N N Darwin to Parafield Table 2.2: Summary of flights for the 2006 ACTIVE campaign. 1) MA flights had the SP2 u-bay fitted; MC flights had the NOX u-bay fitted. 2) MA1 and MC1 flights were through cirrus; MA2 and MC2 flights were survey flights in clear air. 3) FR is the Fuselage REMlet, RR is the Right-wing REMlet, 20 means data is at 20 Hz and 1 means that data is at 1 Hz. 4) "P" indicates partial data is available, missing p s, T a data have been interpolated using the GPS altitude, missing q c data has been synthesised from the pressure difference across the angle of attack ports when available or from ground speed. 6

7 2.3 Standard Aircraft Configuration Instrumentation Table 2.3 lists the quantities recorded by the data logging system on VH-ARA, the sensors used to measure these and their location on the aircraft for the standard configuration flights. "Standard" configuration flights were TE01 to AE14 inclusive (2005), AE17 to AE21 and SE24 to TE31 inclusive (2006). Quantity Sensor Location Pressure ports Rosemount hole probe at tip of 2.5m mast mounted to right pylon Static pressure Rosemount1201 right pylon Dynamic pressure Rosemount 1221 right pylon Angle of attack Rosemount 1221 right pylon Angle of side slip Rosemount 1221 right pylon Air temperature Rosemount 102 de-iced right pylon Aircraft position Novatel GPS fuselage behind rear cabin Aircraft velocity Novatel GPS fuselage behind rear cabin Aircraft attitude Trimble TANS fuselage behind rear cabin Water vapour Closed path TDL hygrometer right gear pod Open path TDL hygrometer right gear pod Buck Research CR-2 rear fuselage compartment Ozone Thermo Electron TE-49C rear fuselage compartment Condensation particles 2 x TSI-3010 u-bay Table 2.3: Summary of instrumentation recorded by the VH-ARA data logging system during the standard configuration flights (see text for details). The open path TDL (TE05 onwards) required information on the pressure and temperature in real-time during the flight in order to calculate its internal calibration coefficients. The original method of supplying this information from the aircraft logging system was via a digital-to-analogue converter that accepted a serial data stream from the aircraft display computer and fed an analogue output to the TDL. This method was abandoned when it proved to be inoperable during testing of the system between AE04 and TE05. In its place, the analogue signal from the Rosemount pressure and temperature sensors was fed direct to the TDL for TE05. Examination of the data after the flight showed that the pressure recorded by the TDL was approximately 20 hpa lower than that recorded by the aircraft data logger. This was due to a combination of low input impedance for the TDL analogue input and a slow rise time for the Rosemount pressure sensor after the TDL input impedance became high when the TDL analogue inputs were sampled. This meant that the pressure signal was still recovering for being loaded down when the TDL sampled the pressure input. The direct analogue connection between the TDL and the Rosemount pressure and temperature sensors was removed for AE06 pending the construction of an electronic 7

