GNSS-Reflectometry: Fundamentals, Methods and Applications

The Global Navigation Satellite System (GNSS) includes the American Global Positioning System (GPS), the Chinese BeiDou Navigation Satellite System (BDS), the Russian GLONASS system, and the European Galileo Satellite Navigation System, as well as regional augmentation systems such as the Japanese Quasi-Zenith Satellite System (QZSS) and the Indian Regional Navigation Satellite (IRNSS). GNSS features all-weather, all-time, real-time, and high-precision capabilities with continuously transmitting L-band signals, and has been widely used in positioning, navigation, and timing (PNT). With the improvement of satellite navigation systems and the increase in constellations and observation stations, its application prospects are becoming increasingly broad. Today, GNSS has been widely used in transportation, power telecommunications, public safety, precision agriculture, basic surveying and mapping, resource surveys, Earth science, space science, and some new fields that are yet to be developed.

The GNSS signal is a modulated wave broadcasted by GNSS satellites to users for navigation, positioning and timing. There are three reasons for using modulated waves: (1) The ranging code and navigation message used for user positioning are digital signals, which cannot be directly transmitted in wireless channels; (2) It can effectively shift the spectrum and make rational use of spectrum resources; (3) Modulating low-frequency navigation information into high-frequency signals is beneficial for signal propagation, which improves anti-interference ability, and reduces the size of the receiver antenna. Since the GNSS signal is a modulated wave, it should include a low-frequency data wave and a high-frequency signal carrier wave. The low-frequency data wave includes a pseudorandom ranging code and a navigation message.

GNSS reflection measurements are divided into ground-based and space-based reflectometry depending on the location of the receiver. For ground-based reflection measurements, if we want to fully receive the reflected signal, we need to use a specially developed receiver that conforms to the polarization of the reflected signal. In addition, the receiver antenna should face the horizontal direction to receive the reflected signal from the ground. However, most of the GNSS monitoring networks or control networks currently use geodetic receivers, with the receiver antenna facing the zenith and mainly receiving right-hand circularly polarized signals. Although the geodetic receiver suppresses the acquisition of reflected signals, reflected signals can still be received at low satellite elevation angles and will interfere with the direct signals from the satellite, affecting GNSS measurements and causing the so-called “multipath error”. This interference delay effect caused by the propagation of signals through multiple paths is called multipath effect. For traditional GNSS receivers, the reflected signal and the direct signal are coupled together with causing the multipath effect, then the surface characteristic information carried by the reflected signal will be reflected in the multipath, so we can use the extracted GNSS multipath to invert the surface characteristics. For example, Fig. 4.1 shows the reflection process of GPS L1 and L2 signals’ RHCP and LHCP on a generally selected surface.

The Interference Pattern Technique (IPT) is a new GNSS-R technology, which mainly utilizes the interference of GNSS direct signals and reflected signals. This technique requires a specially made interference-type receiver, which utilizes the polarization characteristics of the reflected signal. In addition, based on the Z-V model and MSS model, the delay-Doppler map (DDM), and the inversion of the scattering coefficient σ⁰, surface roughness parameters can be obtained, such as sea surface roughness (the higher the wind speed is, the greater the surface roughness is, the larger the DDM in the delay and frequency domain is), ocean wind field, sea surface temperature, ocean salinity, hurricane detection, sea surface oil spill, etc. This chapter introduces the interference pattern technique and delay Doppler map as well as their applications, starting from the Fresnel reflection coefficient.

Over 70% of the Earth’s surface is covered by oceans, and the majority of the world’s population resides in coastal areas. Therefore, changes in sea level significantly impact the stable development of human society. Ocean altimetry helps us effectively grasp information about global sea level changes, polar sea ice melting, thermal expansion of seawater, and changes in the atmospheric ocean circulation system. These changes could potentially trigger extreme weather events, such as hurricanes and tsunamis, posing threats and obstacles to normal human life and socio-economic development. Major disasters, such as the tsunami in Indonesia and Hurricane Katrina, have resulted in significant casualties. Therefore, to predict and prevent ocean disasters in advance, it is necessary to establish a comprehensive early warning system, which requires data on global sea level changes and instantaneous sea conditions. In addition, the water resources of the ocean can affect the exchange of water vapor energy, thereby affecting atmospheric activities. Therefore, even minor changes in sea conditions can cause weather changes in some areas. To meet the needs of ocean and climate observations, it is essential to monitor sea level change.

GNSS-R technology uses the GNSS reflected signals for remote sensing of the Earth. It uses delay Doppler mapping equipment (DDMI) carried on different platforms to receive the reflected signals from the Earth’s surface and the receiver’s local replica code, and measured the scattering power of the Earth’s surface as well as inverted the geophysical parameters of interest. Recently, there have also been experimental studies that directly coherently processing the reflected signals and direct signals to augment the radar signals with improving ranging accuracy and spatial resolution. Because GNSS-R uses the working mode of bistatic radar, it directly uses the GNSS signals that are ubiquitous in space. The deployment of the GNSS-R observation system has a very high cost-effectiveness ratio, and especially the space-based GNSS-R can achieve high temporal and spatial resolution remote sensing of the Earth globally. Like traditional remote sensing radar technology, GNSS-R can be used for ocean altimetry, sea ice detection, sea surface wind speed, sea surface roughness, soil moisture estimation on the Earth’s surface, etc., and therefore has a very broad application prospect.

