NASA - National Aeronautics and Space Administration

Medusa collects 32 cryogenically dried, flow and pressure controlled samples per flight. The samples are collected by an automated sampler into 1.5 L glass flasks that integrate over 25-s (1 e-fold) periods. Medusa provides discretely-sampled comparisons for onboard in situ O2/N2 ratio and CO2 measurements and unique measurements of Ar/N2 and ¹³C, ¹⁴C, and ¹⁸O isotopologues of CO2. The complementary measurements allow ground-truthing of onboard instrument measurements in a laboratory setting, where analysis conditions can often be more stringently controlled and carefully monitored. Isotope and argon measurements can provide additional information about land and ocean controls over the carbon cycle, and about the age and source of the air sampled.

Medusa consists of an onboard computer, two pressure controllers, two
 pumps, three multi-position selector
valves, and a host of other hardware that
control and direct the air samples. All air
is dried by passing it through traps
immersed in a -78 C dry ice bath, adjusted to match atmospheric pressure
at sea level, and then automatically isolated in a flask. Medusa flasks are analyzed on a sector-magnet mass spectrometer and a LiCor non-dispersive infrared CO2 analyzer by the Scripps O2 Program at Scripps Institution of Oceanography.

The NCAR Airborne Oxygen Instrument measures O2 concentration using a vacuum-ultraviolet absorption technique.
 AO2 is based on earlier ship-board and laboratory instruments using the same technique, but has been designed specifically for airborne use to minimize
motion and thermal sensitivity and with a pressure and flow controlled inlet system. To achieve the high levels of precision needed, AO2 switches between sample gas and air
from a high-pressure reference cylinder
every 2.5 seconds. Atmospheric O2 concentrations are typically reported in units
of one part in 1,000,000 relative deviations
in the O2/N2 ratio, which are referred to as
 "per meg." AO2 has a 1-sigma precision of
± 2 per meg on a 5 second measurement.
 For comparison, this is equivalent to detecting the removal of one O2 molecule
 from 2.5 million molecules of air. At typical
flight speeds of 300 kts or climb/descent
rates of 1500 fpm, 5-seconds correspond to
a horizontal resolution of 750 m and a
vertical resolution of 40 m. The instrument includes an internal single-cell CO2 sensor (LI-840), which is used to correct the O2 measurements for dilution by CO2 and for scientific purposes. To minimize inlet
surface effects, the pressure in the inlet line
is actively controlled at the aircraft bulkhead.
The sample air is cryogenically dried in a
series of electropolished stainless steel traps immersed in a dry ice Fluorinert slurry. The
 AO2 system consists of a pump module, a cylinder module, an instrument module, and a dewar.

The Harvard QCLS (DUAL and CO2) instrument package contains 2 separate optical assemblies and calibration systems, and a common data system and power supply. The two systems are mounted in a single standard HIAPER rack, and are described separately below:

The Harvard QCL DUAL instrument simultaneously measures CO, CH4, and N2O concentrations in situ using two thermoelectrically cooled pulsed-quantum cascade lasers (QCL) light sources, a multiple pass absorption cell, and two liquid nitrogen-cooled solid-state detectors. These components are mounted on a temperature-stabilized, vibrationally isolated optical bench with heated cover. The sample air is preconditioned using a Nafion drier (to remove water vapor), and is reduced in pressure to 60 mbar using a Teflon diaphragm pump. The trace gas mixing ratios of air flowing through the multiple pass absorption cell are determined by measuring absorption from their infrared transition lines at 4.59 microns for CO and 7.87 microns for CH4 and N2O using molecular line parameters from the HITRAN data base. In-flight calibrations are performed by replacing the air sample with reference gas every 10 minutes, with a low-span and a high-span gas every 20 minutes. A prototype of this instrument was flown on the NOAA P3 in the summer of 2004.