8 buffer to prevent loading of the pressure sensor output. For AE07, the pressure and temperature for the open path TDL were set internally at constant values of 190 hpa and -53 C respectively. An instrumentation amplifier was placed between the Rosemount pressure and temperature sensor and the TDL analogue inputs before flight AE08. As a result, the aircraft and TDL pressures agree to within approximately 1 hpa when the aircraft is on the ground before take-off and after landing. However, at the maximum altitude of the TDL pressure reads approximately 20 hpa higher than the pressure recorded by the aircraft. This was due to the use of a voltage divider on the output of the instrumentation amplifier that was also being loaded down by the low input impedance of the TDL analogue input. The voltage divider was removed before flight AE09. Note that the pressures recorded by the TDL differ from those recorded by the aircraft data logger and given in the final data set. The values from the aircraft data logger have been corrected for an observed offset of 4.6 hpa (comparison with Darwin AWS data) and low-pass filtered with a cut-off frequency of 0.2 Hz Data Logging The data logging system employed for the standard configuration flights consisted of two remote analogue-to-digital converters (REMlets), two data loggers and a display computer. One REMlet was installed in the pylon mounted beneath the right wing and the second was installed in the rear fuselage compartment. The right pylon REMlet sampled the Rosemount pressure and temperature sensors. The fuselage REMlet sampled the CR-2 frost point hygrometer, the TE-49C O 3 instrument, the TSI analogue outputs and the aircraft radio transmitter status (2 VHF channels and 1 UHF channel). The two data loggers were located in the rear fuselage compartment, one on the lefthand side of the aircraft and the other on the right. Data logger serial number (S/N) 03 was installed in the right hand location and S/N 04 was installed in the left-hand location for flights TE01 to AE04 inclusive. The location of the loggers was swapped between flights AE04 and TE05 in an unsuccessful effort to prevent the repeated loss of the analogue data stream from the REMlets that occurred on flights TE01 to AE04 (see below). From TE05 onwards, S/N 03 was in the left-hand location and S/N 04 was in the right hand location. The display computer was mounted in the rear cabin and drove an LCD display positioned over the aircraft instrument panel in front of the mission scientist. There were a number of changes to the data logging for the standard configuration flights during the ACTIVE campaign. The changes were made in an effort to isolate and remove the cause for the repeated loss of analogue data from the REMlets that occurred for flights TE01 to AE07 inclusive. The chronological order of the changes is given in the following paragraphs The original configuration of the data logging system, used for flights TE01 to AE04, had the REMlet, Novatel GPS, Novatel 1 pulse-per-second (PPS) and TANS GPS data recorded on the right fuselage logger. The left fuselage logger was dedicated to the TSI and TDL serial data streams. The left and right data loggers both recorded the 1 PPS signal from their internal GPS cards to provide a means for synchronising the time stamp given to the incoming data by each logger. A 20 Hz pulse generated by 8

9 the right fuselage logger from the internal 1 PPS timing signal was used to trigger the REMlets in this configuration. The analogue data streams from the REMlets repeatedly halted during flights TE01 to AE04 with this configuration. The only recourse available to the mission scientist during these flights was to cycle power to the loggers and REMlets in an attempt to restart the logging system. The system halted after each restart with no apparent pattern in the time before failure. The data logging system was tested on the ground between flights AE04 and TE05. This showed that both data loggers were halting after processing a random number of 1 PPS timing signals from their internal GPS units. The logging programme was then modified to disable the reading of the internal GPS time and the Novatel GPS data stream was routed to both right and left loggers. The recording of the Novatel GPS data by both loggers provided the means to synchronise the time stamp given to incoming data by each logger. The REMlets continued to be triggered by a 20 Hz pulse sent by the right hand data logger. This configuration (internal GPS disabled, Novatel GPS data routed to both loggers, REMlets triggered at 20 Hz) was used for flights TE05 to AE07 inclusive. Both loggers ran without interruption for the relatively short test flight (TE05) but the loss of the analogue data stream from the REMlets occurred again on flights AE06 and AE07. The left-hand data logger appears to have run continuously for these flights. A further attempt to fix the problem with the right hand data logger and REMlets was made on 2/12/2005 between flights AE07 and AE08. The data logger and REMlets were linked via the 20 Hz trigger pulse in the configurations described above. This meant that it was not possible to decide if the loss of analogue data was caused by a problem with the logger (eg loss of trigger pulse) or with the REMlets (eg loss of response to a trigger pulse). In an attempt to resolve this ambiguity, the 20 Hz trigger pulse for the REMlets was disabled and replaced by the Novatel GPS 1 PPS. This modification also required a change to the data logger configuration so that incoming data from the REMlets were read as generic serial data streams. The size of the data packet expected from the REMlets, required when the data is read as a generic serial stream, was initially set to 64 bytes. This configuration (internal GPS disabled, Novatel GPS data to both loggers, REMlets triggered by Novatel 1 PPS) was used for flight AE08. The modifications to the logger configuration meant that the logger operation could not be checked during this flight. However, both loggers were found to be running when checked after parking in the hangar at the conclusion of flight AE08 and the REMlet data was still being received by the right logger. Examination of the REMlet data recorded by the right hand logger showed that the packet size of 64 bytes was 1 byte too small, causing a byte shift in the raw data. This was corrected in subsequent processing and all data from AE08 were recovered. The right hand data logger configuration was changed to specify the correct REMlet data packet size of 65 bytes before flight AE09 and this data logger configuration was used up to and including flight FE15. The reason for the loss of analogue data was traced to the use of a component in the data loggers that was different from the one specified by Airborne Research Australia. The problem was identified and fixed while the aircraft was in Adelaide between the 2005 and 2006 field campaigns. 9

