Estimating Surface Soil Moisture Using L-Band Reflectivity Measurements From The CYGNSS Microsat Constellation

Contemporary investigations of the Earth’s land surfaces are often facilitated by detailed measurements or measurement-informed models of surface soil moisture (SSM) content. SSM is a key component in models of heat exchange between the Earth's surface and the Atmospheric Boundary Layer. SSM modulates atmospheric conditions through its role in dividing incoming radiation into latent and sensible heat flux, which determine surface/atmospheric temperature and evapotranspiration. SSM data thus provides critical information to civilian and DoD weather modelers and forecasters.

The use of passive microwave radiometers (e.g. Space Force’s WSF-M), presently remains the state-of-the-art in satellite-based SSM remote sensing, though this strategy is not without shortcomings. Dedicated SSM-sensing radiometer instruments are typically mounted on polar-orbiting spacecraft with coarse revisit times on the order of three days near the equator. The placement of eight (seven still operational) low-cost Global Navigation Satellite System Reflectometry (GNSS-R) receivers in low-Earth orbit, as part of NASA’s 2016 Cyclone GNSS (CYGNSS) mission, provides the opportunity to explore the potential utility of using these emergent systems for the estimation of SSM as a means of achieving comparatively shorter revisit times.

There are currently several algorithms utilizing the CYGNSS constellation to retrieve land surface properties including SSM. These algorithms can generally be sorted into one of a few categories: 1) empirical or curve-fitting, 2) forward-modeling, 3) machine learning and neural networks, and 4) change detection. The algorithm described herein is part of the lattermost category.

This work makes use of a dense temporal time series of small-satellite measurements using GNSS-R to augment the use of satellite radiometers for SSM remote sensing. The technique is based on ratios between consecutive angle-normalized reflectivity measurements. The strength of the retrieval algorithm considered here lies in its lack of requirement for: 1) ancillary data on vegetation or surface roughness, 2) training data, and 3) forward modeling parameters.

Error statistics from the SSM retrieval algorithm will be reported for a variety of validation sites, for which analysis will include examination of both trends and absolute retrieved SSM values. Conclusions will be presented based on these quantitative analyses, and future work discussed.

PRESENTER

Ouellette, Jeffrey
jeffrey.d.ouellette4.civ@us.navy.mil
202-767-2526

US Naval Research Laboratory

CATEGORY

Climate, Weather, Ocean Modeling

SYSTEMS USED

Narwhal

SECRET

No