The Harvard QCL CO2 instrument measures CO2 concentrations in situ using a thermoelectrically cooled pulsed-quantum cascade laser (QCL) light source, gas cells, and liquid nitrogen cooled solid-state detectors. These components are stabilized along the detection axis, vibrationally isolated, and housed in a temperature-controlled pressure vessel. Sample air enters a rear-facing inlet, is preconditioned using a Nafion drier (to remove water vapor), then is reduced in pressure to 60 mbar using a Teflon diaphragm pump. A second water trap, using dry ice, reduces the sample air dewpoint to less than –70C prior to detection. The CO2 mixing ratio of air flowing through the sample gas cell is determined by measuring absorption from a single infrared transition line at 4.32 microns relative to a reference gas of known concentration. In-flight calibrations are performed by replacing the air sample with reference gas every 10 minutes, and with a low-span and a high-span gas every 20 minutes.

The NASA Langley CO2 sampling system (AVOCET) has an extensive measurement heritage in tropospheric field campaigns, delivering high reliability over 3400 flight hours (452 science flights) and is recognized within the CO2 community as a benchmark for evaluating newly evolving remote CO2. This instrument was adapted by the investigators for airborne sampling and has been successfully deployed aboard NASA research aircraft beginning with the PEM-West A mission in 1992, and more recently during the 2016 KORUS-AQ, 2017 ACSENDS/ABoVE, and 2019 FIREX-AQ missions. The newest iteration of the technique as of 2017 has at its core a modified LI-COR model 7000 non-dispersive infrared spectrometer (NDIR). The basic instrument is small (13 x 25 x 37 cm) and composed of dual 11.9 cm^3 sample/reference cells, a feedback stabilized infrared source, 500 Hz chopper, thermoelectrically-cooled solid state PbSe detector, and a narrow band (150 nm) interference filter centered on the 4.26 μm CO2 absorption band. Using synchronous signal detection techniques, it operates by sensing the difference in light absorption between the continuously flowing sample and reference gases occupying each side of the dual absorption cell. Thus, by selecting a reference gas of approximately the same concentration as background air (~405 ppm), minute fluctuations in atmospheric concentration can be quantified with high precision. Calibrations are performed frequently during flight using WMO-traceable standards from NOAA ESRL. Precisions of ≤ 0.1 ppm (±1σ) for 1 Hz sampling rates are typical for our present airborne CO2 system when operated at 600 torr sample pressure.

The absorption of infrared solar radiation along a slant path to the sun is recorded from 2 to 15 micrometers. Six spectral filters are used to cover the region from 2-15 microns. An interferogram is recorded in about 10 seconds. Interferograms are transformed to produce spectra. Column amounts are retrieved by fitting the observed spectra using the non-linear least squares fitting code SFIT2 that employs an Optimal Estimation retrieval algorithm.

The major chlorine reservoirs (HCl and ClONO2), the important nitrogen-containing gases in the stratosphere (N2O, NO, NO2, and HNO3), stratospheric and tropospheric tracers (HF, CH4, C2H6, H2O, CO2), a major source CFC (CF2Cl2) and ozone may be routinely retrieved.

A Fabry-Perot interferometer is constructed of two very flat, partially reflecting mirrors held parallel to one another at a fixed distance. Interference occurs among the multiple reflections leading to the condition that wavelengths that exactly divide the spacing between the mirrors by an integer are transmitted very efficiently and all other wavelengths are reflected. Thus if the plates are held fixed at a separation of 10 μm, then radiation at 10, 5, 3.333, … μm will be transmitted. Note that these wavelengths are equally spaced in energy according to the relationship E=hc/l, where l is the wavelength of the light and h and c are Planck’s constant and the speed of light, respectively. This particular FPI technique makes use of these multiple passbands to increase the measurement signal and the resulting signal to noise ratio.

A Fabry-Perot can be tuned to transmit different wavelengths by changing the (optical) spacing between the mirrors. This is commonly done by employing piezo-electric transducers to translate the mirrors by very small distances, while maintaining the very precise parallelism between them. Fixed gap Fabry-Perots can be tuned by tilting, which changes the effective path length between the plates; by using the thermal expansion and contraction of the spacers between the mirrors; and by changing the composition or pressure of the gas that fills the space between the plates, which alters the index of refraction thereby changing the optical separation. Finally, Fabry-Perots can be constructed using a solid substrate of fused silica or optical quality glass onto which reflective coatings are deposited. These devices can be angle tuned or temperature tuned.

