The cloudy, humid atmosphere remains our most important weather and climate forecasting challenge. Airborne remote sensors such as radars and lidars have revolutionized information on aerosol, moisture, cloud, and precipitation (liquid and ice) vertical structure, far increasing the information gathered from in situ measurements alone. In recognition, the National Science Foundation (NSF) has expanded its aircraft deployment resources to the remote sensors listed in Table 1. These also include a profiling microwave radiometer sensitive to the atmospheric temperature structure. However, microwave observations from which integrated water vapor and liquid water paths and free-tropospheric humidity profiles can be retrieved are not yet available. Clouds and water vapor are semitransparent in the micro/millimeter wavelength spectral range, in contrast to infrared. The atmospheric emission can be used to infer atmospheric thermodynamics and cloud information in almost all conditions. The integrated water-phase measurements provide important geophysical constraints on hydrometeor and vapor profiles derived from active sensors, and the profiling and mapping of the atmosphere is more comprehensive than that available from in situ observations. Simultaneously, technological improvements are producing ever more miniaturized, modular, and electronically stable designs that consume less power, allowing micro- and millimeter-wavelength radiometers to fit into standard wing-mounted canisters. The National Aeronautics and Space Administration (NASA) and several European agencies operate profiling radiometers, but, along with the Department of Energy (DOE), lack airborne radiometers capable of sensing integrated vapor and liquid that are compact enough to easily integrate into a synergistic instrumental suite.
These considerations motivated a workshop on airborne radiometry for water vapor and liquid water retrievals held at the National Center for Atmospheric Research (NCAR) on 23–24 September 2014. The workshop provided a critical opportunity to revisit U.S. NSF capabilities, forge a community consensus on scientific requirements, and discuss infrastructure and institutional resources. These discussions culminated in a set of technical and institutional process recommendations.
The science goals were distinguished through two foci. The first is a critical need for path-integrated liquid water and water vapor measurements for clouds occupying well-mixed layers, possibly in conditions cold enough to support mixed-phase microphysics. Such cloud characterization, useful in itself, also geophysically constrains retrievals of cloud vertical structure from active sensors. Multifrequency measurements deliver robust retrievals within a range of humidity environments, spanning the low latitudes to the Arctic. The second focus is on three-dimensional humidity mapping, driven by mesoscale research into Earth’s water cycle but also applicable to climate, where new understanding is linking moisture transport to the low cloud feedback. Specific science goals include understanding the humidity structure of hurricanes and atmospheric rivers, and the marine boundary layer and its overlying troposphere, understanding the relationship of humidity to convection including orographic redistributions, and improving cloud forecast prediction. Both foci highlight the free troposphere and below.
Microwave emission is illustrated from the vantage point of an aircraft radiometer flying at 0.5 and 1.2 km looking up in Fig. 1, and at 12 km looking down in Fig. 2. The opacity of the cloudy atmosphere increases with frequency as a quasiquadratic function of liquid water path, punctuated by water vapor and oxygen absorption bands (Fig. 1). Sampling about the water vapor bands provides integrated water vapor and liquid water measurements, with measurements between 20 and 30 GHz most useful within moist environments because the absorption band is too weak to saturate, and measurements around 183 GHz suited to dry conditions where the strong sensitivity of the measurements to water vapor can be exploited. Figure 2 illustrates weighting functions (the sensitivity of microwave transmission to the vertical distribution of the water vapor, from which atmospheric humidity profiles can be produced) around 183 GHz. These frequencies are used by satellites as well (e.g., the Advanced Microwave Sounder Unit), in part because higher-frequency channels offer higher resolution and/or can be operated with smaller antennas.
BEST PRACTICES INSTRUMENT MENTORSHIP.
