Rising sea levels, originated from global climate change, lead to increasing number of inhabitants exposed to catastrophe damages. Hence, monitoring and observing sea level and its variation are of great significance, especially for the population living in the coastal regions. Recently, the ground-based Global Navigation Satellite System Reflectometry (GNSS-R) technique has been developed and applied to measure coastal sea level and lake level, to complement contemporary methods such as tide gauges and satellite radar altimetry. Compared with conventional techniques of tide gauge, this GNSS-R altimetry is capable of measuring absolute or geocentric sea level, or lake level height without land vertical motion contaminations. Additionally, it behaves much better in the coastal regions than traditional pulse-limited radar altimetry. As a result, the GNSS-R altimetry technique can potentially mitigate the temporal and spatial deficiency of historical and current sea level records.
A list of concisely stated study objectives is as follows:
The GNSS-R altimetry operates in a bistatic radar configuration, and its forward-scattering signal that is an electromagnetic (EM) wave is impacted by surface scattering properties, in addition to other error sources such as media delay. As one of the primary error sources, the EM bias resulted from the lesser reflectivity of sea wave crests rather than troughs, results in the underestimation of sea level height. To model the EM bias, a numerical simulation was initially conducted using linear and nonlinear wave models. The modeling results confirmed that GNSS-R altimetry measurement EM bias increases with decreasing incidence angle and increasing wind speed, with a constant
GNSS antenna height above the reflected sea surface. We used two realistic GNSS-R sea level measurements from two GPS sites located in the Gulf of Mexico for empirical EM bias modeling, which is a function of wind speed and the elevation angles along which GNSS reflected signals were collected by the GNSS antennae. We used the wind speed data to generate empirical GNSS-R EM bias models, which resulted in the improvement of the GNSS-R sea level accuracy. Also, the empirical EM bias models were shown to be more effective in improving GNSS-R sea level accuracy than the theoretical EM bias model. When the simulated and empirical models were applied to the original GNSS-R sea levels, it demonstrated that the RMS error decreased from ~7.3 cm to ~ 4.8 cm and ~3.4 cm, respectively.
To comprehensively assess the accuracy of the in situ GNSS-R sea levels, two adjacent geodetic-quality GPS sites 30 m apart at Robinson Point, and the closest tide gauge 13-km away at Tacoma, Washington, were selected for our validation study. The GNSS-R sea level time series has an 8-year sea level data span. The consistency between two adjacent GNSS-R sea level time series was significantly closer than the cases between either of the GNSS-R sea level and tide gauge time series. The root-mean-squares (RMS) errors between the adjacent 8-year GNSS-R altimetry time series were 4.6 cm and 1.1 cm, for hourly and weekly sampling, respectively, indicated excellent agreement and a robust error estimate. When the GPS-derived sea level was compared with tide gauge sea level 13-km away, the results illustrated that the GPS-derived hourly sea level time series has a consistency of 8.0 cm RMS, while the weekly smoothed sea level time series increased the consistency accuracy to 1.5 cm RMS. To further the detection, the tidal harmonic analysis was performed for the two adjacent GPS sites and the tide gauge. The results showed the largest differences occurred by the underestimation of amplitudes of large tidal constituents in GPS-derived sea levels and the phase difference in smaller tidal constituents.
As one of the most populated coastal regions of the world, the annual freezing and thawing of the Great Lakes and its influence on regional severe weather patterns significantly impact the local economy and ecosystems. Thus, accurate knowledge for the extent and thickness variations of lake ice in the Great Lakes, and their roles in the severity of winter storm and lake effect patterns are of importance to mitigate winter weathers for the people and the economies in the region. In this study, a novel method for lake ice thickness retrieval was proposed and developed, which was mainly based on a combination of a single geodetic quality GNSS receiver and a collocated water level lake gauge. For the first time, we estimated a 12-year lake ice thickness time series near the vicinity Harbor Beach on Lake Huron, and its reliability was validated by ice coverage product. The result initiates the opportunity for the potential use of this new measurement type of near shore lake ice thickness variations, in the assimilative modeling of Great Lakes Forecasting System, to improve the predictability of lake effects and severe winter storms.