3. DATA COLLECTION AND PROCESSING

3.1 TYPES OF INFORMATION COLLECTED

The types of information collected for flights monitored during the study included: (1) activities related to the flight, such as smoking information and passenger data, (2) continuous monitoring information for pollutants and other parameters, and (3) concentrations of contaminants collected as time-integrated samples. Sections 3.1.1, 3.1.2, and 3.1.3 contain descriptions of information from daily flight documentation logs, continuous monitoring and integrated sampling.

3.1.1 Daily Flight Documentation Log

TABLE 3-1. FLIGHT CHARACTERISTICS, AIRCRAFT
INFORMATION AND PASSENGER DATA FROM
THE FLIGHT DOCUMENTATION LOG

Parameter	Page Location in Flight Documentation Log
1. Flight date 2	2
2. Airline 2	2
3. Flight number 2	2
4. Airport of departure 2	2
5. Airport of arrival 2	2
6. Aircraft model number 2	2
7. Aircraft registration number 2	2
8. Number of passengers2	2

As described in Section 2.5, the Daily Log was divided into 3 major sections: (1) Start of Day Documentation Log, (2) Flight Documentation Log, and (3) End of Day Documentation Log. The information collected in the Flight Documentation Log and the End of Day Documentation Log was grouped in the following four categories for purposes of data entry and processing:

• Flight characteristics, aircraft information, and passenger data
• Smoking information
• Time of particular flight activities and technician locations
• Information relating to instrumentation and sampling media.

Flight characteristics (Table 3-1) included the date of the flight, the airline and flight number, and the airports of departure and arrival. Aircraft information included the model of the airplane (e.g., Boeing 727-200 or DC10-30) and the registration number of the plane found on the outside of the aircraft. The primary passenger information was the total number of passengers, excluding the crew, on the plane.

TABLE 3-2. SMOKING INFORMATION FROM THE FLIGHT DOCUMENTATION LOG

Parameter	Page Location in Flight Documentation Log
1. Ashtrays emptied at start of flight?	4
2. Smoking rows	4
3. Number of passengers in smoking section	4
4. Number of passengers in boundary section	4
5. Smoking counts during one-minute intervals every 15 minutes	4, 5
6. Number of seats from which cigarette butts were collected	8
7. Total number of cigarette butts collected	8
8. Was previous flight smoking?	8

The smoking information (Table 3-2) collected by the field technicians and recorded in the Flight Documentation Logs included the identification of coach smoking rows and an observation at the beginning of the flight on whether or not the ashtrays were emptied prior to boarding. Additionally, the technician in the coach smoking section was required to count the number of cigarettes smoked during a one-minute interval every 15 minutes. These observations were recorded on pages 4 and 5 of the Flight Documentation Log and the observed counts were used as an input to estimation of total cigarettes smoked during each smoking flight. Procedures for estimating total smoking in the coach section are described in Section 3.2.

Also included as smoking information was an indication of whether the previous flight was smoking, as reported by the flight attendant. At the end of the flight, cigarette butts were collected from seats in the coach smoking section. The number of seats from which butts were collected and the total number of butts collected were recorded in the logbook.

TABLE 3-3. VARIABLES WITHIN THE FLIGHT DOCUMENTATION
LOG RELATED TO TIMES OF VARIOUS FLIGHT MILESTONES
AND LOCATION OF TECHNICIAN

Parameter	Page Location in Flight Documentation Log
1. Seat number	2
2. Section number	2
3. Boarding time	3
4. Time of departure from gate	3
5. Time of takeoff	3
6. Time when no-smoking light was turned off	3
7. Time when cruise altitude was reached	3
8. Time of cruise descent	8
9. Time when no-smoking light was turned on	8
10. Time of arrival at gate	8

Table 3-3 lists the information recorded about technician locations and the times of various events during flights. The location of the technician included the seat number and the section number. The target locations for technician seating on a smoking flight were the snaking section, one of the boundary rows, the middle of the nonsmoking section, and the remote location (typically near the front) in the nonsmoking section. On international flights, a boundary seat in the business class was substituted for the nonsmoking remote location, and on nonsmoking flights technicians were seated in the section of the plane where smokers would be assigned on smoking flights (usually the rear) and the middle of the nonsmoking section. Flight events that were recorded included the time when the aircraft was boarded and the time when cruise altitude was reached. Of particular importance were the times when the no-smoking light was turned off and turned on. The interval between these two events was used as the timeframe for averaging temperature, relative humidity, pressure, and pollutant measurements that were recorded with continuous monitoring devices.

