The dynamic distribution and change of terrestrial water, manifested in wetlands that support a wide range of vegetation types and ecosystems, and also in floodplains that are prone to inundations, are very important to the understanding of our changing climate and our ability to mitigate risk. At present all of the existing measurement techniques have serious limitations in the ability to observe terrestrial water bodies globally and at the spatial and temporal scales required to fully capture their dynamics. In the last few years a number of studies in the community of Global Navigation Satellite Systems (GNSS) Reflectometry have been focusing on reflections over wetlands and inundated areas [Zuffada et al, 2016; Nghiem et al, 2017; Chew et al., 2018; Morris et al., 2019], particularly since data analysis of the CYclone GNSS (CYGNSS) mission began showing the ability to resolve small-scale land features such as rivers and bodies of water even partially obstructed by vegetation. CYGNSS is a constellation of 8 microsatellites, that provides significantly increased temporal sampling and revisit rate in the tropical latitudinal band as compared to traditional monolithic instruments, thus enabling a new observing strategy for capturing dynamic water events. In [Zuffada et al., 2017; Nghiem et al., 2017], based on reflected signal characteristics such as peaked (limited spread in delay and doppler) and symmetric shape, and very high reflected peak power, it was hypothesized that over wetlands there are strong coherent specular reflections in the collection area of the signal, originating from (even small) areas of standing water, resulting in the measurements’ magnified sensitivity to water because of its high electric permittivity compared to dry land and/or vegetation. Plots of peak power, corresponding to CYGNSS measurements’ specular points, aggregated over a period of time, and displayed over large regions with complex hydrology such as the Amazon basin clearly showed the potential of CYGNSS to map surface hydrology of intricate scenes at the continental level. At the regional scale, availability of in-situ and other correlative data have led to introduce thresholds in peak power values that differentiate between two binary states, i.e. dry and wet associated with inundations [Chew et al., 2018; Morris et al., 2019]. The limitations of the assumptions used to map wetlands have been analyzed in [Loria et al., 2019 (in preparation)] and shown to be traceable to the complex nature of the scattering from inhomogeneous scenes where the local water topology, surface topography and local meteorology can affect the mix of coherent and incoherent scattering, thus producing highly variable peak power that confounds the measurements. This communication summarizes our best understanding of the retrieval accuracy of GNSS reflectometry for dynamic monitoring of wetlands and inundations. It is based on analysis of actual CYGNSS constellation data acquired over terrestrial water bodiesand discusses the limitations in estimating surface water globally when inhomogeneities at the small scales are at play. Comparisons and cross-validations have been performed with measurements from ALOS-2 and Sentinel-1 data. References Chew, C., Reager, J.T. and Small, E., 2018. CYGNSS data map flood inundation during the 2017 Atlantic hurricane season. Scientific Reports (Nature Publisher Group), 8, pp.1-8. Loria, E. et al. Analysis of Wetland Extent Retrieval Accuracy Using CYGNSS Data. In preparation, 2019. Morris, M., C. Chew, J.T. Reager, R. Shah, C. Zuffada. A Novel Approach to Monitoring Wetland Dynamics using CYGNSS: Everglades Case Study. Remote Sensing of Environment, in press. 2019. Nghiem SV, Zuffada C, Shah R, Chew C, Lowe ST, Mannucci AJ, Cardellach E, Brakenridge GR, Geller G, Rosenqvist A. Wetland monitoring with Global Navigation Satellite System reflectometry. Earth and Space Science. 2017 Jan 1;4(1):16-39. Zuffada, C., Chew, C., Nghiem, S.V., Shah, R., Podest, E., Bloom, A.A., Koning, A., Small, E., Schimel, D., Reager, J.T. and Mannucci, A., 2016, August. Advancing Wetlands Mapping and Monitoring with GNSS Reflectometry. In Living Planet Symposium (Vol. 740, p. 83). Zuffada, C., Chew, C. and Nghiem, S.V., 2017. GNSS-R algorithms for wetlands observations. In IGARSS 2017-2017 IEEE International Geoscience and Remote Sensing Symposium (pp. 1126 – 1129), Fort Worth, TX.

Wetlands represent an essential ecosystem, providing flood control, carbon storage, and supporting biodiversity. In particular, the Everglades is a Ramsar Wetland of International Importance, supporting several threatened and endangered species of flora and fauna, and is especially important for wintering birds. Understanding and monitoring wetlands like the Everglades requires the ability to accurately identify and measure wetland extent and change in extent on short time scales. However, in situ methods are difficult given the nature of the surrounding environment, and optical methods of remote sensing are unable to see through dense vegetation. NASA’s Cyclone Global Navigation Satellite System (CYGNSS) has shown promising results using GNSS Reflectometry to identify the presence and extent of inland water. Utilizing GNSS as a signal of opportunity in an L-band passive bistatic radar, it can penetrate rain, clouds, and vegetation. Its 8-satellite constellation exhibits daily or sub-daily revisit rates, enabling the observation of dynamic changes on short time scales. In this work, we utilize a combination of CYGNSS data, ancillary information, and simulations to understand the observability of inundation beneath vegetation. Simulations were used to predict the received power using a water mask derived from Landsat imagery over the Everglades. By analyzing the differences between expected and actual received power, we identified areas of flooded vegetation. These differences were then combined with ancillary data sets to measure seasonal changes and create a seasonal map of open water and inundated vegetation throughout the Everglades network. We also investigated the ability of CYGNSS to discern and measure different vegetation types. Results were then compared with optical and radar imagery and verified with truth data from the Everglades Depth Estimation Network (EDEN) and littoral vegetation maps from the South Florida Water Management District. By leveraging CYGNSS’s high temporal frequency of observations and ability to see under vegetation, measurements of inundated vegetation and its change can complement other remote sensing and in situ methods of wetland monitoring.