Low-latency flood inundation mapping with airborne GNSS-R

The Rongowai project, based in New Zealand, represents a groundbreaking initiative in earth observation using next-generation Global Navigation Satellite Systems Reflectometry (GNSS-R) sensors. A NASA-developed sensor mounted on an Air New Zealand Q300 passenger aircraft collects land-surface and coastal data daily between airport hubs across the country. This project builds upon NASA's CYGNSS constellation, initially designed for sensing ocean surface winds but later expanded to terrestrial sensing due to the sensitivity of GNSS-R measurements to various surface properties of water. The next-generation GNSS-R receiver (NGRx) offers enhanced capabilities beyond CYGNSS, providing increased simultaneous measurements and introducing new measurement capabilities like polarimetry for improved land characterization. The unique mission model of Rongowai emphasizes sustainability while maintaining high-quality observations, utilizing an existing commercial Air New Zealand aircraft for data collection, thereby achieving unprecedented spatio-temporal sampling throughout New Zealand. The Air New Zealand Q300 operates approximately 7-8 flights daily in a hub-and-spoke pattern across major centers in New Zealand, offering near-ideal operational characteristics for capturing dynamic events. Here, we present a system that leverages the flight characteristics of the Q300 to deliver low-latency inundation observations immediately after landing, providing near real-time data transmission from the preceding flight. The framework, named the Flood Assessment Spatial Triage (FAST) addresses the challenge of data latency in flood reconnaissance by providing rapid inundation detection and visualization on an on-demand flight-by-flight basis within an hour after landing. The processing chain of FAST involves geolocation of specular points, coherence detection, and overlaying transects on a high-resolution digital elevation model (DEM) using a simplified flood inundation model. Analysis of GNSS-R waveforms demonstrates the ability to robustly observe inundation even in challenging conditions such as cloud cover, nighttime, and vegetated areas. Our study period captured flooding events in New Zealand's North Island during the Southern hemisphere summer of 2023, particularly in areas affected by Cyclone Gabrielle. The inundation observations from February 2023 depicted regions with surface water not classified as permanent water bodies, and a combination with a physically-based algorithm allowed for mapping flood inundation from the relatively sparse Rongowai observations. Our results align with ground reports of flooding, highlighting the potential for valuable reconnaissance information from GNSS-R when transiting affected regions. Rongowai's higher spatial resolution, combined with its hub-and-spoke flight pattern, enables rapid revisits over affected regions, making it well-suited for dynamic and rapidly evolving processes like floods.