Information related to instrumentation and sampling media, shown in Table 3-4, included identification numbers of sampling devices and the times when sampling pumps were turned on and off. Within the End of the Day Documentation Log, the MINIRAM zero values, the MINIRAM pump flow rate, the nicotine pump flow rate, and the ozone pump flow rate were recorded. Each of these items was ultimately used in the computation of measured concentrations.

3.1 Continuous Monitoring Data

TABLE 3-4. VARIABLES WITHIN THE FLIGHT DOCUMENTATION
LOG RELATED TO INSTRUMENTATION AND SAMPLING MEDIA

Parameter	Page Location in Flight Documentation Log
1. Nicotine/RSP cassette ID numbers	1
2. C02 tube ID number	1
3. Ozone cassette ID number	1
4. CAT sampler ID number	1
5. PFT sources	1
6. Time sensors turned on	1
7. Instrument package number	2
8. SAS package number	2
9. Time PFT sources deployed	3
10. Time MINIRAM pump turned on	3
11. Time nicotine pump turned on	3
12. Time C02 tube opened	3
13. Time ozone pump turned on	3
14. Start time of bioaerosol sampling	7
15. Bioaerosol plate ID numbers	7
16. Stop time of bioaerosol sampling	7
17. Time ozone pump turned off	8
18. Time nicotine pump turned off	8
19. Time C02 tube capped	8
20. C02 tube reading/time of reading	8
21. Time MINIRAM pump turned off	8
22. MINIRAM zero checks	3
23. MINIRAM flow rate	4
24. Nicotine pump flow rate	5
25. Duplicate nicotine pump flow rate	6
26. Ozone pump flow rate	7

Continuous monitoring data were collected at all four locations on smoking flights and at both locations on nonsmoking flights. A data logger was programmed to compute and record average measurement values every minute. The Julian date, hour, minute, RSP, C0, temperature, relative humidity, and pressure values were recorded on individual channels. This information was stored in the internal memory of the data logger and transferred to computer diskettes at the end of each day. The file-naming convention was keyed to the Julian date and the identification number of the data logger used by a particular technician (e.g., 102-1477.PRN). Following file transfers at the end of each day, a backup of each transferred file was made.

3.1.3 Integrated Sampling Media

TABLE 3-5. PLACEMENT LOCATIONS FOR INTEGRATED
SAMPLING DEVICES

Measurement Parameter	Location
	Smoking	Nonsmoking
		Boundary	Middle	Remote
A. Smoking Flights
Nicotine/RSP	X	X	X	X
Ozone			X	X
C02	X			X
Microbial aerosols	X			X
PFT sources	X			X
PFT samplers			X	X
B. Nonsmoking Flights*
Nicotine/RSP			X	X
Ozone			X	X
C02				X
Microbial aerosols				X
PFT sources				X
PFT samplers				X

As described in Section 2.3, integrated sampling devices were used to collect samples for nicotine, RSP, ozone, C02, microbial aerosols, and air exchange rates. Nicotine and RSP samples were collected at all locations on every flight. Ozone, C02, and microbial aerosols were collected at two sites on smoking flights and international flights and at one site on nonsmoking flights. Table 3-5 summarizes the locations of integrated sampling devices on smoking and nonsmoking flights. PFT sources for air exchange measurements were deployed in the remote and smoking locations, whereas samplers (CATs) were deployed in the boundary and central nonsmoking locations.

Table 3-6 lists the laboratory destination for each type of sampling device. C02 concentrations were read by the technicians during each flight; the time of the analysis and the concentration were recorded in the Flight Documentation Log. The ozone samples were analyzed by GEOMET’s laboratory, whereas the nicotine RSP samples, microbial aerosol samples, and PFTs were analyzed by external laboratories.