After careful discussion, participants concluded that the calibration and retrieval development is best done by the scientists themselves, rather than by the vendor or developer. This is because vendors differ in their calibration approaches, and a consensus on the best approach is still lacking, with information sharing limited by patent rights in some cases. Instead, calibration and subsequent model evaluation is best achieved by comparison to the many DOE radiometers at its Southern Great Plains site because of DOE’s long history of evaluating multiple-vendor instruments. The retrieval development and evaluation by scientists is also not hampered by proprietary software. Overall, scientists are best motivated to develop and document retrievals meeting standards of accuracy, reproducibility, and independent verification. Such motivations, for example, encourage retrievals that incorporate the latest gaseous absorption coefficients, still an area of active research at the higher frequencies.
The needed retrieval products can also be divided into the integrated amounts (liquid water and water vapor) and profiles of humidity. In both cases, the challenge is to determine a process for making the best performing radiometer available to the research community. Once a requestable instrument is available for path-integrated measurements, responsibility for it and its retrievals could reside with a university scientist as well as within NCAR. Airborne radiometers capable of humidity profiling or scanning require more support. The retrieval development of profiles requires more sophistication than of path-integrated values. Retrieval methods can in theory be developed from early, dedicated test flight datasets and provided as part of a subsequent campaign instrument request. The initial retrieval development, not linked to any particular campaign, would most likely require collaboration between university or government laboratory scientists, instrument developers, and NCAR scientists (or else need to motivate new activity within NCAR).
Technological improvements are producing passive radiometer designs that enhance the overall synergism of airborne measurements, providing new insights into Earth’s cloudy atmosphere. No radiometers capable of sensing any of water’s phases are available within the NSF deployment pool. We recommend against the leasing arrangements that satisfy short-term campaign requirements, as they do not encourage the development of an optimal radiometer and do not provide good instrument mentorship, nor support practices best suited for scientific advancement. Instead, we recommend additional investment in microwave and millimeter-wavelength radiometers in the United States toward providing critical cloud and atmospheric characterization and to support a future three-dimensional thermodynamic mapping capability.
Radiometers emphasizing a compact design should fit into a standard cloud probe canister. Irrespective of size, radiometers should possess
multifrequency humidity and liquid water sensing (20–30, 90, 183 GHz) capability;
modularity of frequency components;
the ability to be deployed upon multiple aircraft;
views above and below the aircraft;
complete documentation of internal measurements and calibration procedures;
vendor-independent calibration and retrieval development plan; and,
ground-based deployability to aid radiometer assessment at the Southern Great Plains DOE Atmospheric Radiation Measurement site.
Processes by which NSF might promote the development and/or acquisition of a new sensor include a joint development effort between NCAR and another entity (university, other government laboratory, or private-sector company) that already possesses the development capability. If a prototype sensor of interest already exists, we encourage further investment to develop the appropriate technical readiness for scientific fieldwork. This could include an expansion of NCAR’s Airborne Research Instrumentation Testing Opportunity (ARISTO) program to include vendor-independent calibration and retrieval development. Sensor maintenance, calibration, operation, and retrieval development could occur at either NCAR, another government laboratory, or at a university, and follow “best practices” for instrument mentorship.
The full workshop report is available at www.eol.ucar.edu/content/workshop-airborne-radiometry-water-vapor-and-liquid-water-retrievals. Measurement integration based on mixed-layer cloud science is discussed in Wood et al. (2011), Wang et al. (2012), and Zuidema et al. (2012). Further background on available temperature profiling capabilities can be found in Haggerty et al. (2014), while German capabilities in free-tropospheric thermodynamic profiling are detailed in Mech et al. (2014). An example of NASA high-altitude thermodynamic profiling integrated with active microwave sensing and in situ measurements to understand hurricane evolution is given in Braun et al. (2013). A previous NSF workshop report emphasizing the value of humidity profiling can be found at www.eol.ucar.edu/system/files/LAOF_Workshop_Report_FINAL_06112013.pdf.
We thank the NSF Lower Atmospheric Observing Facilities Division and NCAR Earth Observing Laboratory for supporting the workshop on airborne microwave radiometry. Three anonymous reviewers and Ian McAdams of the Naval Research Laboratory provided comments that clarified and improved the final document.