10 On return to Darwin, the fuselage REMlet (CR-2, TE-49C O 3, analogue TSI, radio markers) was triggered by the 20 Hz pulse sent from the right hand data logger but the right wing REMlet was still triggered by the Novatel GPS 1 PPS. This was discovered after flight AE20 and the correct triggering of the right wing REMlet by the 20 Hz pulse from the logger was reintroduced from AE21 onwards. As a result, there are 20 Hz data for the analogue channels sampled by the fuselage REMlet (CR- 2, TE-49C O 3, analogue TSIs and radio markers) for all flights in However, the analogue data from the right wing REMlet (pressures and temperature) is only available at 1 Hz for flights AE17 to AE20. The pressure and temperature data is available at 20 Hz for flights AE21 and SE24 to FE32 inclusive. 2.4 Ferry/LIDAR Aircraft Configuration Instrumentation Table 2.4 lists the quantities recorded by the data logging system on VH-ARA, the sensors used to measure these and their location on the aircraft for the ferry and LIDAR configuration flights. Ferry and LIDAR configuration flights were FE15 (2005), FE16, LE22, LE23 and FE32 (2006). FE15 differed from other ferry/lidar configuration flights in that the right pylon was mounted on the aircraft for this flight. This means that FE15 used the same static pressure sensor as used for the standard configuration flights and that the static pressure and air temperature data were sampled by the right wing REMlet. The right pylon and the associated pressure sensors were not mounted on the aircraft for flights FE16, LE22, LE23 and FE32. An alternate pressure sensor, mounted inside the rear fuselage compartment, was used to measure static pressure for these flights. The manufacturer's calibration was not available for this pressure sensor. A calibration was inferred from the "standard" configuration sensor via the chamber pressure measurement of the CR-2 frost point hygrometer as follows. First, a calibration for the CR-2 chamber pressure was obtained by least-squares fit of the CR- 2 pressure output voltage against the static pressure for all flights with the right pylon fitted to the aircraft. This calibration of the CR-2 chamber pressure was then transferred to the static pressure sensor used on the ferry and LIDAR flights. Temperature was measured using the same Rosemount 102 de-iced sensor as for the "standard" configuration flights but the sensor was mounted on the left-hand side of the u-bay with the associated electronics in the rear fuselage compartment. 10

11 Quantity Sensor Location Static pressure Rosemount1201 rear fuselage compartment (FE15 used right pylon) Air temperature Rosemount 102 de-iced rear fuselage compartment (FE15 used right pylon) Aircraft position Novatel GPS fuselage behind rear cabin Aircraft velocity Novatel GPS fuselage behind rear cabin Aircraft attitude Trimble TANS fuselage behind rear cabin Water vapour Closed path TDL hygrometer right gear pod (not used for FE15 or FE16) Open path TDL hygrometer Buck Research CR-2 right gear pod rear fuselage compartment (LIDAR flights only) Ozone Thermo Electron TE-49C rear fuselage compartment (LIDAR flights only) Table 2.4: Summary of instrumentation recorded by the VH-ARA data logging system during the ferry and LIDAR configuration flights (see text for details). The open path TDL was supplied with analogue signals for static pressure and temperature as described in Section Data Logging The data logging system employed for the ferry/lidar configuration flights consisted of one (FE16, LE22, LE23 and FE32) or two (FE15) remote analogue-todigital converters (REMlets), two data loggers and a display computer. For FE15, one REMlet was installed in the pylon mounted beneath the right wing and the second was installed in the rear fuselage compartment. The right pylon REMlet sampled the Rosemount pressure and temperature sensors. The fuselage REMlet sampled the aircraft radio transmitter status (2 VHF channels and 1 UHF channel). Note that the CR-2 frost point hygrometer, the TE-49C O 3 instrument and the TSI analogue outputs were sampled as well but the instruments were not fitted to the aircraft. For FE16, LE22, LE23 and FE32, the REMlet was installed in the rear fuselage compartment. This REMlet sampled the Rosemount static pressure and temperature sensors and the radio markers on flights FE16, LE22, LE23 and FE32. The CR-2 and TE-49C O 3 instruments were also sampled on flights LE22 and LE23. The two data loggers were located in the rear fuselage compartment, S/N 03 on the left-hand side of the aircraft and S/N 04 on the right. The display computer was mounted in the rear cabin and drove an LCD display positioned over the aircraft instrument panel in front of the mission scientist. This data logger configuration used for flights AE09 to SE14 (internal GPS disabled, Novatel GPS data to both loggers, REMlets triggered by Novatel 1 PPS) was used for flight FE15. The use of the same configuration as the standard flights means that the data from flight FE15 can be processed using the same techniques as used for flights AE08 to SE14. 11