TABLE 3-6. LABORATORY ANALYSIS RESPONSIBILITY
FOR INTEGRATED SAMPLES

Type of Sample	Laboratory Responsible for Analysis
Nicotine/RSP	University of Massachusetts
Ozone	GEOMET
C02 diffusion tubes	Technicians (during flight)
Microbial aerosols	Pathogen Control Associates
PFT samples	Brookhaven National Laboratory

3.2 DATA PROCESSING PROCEDURES

The field documentation collected by the technicians was returned to GEOMET for processing and analysis. Several different software packages were used during processing including dBase III Plus, Lotus 1-2-3, Microsoft QUICKBASIC, and SPSS/PC. Section 3.2.1 reviews the processing of data recorded in the Daily Flight Documentation Logs, Section 3.2.2 includes an explanation of continuous monitoring data processing, and Section 3.2.3 discusses processing of integrated sample data. Procedures for estimating total smoking rates in the coach smoking section, based on technician observations, are described in Section 3.2.4. Supplemental information that was gathered independently is described in Section 3.2.5.

3.2.1 Daily Flight Documentation Logs

Information collected by the field technicians and recorded in the Daily Log was entered into a database using dBase III Plus software. The database contained one record for each technician location on each flight. Data were entered from the Flight Documentation Logs and the End of the Day Documentation Logs. Information from the end of the day was entered for each flight during the day to which it applied. Each daily log was assigned an identification number, and this number was also entered into the database; this practice enabled easy reference to a particular log in the event that further review was needed.

Initially, each chain of flights was entered in a entered database for easy reference. Information from all ten chains was ultimately united in a single database.

The sampler identification numbers were entered into the database twice, as they appeared in the logbook. This practice enabled additional quality control checks to ensure that the first reported identification number matched the final reported number. Additional fields were provided to capture information relating to duplicate samplers.

3.2.2 Continuous Monitoring Data

The continuous monitoring data were processed using a BASIC program that combined (1) data logger outputs (voltages) for each channel, (2) calibration factors for converting the voltages to engineering units, and (3) selected information extracted from the dBase III Plus file for Flight Documentation Logs.

CO multipoint calibrations were performed at the beginning and the end of each chain at GEOMET’s laboratory. Regression analysis was applied to the beginning and ending calibrations to calculate beginning and ending slopes and intercepts. This information was entered into an ASCII file along with the data logger identification number, MINIRAM identification number, initial zero value (MINIRAM), and the CO monitor identification number. These files were specific to a chain and were referred to as ‘set’ files. The file was sorted by instrument package identification number, and the final line in the file indicated the date and time of the initial and final CO calibrations.

In addition to the set files, the following information was extracted from the dBase III Plus tile described previously:

• Daily Log identification number
• Flight date
• Instrumentation package number
• Airline
• Flight number
• Sent number
• Section location
• MINIRAM on time
• No-smoking light off time
• No-smoking flight on time
• MINIRAM off time
• First MINIRAM zero reading
• Second MINIRAM zero rending
• Third MINIRAM zero reading

These files, each specific to a chain of flights, were referred to as “case” files. The case files and set files were used together as inputs to the processing routine.

The raw data files, as discussed in Section 3.1.2, were named according to the Julian date on which the data were collected and the technician’s data logger identification number. The raw data files for a particular chain, along with the set file and the case file, were inputs to a BASIC program. Figure 3-1 contains a flow diagram depicting the procedure that was followed during processing of the continuous monitoring data. The program read the first line of the case file and identified the flight date and the instrument package identification number. within the program, the data logger assigned to each package was identified. Based on flight date and data logger number, the proper raw data file was retrieved. Then the MINIRAM identification number, the flight number, and the seat number were identified from the set and case files. The CO data were calibrated using the slope/intercept information contained in the set file; a linear drift between beginning and ending calibrations was assumed.

Reports were produced for each technician location on each flight. Figure 3-2 is an example of a report produced through the combination of the three inputs (raw data, set file, case file). Selected information from the set file and the case file is listed at the top of each report, including the instrument identification numbers, the flight date, the seat number and section location, the time when the MINIRAM pump was turned on and off, the time when the no-smoking light was turned off and on, and MINIRAM zero values. Program outputs included the input and output file names and the average, minimum, and maximum values for RSP, C0, temperature, relative humidity, and pressure. These values were reported for the entire period when smoking was allowed as well as the periods before and after the smoking period and successive hours during the smoking period.

