My primary research effort is an ongoing project to determine the pathways by which water vapor reaches the stratosphere using high-altitude measurements of a new atmospheric tracer. The tracer is the isotopic composition of that water: the tiny amounts of naturally occurring HDO or H₂¹⁸O relative to ordinary H₂¹⁶O. This strategy was proposed in 1996 (Moyer et al., 1996), and since 2001 I have been working as part of the Anderson group at Harvard University on developing techniques and instrumentation to make the measurement. The resulting instrument (the Harvard ICOS Isotope Instrument) made its first flights in Dec. 2004 on NASA's WB-57 high-altitude research aircraft and has been selected for a 2006 science mission to the tropics, the region where most air ascends to the stratosphere. We expect to obtain the first high-resolution, high-precision profiles of water vapor isotopic composition across the tropical tropopause.

Understanding the processes that distribute water in the atmosphere, and possible future changes in that water, is especially important as human-produced greenhouse gases continue to build up in the atmosphere. Water vapor is itself the most important greenhouse gas, and water in the dry stratosphere plays a crucial role in ozone destruction. Measurements of the isotopic composition of water vapor can provide a crucial contribution to this understanding. Because the heavier isotopologues (HDO and H₂¹⁸O) "prefer" to condense out, water vapor isotopic composition carries with it a record of the entire condensation and evaporation history of an air parcel. Water in the dry stratosphere is strongly depleted of its heavier isotopologues: HDO/H₂O there is ~65% less than at the sea surface. (In geochemical terms, dD ~ -650 per mil) Fast-rising convective systems, on the other hand, carry with them a large amount of ice that has condensed at lower altitudes and is therefore much less isotopically depleted. When that ice subsequently evaporates, it leaves behind a characteristic isotopic signature. This effect means that isotopic measurements can address a major question that has troubled atmospheric science for over fifty years. Deep convective systems are frequently observed punching through the tropopause and depositing ice in the stratosphere proper, but the importance of this to the total water budget is still unknown. In winter 2006 the Harvard ICOS instrument will fly out of Costa Rica, in the tropics, and will look for isotopic markers of convective water in the stratosphere.

Measuring HDO in the dry stratsophere is extremely difficult: stratospheric water vapor mixing ratio is only ~5 ppm, and the concentration of HDO is 10,000 times less. Traditional spectroscopic techniques do not have the sensitivity needed to measure small changes in isotopic ratio. For this reason we worked to develop a new, more sensitive measurement technique (ICOS, or Integrated Cavity Output Spectroscopy). The sensitivity of an absorption spectroscopy measurement is directly related to its optical pathlength: the longer that light passes through an air sample, the stronger the resulting absorption features. ICOS is a variant of cavity ringdown spectroscopy, in which pulses of light are coupled into a high-finesse cavity and trapped for thousands of bounces between mirror surfaces. ICOS uses the same optical layout but with a cw instead of a pulsed laser. When the laser frequency is scanned slowly across an absorption feature of interest, the result is effectively an absorption spectrum whose optical pathlength is the distance traveled during the cavity ringdown time: in our case, a distance of 4.8 km, nearly 50 times longer than traditional spectroscopic instruments, with a resulting 50 times improvement in sensitivity. The Harvard ICOS instrument is now the most sensitive mid-IR spectrometer ever flown on an airplane. Collaborators on this effort include current and former postdocs and research associates Frank Keutsch, Tom Hanisco, Josh Paul, and Jesse Kroll, graduate students Greg Engel, David Sayres and Jason St. Clair, and engineers Norton Allen, Larry Lapson, Mike Greenberg, Joe Demusz, Marco Rivero, and Terry Martin. Funding was provided by NASA's Instrument Incubator Program.

The Harvard ICOS Isotope Instrument flies on NASA's WB-57 aircraft, a former surveillance plane that can reach 20 km (60,000 feet) altitude carrying 4000 pounds of payload. ICOS currently flies in a pallet in the belly of the aircraft, with a rear-facing inlet pulling in particle-free air, so that we obtain measurements of water vapor uncontaminated by ice even when flying through thick cloud. Cell pressure is regulated at 40 torr by a pump and throttle valve. The spatial resolution of the resulting measurements is the distance flown by the aircraft during the flush time of the optical cell (~3 s, for a resolution of 600 m, at or below the size of most convective features). While ambient temperature during ascent drops by 100C, temperature within the optical enclosure is regulated to within a few degrees of 25C, with temperature gradients across the optical cell of less than 0.5 degrees. Laser frequency is monitored continuously by a Ge etalon, and the base laser power curve is measured during periodic calibration cycles by flushing the optical cell with dry air. Calibration cycles are computer-controlled: the instrument operates autonomously, controlled only by an on/off switch in the cockpit. To date the instrument has made eight flights, all out of Houston, TX: four during test flights in winter 2004 and four during the AVE-WIIF mission in Jun.-Jul. 2005, both times in conjunction with other Harvard instruments measuring water vapor and HDO. ICOS in-flight sensitivity was comparable to that in the laboratory and met our science target, with s dD of 15 per mil in the troposphere and 40 per mil in the dry stratosphere (or 25 per mil with 10 s averaging). For geochemists, these numbers may seem large, but remember that the variations in atmospheric water dD are also large by geochemical standards, at ~70-100's per mil. The strength of the water fractionation makes isotopic science on atmospheric water vapor compatible with the high-data-rate in-situ instruments that provide the requisite spatial resolution.

Data from the Houston flights have provided considerable insight into the isotopic effects of convective processes. The data show the essential utility of water isotopic measurements, revealing clearly events such as the mixing of different airmasses and the evaporation of convective ice. During AVE-WIIF, in the summer, a time of strong convective activity, the data show that significant convective detrainment begins well below the tropopause, and that water isotopic composition throughout the upper troposphere is enriched substantially beyond the Rayleigh prediction. Convective influence clearly extends above the tropopause, and the data allow the first-ever isotopic identification of an evaporation convective remnant in the stratosphere proper. Throughout the atmosphere the isotopic ratios H₂¹⁸O/H₂O and HDO/H₂O follow each other as predicted by theory, showing no evidence of unforeseen physical processes, condensation in strongly supersaturated conditions, or mixing between extreme endmember airmasses (click on figure at right for detail). Analysis of the results is ongoing and preprints will be made available on this site. Given the richness of the Houston data we have high expectations for the upcoming tropical mission.