12 For flights FE16, LE22, LE23 and FE32, the data loggers were configured to disable the internal GPS timing, the Novatel GPS data stream was fed to both loggers and the fuselage REMlet was triggered by a 20 Hz pulse derived from the right hand logger. 12

13 3 Flight Details /11/2005 : TE General This was the first test flight of VH-ARA in Darwin. The NOX u-bay was fitted to the aircraft. Logger S/N 03 was in the right fuselage position and S/N 04 was in the left fuselage position. Take-off was approximately 06:44 UTC, landing at approximately 09:45 UTC. The mission scientist was Jörg Hacker Instrumentation and Data Logging The NOX u-bay was flown with both TSI instruments in place. The standard aircraft configuration consisting of the right pylon, Rosemount boom, Novatel GPS and TANS GPS was used. The right wing and fuselage REMlets (remote analogue to digital converters) were triggered by a 20 Hz sample pulse from the right fuselage data logger. The REMlet, Novatel GPS and TANS GPS data were recorded on the right fuselage logger. The digital data from TDL1, TSI1 and TS2 were recorded on the left fuselage logger. The right and left fuselage data loggers both recorded the 1 second timing pulse from their internal GPS systems to provide a means for synchronising the time stamp assigned to the various data streams by the loggers. Both loggers and REMlets stopped repeatedly during the flight. For example, the right fuselage data logger was restarted 6 times during the flight but there are 17 gaps longer than 5 secs in the analogue data from the right wing (temperatures and pressures). The left fuselage data logger stopped completely after 4,250 seconds, about one third of the flight, and there is no data (digital TDL and TSI) from this logger for the remainder of the flight. The gaps in the data caused by the failure of the right fuselage data logger have been filled using different methods depending on the data type. No attempt has been made to fill gaps in the data from the left fuselage data logger. Gaps in the Novatel GPS position data were filled by extracting the required values from the aircraft position recorded in the OziExplorer track file. Gaps in the Novatel GPS velocity data were filled by differentiating the position data obtained from the OziExplorer track file. Gaps in the temperature and pressure data were filled as follows. First, the decreases in temperature and pressure with height were calculated using all available data. The change in temperature or pressure from the last good measurement was then calculated using these lapse rates and the change in aircraft altitude from the Novatel GPS data. The values used to fill the gaps in T a and p s were then calculated by adding the running integral of the temperature and pressure changes to the last good measurement. The difference between the last value calculated for the gap and the first good measurement of the next data block is recorded and used to characterise the uncertainty of the method. A linear correction is then applied to the values used to fill the gaps to remove any discontinuity between 13

14 the end of the gap and the start of the next data block. Details of the method and the uncertainty in the values used to fill the gaps are discussed in Section 5. The pressure altitude begins to deviate from the GPS altitude at 09:42:10 UTC. The aircraft was on descent into Darwin at an altitude of 400 m at this time. The problem is due to erroneous static pressure data, probably caused by water in the static port or pressure line. There are no other known problems with the meteorological data at this time other than those associated with the gaps caused by the logger failures. There is digital data for TDL1, TSI1 and TSI2 for the first third of the flight. There is no data for TDL2. The digital data for TSI1 and TSI2, which appear to be diagnostic messages rather than measurements, are scrambled because the wrong serial port configuration was specified in the data logger initialisation file Flight Track After taking off from Darwin airport to the north, the aircraft tracked northeast before turning northwest and continuing to climb, crossing the coast over Lee Point at 2,000 m. The aircraft then tracked west to a point 8 km northwest of Charles Point before tracking northwest and climbing to 11,900 m at a point 156 km west-northwest of Darwin. From here the aircraft turned southeast until intersecting latitude 12 20'. The aircraft then completed three east/west oriented transects along this latitude at 12,600, 13,500 and 14,400 m. At the end of the last transect from west to east the aircraft began a spiral descent 25 km northwest of Darwin before tracking west of Darwin and landing from the southeast. A plot of the flight track is shown in Figure