The second type of output produced by the BASIC program was continuous calibrated data saved in files specific to each flight and seat location. Data in these files were later used for more in-depth analyses (e.g., peak versus average concentrations)

3.2.3 Integrated Samples

Results from laboratory analysis of integrated samples were combined with information from the flight documentation logs to calculate measured concentrations. The logs provided the flight information and the length of the sampling period. In calculating the concentrations from the integrated sampling results, selected outputs (e.g., temperature, pressure) from the continuous monitoring were also needed in some instances.

The concentrations for Gravimetric nicotine/RSP were calculated using sample mass from the laboratory, flow rates and sampling duration from the flight documentation log, and temperature and pressure during the smoking period from the continuous monitoring data. Figure 3-3 illustrates the procedures used to calculate the nicotine and RSP concentrations in ug/m3.

Ozone concentrations were calculated in parts per million (ppm) in much the same way as nicotine and RSP concentrations were calculated. The duration of the sampling period and the pump flow rate were extracted from the Daily Log database and the sample mass was provided by the laboratory. The pump flow rate was measured by the field technician.

Draeger Tubes were used to collect C02 in two locations in the plane. The diffusion tubes were filled with a blue indicator compound that gradually turned white as C02 diffused into the tube. At the end of the flight the field technician read the C02 level by noting where the white coloration stopped. The analytical range for these tubes was from 500 ppm/hr to 20,000 ppm/hr.

Most of the data required to calculate the C02 concentration were extracted from the Daily Log. One field (pressure) was extracted from the continuous monitoring data; the average pressure during the smoking portion of the flight was used. The average C02 concentration, in ppm, was calculated by (1) applying a correction factor derived from the average pressure measurement to the raw integrated value and (2) dividing the corrected integrated value by measurement durations.

Figure 3-4 is a flow chart depicting the data processing procedure for calculating the air exchange rate on each flight. Samplers (CATs) were analyzed at Brookhaven National Laboratory and the quantity of each tracer gas found in each tube was reported. This information was combined with data extracted from the Daily Log database on the type of source deployed on a particular flight and the length of exposure for the CAT. The final inputs to the calculation were average temperature and pressure during the smoking period. A file containing the above inputs was processed using a BASIC program; the output was a report including the air exchange rate per location and the average airflow rate between locations for the flight.

The results of the bioaerosol sampling were reported by the laboratory in colony-forming units per cubic meter (CFU/m3). These results were linked with the flight date, airline, flight number, seat number, and section. Total bacterial concentrations as well as concentrations of Staphylococcus aureus and Streptococcus pyogenes were reported for each sample. In addition, concentrations of several other types of bacteria were reported. For example, among the most prevalent types were:

• Staphylococcus not aureus
• Micrococcus varians
• Micrococcus luteus
• Micrococcus lylae
• Corynebacterium .

Total fungi were also reported together with the most prominent genera.

3.2.4 Estimation of Smoking Rates

Estimated smoking rates were calculated using the data recorded by the technician seated in the coach smoking section. One of the inputs to the calculation of smoking rates was the number of people smoking during a one-minute period every 15 minutes as recorded by the technician in the smoking section. The estimated quantity of cigarettes smoked during the flight, in cigarette-minutes, was, calculated by the following formula:

(Smoking Duration x 60)
———————————– x (Smoking Count)
(Number of 15-minute intervals)

The result of this calculation was divided by 6, the typical number of minutes a cigarette was lit in the cabin environment (based on technician observations during the pretest), to obtain an estimate for the number of cigarettes smoked during the flight.

3.2.5 Supplemental Information

Additional information on aircraft characteristics was gathered from archived data and keyed into a separate database. This information included such aircraft features as the volume of the plane, whether a plane was a wide body or narrow body, and the nominal extent of cabin air recirculation for each type of aircraft. The total seating capacity was obtained for the aircraft specific to each monitored flight. The contacts with each airline were also used to verify certain data that were collected by field technicians. Each airline was requested to provide the aircraft type, aircraft registration number, total passenger count, and smoking rows for each monitored flight involving the airline.