Remote Sensing of Supercooled Liquid Water to Enhance Safe Operational Environments for Crewed and Uncrewed Aircraft Systems

This research explored the novel application of uncrewed aircraft systems (UASs) for atmospheric observations in cold season clouds characterized by supercooled liquid water (SLW). The primary aim was to enhance our understanding of cloud microphysical properties and flight dynamics in and around these challenging environments. SLW plays a crucial role in cloud formation, precipitation processes, and weather phenomena, making its accurate measurement essential for both meteorological forecasting and climate modeling. Traditional observational methods, such as balloon-borne instruments and ground-based radar measurements, offer valuable data but are often limited in spatial resolution and real-time dynamics. To address these limitations, our study leverages the UAS’s capability to collect high-resolution, spatially distributed data within these complex environments.

The fixed wing UAS, equipped with instrumentation to measure variables of state, was flown under clouds containing SLW. Key flight dynamics parameters, including altitude, velocity, thrust, and orientation, were recorded alongside atmospheric variables such as temperature, and humidity.

These UAS-derived data were compared to measurements obtained from two remote sensing methods - supercooled liquid water content (SLWC, in units of gm-3) sondes on weather balloons and output from a machine learning algorithm that utilizes dual-polarization S-band radar fields for SLW detection. The SLW-sonde sensor operated by measuring the buildup of ice on a thin vibrating wire. Accrued ice changes the vibration frequency of the wire and the change of vibration frequency can be used to quantitatively determine the supercooled liquid water content (SLWC, in units of gm-3) of the cloud layers. The SLW-sonde provided quasi-vertical profiles of the cloud's SLWC and variables of state such as temperature, pressure, and relative humidity. The second SLW remote sensing method utilized fuzzy logic machine learning to categorically identify the presence and absence of small and large drop SLW within a radar volume.

By integrating and comparing these datasets, the study addresses several key research questions: How do UAS measurements of flight dynamics correlate with those obtained from these remote sensing platforms? What are the advantages and limitations of UAVs in capturing real-time variations in cloud properties compared to balloon-borne and ground-based methods? How can UAS data be used to refine our understanding of cloud microphysics and improve atmospheric models?

Preliminary results from two recent case studies suggest that UASs provide a valuable complement to traditional observational tools by offering more dynamic and spatially resolved insights into cloud conditions. The comparison of UAS data with balloon-borne and vibrating wire sonde measurements enable us to validate and calibrate UAS instruments, enhancing their accuracy and reliability. Additionally, this research contributes to the development of more sophisticated atmospheric observation techniques, potentially improving weather forecasting and climate prediction capabilities, and enables precise targeting of SLW for cloud seeding.

The implications of this study are significant for both meteorological research and practical applications. By bridging the gap between traditional observational methods and cutting-edge UAS technology, this study paves the way for more precise and comprehensive atmospheric measurements, ultimately advancing our knowledge of cloud processes and their impacts on weather and climate. Furthermore, the integration of UAS technology into atmospheric research applications represents a promising advancement in observational capabilities, offering a new perspective on the dynamic and complex nature of cloud environments and increasing the volume of available data. This data source, once more fully developed, will be cheaper than crewed research aircraft per flight hour by roughly an order of magnitude.