15 + CGMP ao paoin e rut dn et DS n ot nbu upa onr indt ty ACTIVE ARA TE :31-09:45 UTC Latitude Darwin Longitude Figure 3.1: Flight track for TE01 on 11/11/2005. Arrows indicate the direction of the aircraft track. Numbers to the left of the aircraft track are the seconds since 0000 UTC in increments of 300 secs. Numbers to the right of the aircraft track are the aircraft altitude in km. 15

16 3.2 13/11/2005 : TE General This was the second test flight of VH-ARA in Darwin. The SP2 u-bay was fitted to the aircraft. Logger S/N 03 was in the right fuselage position and S/N 04 was in the left fuselage position. Take-off was approximately 05:51 UTC, landing at approximately 07:18 UTC. The mission scientist was Jörg Hacker Instrumentation and Data Logging The NOX u-bay was flown and the standard aircraft configuration consisting of the right pylon, Rosemount boom, Novatel GPS and TANS GPS was used. The right wing and fuselage REMlets (remote analogue to digital converters) were triggered by a 20 Hz sample pulse from the right fuselage data logger. The REMlet, Novatel GPS and TANS GPS data were recorded on the right fuselage logger. The digital data from TDL1, TSI1 and TS2 were recorded on the left fuselage logger. The right and left fuselage data loggers both recorded the 1 second timing pulse from their internal GPS systems to provide a means for synchronising the time stamp assigned to the various data streams by the loggers. Both loggers and REMlets stopped repeatedly during the flight. The right fuselage data logger was restarted 5 times during the flight. The left fuselage data logger stopped completely after 103 seconds, before the aircraft had taken off. The gaps in the data caused by the failure of the right fuselage data logger have been filled using different methods depending on the data type. No attempt has been made to fill gaps in the data from the left fuselage data logger. Gaps in the Novatel GPS position data were filled by extracting the required values from the aircraft position recorded in the OziExplorer track file. Gaps in the Novatel GPS velocity data were filled by differentiating the position data obtained from the OziExplorer track file. Gaps in the temperature and pressure data were filled as follows. First, the decreases in temperature and pressure with height were calculated using all available data. The change in temperature or pressure from the last good measurement was then calculated using these lapse rates and the change in aircraft altitude from the Novatel GPS data. The values used to fill the gaps in T a and p s were then calculated by adding the running integral of the temperature and pressure changes to the last good measurement. The difference between the last value calculated for the gap and the first good measurement of the next data block is recorded and used to characterise the uncertainty of the method. A linear correction is then applied to the values used to fill the gaps to remove any discontinuity between the end of the gap and the start of the next data block. Details of the method and the uncertainty in the values used to fill the gaps are discussed in Section 5. There are no known problems with the meteorological data at this time other than those associated with the gaps caused by the logger failures. 16

17 There is digital data for TDL1, TSI1 and TSI2 for the first 100 seconds of the flight (aircraft still in the hangar). There is no data for TDL2. The digital data for TSI1 and TSI2, which appear to be diagnostic messages rather than measurements, are scrambled because the wrong serial port configuration was specified in the data logger initialisation file Flight Track After take-off, the aircraft tracked west-northwest and climbed to an altitude of 6,500 m at 87 km from Darwin. The aircraft then tracked east-northeast climbing to an altitude of 12,100 m at a point 17.5 km east of Cape Gambier before turning northeast and continuing to climb slowly, reaching 12,600 m over Soldier Point on Melville Island. From here, the aircraft completed a right turn and descended southwest back towards Darwin, landing from the southeast. A plot of the flight track is shown in Figure

18 + CMP ao poin e ut n t DS ot nbu ua nr dt y ACTIVE ARA TE :46-07:24 UTC + Garden Point Latitude Darwin Longitude Figure 3.2: Flight track for TE02 on 13/11/2005. Arrows indicate the direction of the aircraft track. Numbers to the left of the aircraft track are the seconds since 0000 UTC in increments of 300 secs. Numbers to the right of the aircraft track are the aircraft altitude in km. 18

