Modeling and Analysis of GNSS-R Measurements in Wetland Areas
Abstract
The distribution and dynamics of surface water plays an important role in understanding our changing climate. The current suite of measurement techniques to observe the global surface water distribution, in particular wetland areas, are inadequate at the spatial and temporal scales required to fully capture the dynamics. Recently, measurements of the reflections from Global Navigation Satellite System (GNSS) signals have demonstrated high sensitivity to inland waters through rain, clouds, and vegetation. These measurements made by GNSS reflectometry (GNSS-R) receivers from spaceborne platforms have the potential to provide drastic increases in the temporal resolution of inundation measurements as compared to traditional monostatic radar systems. The current state of the art GNSS-R system is NASA mission CYGNSS, which consists of 8 small satellites. Over inland waters, coherent specular reflections are often measured by CYGNSS. Coherent reflections typically have higher power and increased spatial resolution in comparison to reflections from rough surfaces like the ocean. In this work, we utilize a scattering model for coherent reflections in heterogeneous wetlands scenes. The scattering model is extended to include the effects of the local wind field and the vegetation on the received signal. Complex interactions between wind, water, and vegetation may lead to large changes in the total reflected signal, potentially obfuscating retrievals of inundation extent. These interactions include the surface roughness due to wind-generated waves and signal attenuation due to vegetation. Partially submerged vegetation will also act to block wind and dampen waves, resulting in a smooth water surface and strong reflected signal. In this work, we model these effects to a first-order and incorporate them into a unified scattering model for arbitrary wetlands scenes. The unified model is then used to examine the effects of wind and vegetation on the received coherent signal in order to examine the accuracy of surface inundation extent and vegetation type/density retrievals. We first use a simulation study to examine these effects. Then, data from CYGNSS is used to validate the unified model using the measured reflections from target lakes.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2020
- Bibcode:
- 2020AGUFMIN019..15L
- Keywords:
-
- 1964 Real-time and responsive information delivery;
- INFORMATICS;
- 1972 Sensor web;
- INFORMATICS;
- 4303 Hydrological;
- NATURAL HAZARDS;
- 4337 Remote sensing and disasters;
- NATURAL HAZARDS