NIRAD consists of three systems: (1) CO2 detector, (2) power and data acquisition, and (3) gas-handling. All three systems have flown previously. The CO2 detector was first flown in 1999 as part of CORE+ instrument during RISO and ACCENT and again in 2004 during PUMA-A. There have been no changes to the detector, other than inspection and routine maintenance. The power and data acquisition system were new for PUMA-A, and are flown here without change, other than to software. The gas-handling system is the same as that flown in May 2004, except that it is now packaged into a single box that contains the detector and power/data system.

The detector is packaged in a vacuum housing to facilitate management of temperature and pressure. At power-up the housing is pumped down to ~300 hPa by one stage of a diaphragm pump and held at this pressure throughout the flight. Thus, at pressure altitudes < 300 hPa the pressure within the housing is above ambient. By design, if the pressure differential is significantly greater than about 5 psi, the O-ring seals leak. A redundant additional mechanical safety relief valve (set for ~15 psi or less) is placed on the housing.

Two 1.2 L epoxy-coated, fiber-wrapped aluminum bottles (DOT rated and certified) are filled to ~1600 psi before flight with zero air doped with CO2. These ‘standards’ are sampled repeatedly during flight to provide an accurate standard for reference to the NOAA/CMDL CO2 scale. Two-stage regulators provide a service pressure of ~25-30 psig throughout flight. The bottles and regulators are backed with safety relief valves.

The diaphragm pump is current-limited for a ‘soft start’ (that is, there is no electrical surge on startup, allowing for use of compact, highly efficient Vicor VI-100 DC/DC converters.

The MkIV interferometer operates in solar absorption mode, meaning that direct sunlight is spectrally analyzed and the amount of various gases at different heights in the Earth's atmosphere is derived from the shapes and depths of their absorption lines. The optical design of the MkIV interferometer is based largely on that of the ATMOS instrument, which has flown four times on the Space Shuttle. The first three mirrors in the optical path comprise the suntracker. Two of these mirrors are servo-controlled in order to compensate for any angular motion of the observation platform. The subsequent wedged KBr plates, flats, and cube-corner retro-reflectors comprise a double-passed Michelson interferometer, whose function is to impart a wavelength-dependent modulation to the solar beam. This is achieved by sliding one of the retro-reflectors at a uniform velocity so that the recombining beams interfere with each other. A paraboloid then focusses the solar beam onto infrared detectors, which measure the interferometrically modulated solar signal. Finally, Fourier transformation of the recorded detector outputs yields the solar spectrum. An important advantage of the MkIV Interferometer is that by employing a dichroic to feed two detectors in parallel, a HgCdTe photoconductor for the low frequencies (650-1850 cm-1) and a InSb photodiode for the high frequencies (1850-5650 cm-1), the entire mid-infrared region can be observed simultaneously with good linearity and signal-to-noise ratio. In this region over 30 different gases have identifiable spectral signatures including H2O, O3, N2O, CO, CH4, NO, NO2, HNO3, HNO4, N2O5, H2O2, ClNO3, HOCl, HCl, HF, COF2, CF4, SF6, CF2ClCFCl2, CHF2Cl, CF2Cl2, CFCl3, CCl4, CH3Cl, C2H2, C2H6, OCS, HCN, N2, O2, CO2 and many isotopic variants. The last three named gases, having well known atmospheric abundances, are important in establishing the observation geometry of each spectrum, which otherwise can be a major source of uncertainty. Similarly, from analysis of T-sensitive CO2 lines, the temperature profile can be accurately determined. The simultaneity of the observations of all these gases greatly simplifies the interpretation of the results, which are used for testing computer models of atmospheric transport and chemistry, validation of satellite data, and trend determination.

Although the MkIV can measure gas column abundances at any time during the day, the highest sensitivity to atmospheric trace gases is obtained by observing sunrise or sunset from a balloon. The very long (~ 400 km) atmospheric paths traversed by incoming rays in this observation geometry also make this so-called solar occultation technique insensitive to local contamination.