19 3.3 15/11/2005 : AE General This was the first science flight (HA1) of VH-ARA in Darwin. The SP2 u-bay was fitted to the aircraft. Logger S/N 03 was in the right fuselage position and S/N 04 was in the left fuselage position. The flight was finished early because Hector did not develop. Take-off was approximately 06:16 UTC, landing at approximately 07:59 UTC. The mission scientist was Jörg Hacker Instrumentation and Data Logging The SP2 u-bay was flown and the standard aircraft configuration consisting of the right pylon, Rosemount boom, Novatel GPS and TANS GPS was used. The right wing and fuselage REMlets (remote analogue to digital converters) were triggered by a 20 Hz sample pulse from the right fuselage data logger. The REMlet, Novatel GPS and TANS GPS data were recorded on the right fuselage logger. The digital data from TDL1 were recorded on the left fuselage logger. The right and left fuselage data loggers both recorded the 1 second timing pulse from their internal GPS systems to provide a means for synchronising the time stamp assigned to the various data streams by the loggers. Both loggers and REMlets stopped repeatedly during the flight. The right fuselage data logger was restarted twice times during the flight. The left fuselage data logger stopped completely after 746 seconds, before the aircraft had taken off. The gaps in the data caused by the failure of the right fuselage data logger have been filled using different methods depending on the data type. No attempt has been made to fill gaps in the data from the left fuselage data logger. Gaps in the Novatel GPS position data were filled by extracting the required values from the aircraft position recorded in the OziExplorer track file. Gaps in the Novatel GPS velocity data were filled by differentiating the position data obtained from the OziExplorer track file. Gaps in the temperature and pressure data were filled as follows. First, the decreases in temperature and pressure with height were calculated using all available data. The change in temperature or pressure from the last good measurement was then calculated using these lapse rates and the change in aircraft altitude from the Novatel GPS data. The values used to fill the gaps in T a and p s were then calculated by adding the running integral of the temperature and pressure changes to the last good measurement. The difference between the last value calculated for the gap and the first good measurement of the next data block is recorded and used to characterise the uncertainty of the method. A linear correction is then applied to the values used to fill the gaps to remove any discontinuity between the end of the gap and the start of the next data block. Details of the method and the uncertainty in the values used to fill the gaps are discussed in Section 5. There are no known problems with the meteorological data at this time other than those associated with the gaps caused by the logger failures. 19

20 There is digital data for TDL1 for the first 746 seconds of the flight (aircraft taxiing before take-off). There is no data for TDL2, TSI1 and TSI Flight Track After take-off, the aircraft tracked northeast and climbed to 8,200 m at a point 27 km south of Cape Keith. The aircraft then turned to the west and continued to climb reaching 12,800 m over the southern coast of Bathurst Island. From here, the aircraft tracked east over the southern coasts of Bathurst and Melville Islands climbing to a maximum altitude of 13,800 m north of Cape Gambier. The aircraft then descended to 13,000 m at a point 26 km southeast of Cape Keith before completing a right turn and descending into Darwin from the northeast and landing from the southeast. A plot of the flight track is shown in Figure

21 + CGMP ao paoin e rut dn et DS n ot nbu upa onr indt ty ACTIVE ARA AE :08-08:02 UTC Latitude Darwin Longitude Figure 3.3: Flight track for AE03 on 15/11/2005. Arrows indicate the direction of the aircraft track. Numbers to the left of the aircraft track are the seconds since 0000 UTC in increments of 300 secs. Numbers to the right of the aircraft track are the aircraft altitude in km. 21

22 3.4 16/11/2005 : AE General This was a science flight (HC1) with the NOX u-bay fitted to the aircraft. Logger S/N 03 was in the right fuselage position and S/N 04 was in the left fuselage position. Oil streaks were found on the canopy after the flight. These were traced to a failed start lock in the propeller and the aircraft did not fly again until 27/11/2005. Take-off was approximately 06:28 UTC, landing at approximately 09:47 UTC. The mission scientist was Jörg Hacker Instrumentation and Data Logging The NOX u-bay was flown and the standard aircraft configuration consisting of the right pylon, Rosemount boom, Novatel GPS and TANS GPS was used. The right wing and fuselage REMlets (remote analogue to digital converters) were triggered by a 20 Hz sample pulse from the right fuselage data logger. The REMlet, Novatel GPS and TANS GPS data were recorded on the right fuselage logger. The digital data from TDL1 were recorded on the left fuselage logger. The right and left fuselage data loggers both recorded the 1 second timing pulse from their internal GPS systems to provide a means for synchronising the time stamp assigned to the various data streams by the loggers. Both loggers and REMlets stopped repeatedly during the flight. The right fuselage data logger was restarted eleven times during the flight. The left fuselage data logger stopped completely after 668 seconds, before the aircraft had taken off. The gaps in the data caused by the failure of the right fuselage data logger have been filled using different methods depending on the data type. No attempt has been made to fill gaps in the data from the left fuselage data logger. Gaps in the Novatel GPS position data were filled by extracting the required values from the aircraft position recorded in the OziExplorer track file. Gaps in the Novatel GPS velocity data were filled by differentiating the position data obtained from the OziExplorer track file. Gaps in the temperature and pressure data were filled as follows. First, the decreases in temperature and pressure with height were calculated using all available data. The change in temperature or pressure from the last good measurement was then calculated using these lapse rates and the change in aircraft altitude from the Novatel GPS data. The values used to fill the gaps in T a and p s were then calculated by adding the running integral of the temperature and pressure changes to the last good measurement. The difference between the last value calculated for the gap and the first good measurement of the next data block is recorded and used to characterise the uncertainty of the method. A linear correction is then applied to the values used to fill the gaps to remove any discontinuity between the end of the gap and the start of the next data block. Details of the method and the uncertainty in the values used to fill the gaps are discussed in Section 5. The dynamic pressure data is unreliable from time 07:59:10 UTC due to ice forming in the pitot tube of the Rosemount 5-hole probe. The missing dynamic pressure data 22

23 has been synthesised from the pressure difference across the angle of attack ports on the 5-hole probe. The errors associated with this approach are discussed in Section 5. There is no data for the CR2 frost point hygrometer for this flight. There is data for TDL1 for the first 668 seconds of the flight (aircraft taxiing before take-off). There is no data for TDL2 and no data (analogue or serial) for TSI1 and TSI Flight Track After take-off, the aircraft tracked northeast, climbed to an altitude of 9,300 m 19 km south of Cape Keith and then tracked west while continuing to climb to 12,900 m 27 km northwest of Cape Gambier. From here the aircraft began a run to the southwest at 13,200 m climbing to 13,800 m at the end of the transect 113 km west of Darwin. The aircraft then began a right turn onto the reverse heading and completed a run to the northeast at 12,800 m ending 29 km northwest of Cape Gambier. From here the aircraft stepped 29 km to the southeast and began a run from Cape Gambier to the southwest at 13,800 m descending to 13,200 m at the end of the transect 72 km west of Darwin. After a left procedure turn the aircraft returned on the reverse heading at 13,200 m to a point 32 km southwest of Cape Gambier. From here the aircraft stepped 29 km northwest back to the original track and completed another northeast to southwest transect at 13,200 m from Buchanan Island to a point due south of Cape Fourcroy. The aircraft then tracked northwest to southeast over the Beagle Gulf at 13,200 m before descending to the west over Cape Hotham and landing into Darwin from the southeast. A plot of the flight track is shown in Figure

24 + CMP ao poin e ut n t DS ot nbu ua nr dt y ACTIVE ARA AE :19-09:47 UTC + Garden Point Latitude Darwin Longitude Figure 3.4: Flight track for AE04 on 16/11/2005. Arrows indicate the direction of the aircraft track. Numbers to the left of the aircraft track are the seconds since 0000 UTC in increments of 300 secs. Numbers to the right of the aircraft track are the aircraft altitude in km. 24

25 3.5 27/11/2005 : TE General This was a test flight (HC1) after the propeller repair with the SP2 u-bay fitted to the aircraft. Logger S/N 03 was swapped from the right fuselage position to the left fuselage and S/N 04 was swapped to the right fuselage position. This was done after checks on the ground suggested that S/N 04 was more reliable than S/N 03. The general flight pattern consisted of an ascent to altitude from Darwin over Van Diemen Gulf to the Cobourg Peninsula followed by a short return leg at altitude back towards Darwin and descent into Darwin. Take-off was approximately 06:42 UTC, landing at approximately 08:10 UTC. The mission scientist was Peter Isaac Instrumentation and Data Logging The SP2 u-bay was flown and the standard aircraft configuration consisting of the right pylon, Rosemount boom, Novatel GPS and TANS GPS was used. The right wing and fuselage REMlets (remote analogue to digital converters) were triggered by a 20 Hz sample pulse from the right fuselage data logger. The REMlet, Novatel GPS and TANS GPS data were recorded on the right fuselage logger. The Novatel GPS, TDL1, TDL2, TSI1 and TSI2 data were recorded on the left fuselage logger. The Novatel GPS data (including the 1 PPS timing pulse) were fed to both data loggers for this flight to allow the two loggers to be synchronised and to provide timing for the various data streams. Both loggers operated throughout the flight. There are no known problems with the meteorological data at this time but there is no TANS GPS attitude data. As a result, there is no wind data for this flight. There is data for TDL1 and TDL2 for the flight. The serial data for TSI1 and TSI2, which appear to be diagnostic messages rather than measurements, are scrambled because the wrong serial port configuration was specified in the data logger initialisation file. The analogue data for TSI1 and TSI2 appear correct Flight Track After take-off the aircraft tracked northeast over Cape Hotham climbing to an altitude of 12,800 m at a point 55 km east of Cape Don. The aircraft then completed a left turn onto the reverse heading and climbed 13,200 m while tracking southwest back towards Darwin. The descent into Darwin for landing began 15 km northeast of Cape Hotham with the aircraft landing from the southeast. A plot of the flight track is shown in Figure

26 + GMP o aoin rut dn et S n t Bu upa onr indt ty -11 ACTIVE ARA TE :39-08:14 UTC + Cape Don Latitude Darwin Longitude Figure 3.5: Flight track for TE05 on 27/11/2005. Arrows indicate the direction of the aircraft track. Numbers to the left of the aircraft track are the seconds since 0000 UTC in increments of 300 secs. Numbers to the right of the aircraft track are the aircraft altitude in km. 26

27 3.6 30/11/2005 : AE General This was a science flight (HC1) with the NOX u-bay fitted to the aircraft. The flight pattern consisted of 7 transects oriented southeast to northwest over Dundas Strait. The first 4 of these transects roughly followed the line joining Soldier Point and Point Jahleel on Melville Island. The remaining 3 of these runs extended northwest from Cape Don on the Cobourg Peninsula. After these, the aircraft competed 3 runs oriented northeast to southwest over the Dundas Strait. The third of these transects extended along the southeastern coast of Melville Island. Take-off was approximately 04:14 UTC, landing at approximately 08:11 UTC. The mission scientist was Peter Isaac Instrumentation and Data Logging The NOX u-bay was flown and the standard aircraft configuration consisting of the right pylon, Rosemount boom, Novatel GPS and TANS GPS was used. The right wing and fuselage REMlets (remote analogue to digital converters) were triggered by a 20 Hz sample pulse from the right fuselage data logger. The REMlet, Novatel GPS and TANS GPS data were recorded on the right fuselage logger. The Novatel GPS, TDL1, TDL2, TSI1 and TSI2 data were recorded on the left fuselage logger. The Novatel GPS data (including the 1 PPS timing pulse) were fed to both data loggers for this flight to allow the two loggers to be synchronised and to provide timing for the various data streams. Both loggers and REMlets stopped repeatedly during the flight. The right fuselage data logger was restarted five times during the flight and the left fuselage data logger was restarted twice. The right wing REMlet data (temperature and pressures) stops at a time of (seconds into UTC day). The gaps in the data caused by the failure of the right fuselage data logger have been filled using different methods depending on the data type, see below for a description of these methods. No attempt has been made to fill gaps in the data from the left fuselage data logger. Gaps in the Novatel GPS position data were filled by extracting the required values from the aircraft position recorded in the OziExplorer track file. Gaps in the Novatel GPS velocity data were filled by differentiating the position data obtained from the OziExplorer track file. Gaps in the temperature and pressure data were filled as follows. First, the decreases in temperature and pressure with height were calculated using all available data. The change in temperature or pressure from the last good measurement was then calculated using these lapse rates and the change in aircraft altitude from the Novatel GPS data. The values used to fill the gaps in T a and p s were then calculated by adding the running integral of the temperature and pressure changes to the last good measurement. The difference between the last value calculated for the gap and the first good measurement of the next data block is recorded and used to characterise the uncertainty of the method. A linear correction is then applied to the values used to fill the gaps to remove any discontinuity between 27