The BeiDou Satellite Navigation System (BDS) has been developing since 1999. Currently, more than 35 BDS satellites are in orbit, including Geostationary Orbit satellites (GEO), Inclined Geosynchronous Orbit satellites (IGSO), and Medium Earth Orbit satellites (MEO). BDS can provide global and regional positioning, navigation, and timing (PNT). In this section, we used an IGS multi-GNSS station MAYG to estimate sea level changes. This station is near the Indian Ocean’s Mayotte. The MAYG station is equipped with a TRIMBLE NETR9 geodetic receiver and a TRM59800.00 antenna. (Latitude: −12.78°, Longitude: 45.26°, Height: −16.35 m). As MAYG is a multi-system GNSS experimental station, it can not only receive GPS and GLONASS signals but also receive tri-frequency BDS signals (L2, L6, L7). We used five available BDS satellites (PRN6-10) to estimate the sea level changes and its local tides from January 2015 to June 2015. In suit sea level data collected by radar sensors at a sampling rate of 1 min were provided by the Intergovernmental Oceanographic Commission (IOC).

The see wind field plays a crucial role in weather and climate research, sea-air interaction studies, and maritime shipping safety. Traditional see wind speed observation techniques struggle to provide high spatiotemporal resolution wind speed data to meet scientific and commercial applications across the vast ocean surface. Marine buoys offer high-precision in-situ wind speed measurements, while their deployment is costly, and most are deployed near coastal regions, leading to significant spatial unevenness in global sea wind field observations. Space-borne microwave scatterometers can remotely sense sea surface wind speeds at specific working frequency bands, but they are susceptible to cloud and precipitation interference. Due to the high deployment costs of these dedicated satellites, denser ocean satellite constellations are challenging and also the timeliness of wind speed observations is constrained by the revisit cycle of individual satellites. The space-borne GNSS-R remote sensing technology acquires GNSS signals reflected from the ocean surface using a GNSS-R receiver mounted on a LEO small satellite, and provides a bistatic radar scatterometer for sea surface wind speed retrieval. GNSS-R equipment is lightweight and energy-efficient, significantly reducing deployment costs of this remote sensing technology. Through optimized satellite networking, it can achieve continuous and rapid observation of global sea surface wind speeds, effectively addressing the low spatiotemporal resolution limitations of traditional space-borne monostatic scatterometers and radiometers.

Soil moisture is an important factor affecting global climate and environment. Its spatial flow in the Earth system are crucial in global energy balance. Meanwhile, soil moisture is a key parameter controlling the exchange of water and thermal energy between land and atmosphere, directly related to atmospheric humidity, and affects the heat flux transfer from land to atmosphere. Atmospheric circulation modelling parameterized by surface parameters shows a strong feedback relationship between soil moisture and abnormal climate. Soil moisture is also an important indicator parameter in the fields of hydrology, meteorology, and agricultural science. Large-scale soil moisture monitoring and understanding are important components of agricultural study and ecological environment evaluation. At the same time, soil moisture is also the link between surface water and groundwater and an important part of the water cycle in terrestrial ecosystems. Therefore, soil moisture plays an important role in improving regional and even predicting global climate models, studying global water cycle, managing water resources, establishing watershed hydrological models, monitoring crop growth, estimating crop yield, monitoring hydrological disasters, and other related natural and ecological environment issues. Traditional monitoring of soil moisture is conducted using discrete sites or corresponding meteorological stations, but the results of this method can only represent a limited observation area and the observation process is time-consuming and laborious, unable to meet the needs of large-scale and high-efficiency soil moisture observation. At the same time, the traditional monitoring method cannot match the corresponding weather and hydrological models (0.1–10 km) in terms of spatial scale and time accuracy, so this method cannot effectively study the impact of soil moisture on environmental changes.

The strong attenuation and scattering of vegetation make the inversion of soil moisture difficult. However, the vegetation biomass plays an important role in carbon cycling and greenhouse effect monitoring. Compared with traditional active and passive remote sensing technologies, such as SAR and radiometric measurement technology, GNSS-R technology has unique advantages such as small size, light weight, low energy consumption, and high spatiotemporal resolution, which provides a new technology for environmental remote sensing. Currently, a number of institutions and researchers are conducting studies on GNSS-R vegetation remote sensing using qualitative analysis methods. Rodriguez-Alvarez et al. [1, 2] have been using ground-based SMIGOL-reflectometer (Soil Moisture Interference Pattern GNSS Observations at L-band) for geophysical parameter inversion. Interference Pattern Technique (IPT) was developed to measure the changes in direct signals caused by multipath effects generated by reflected interference signals. The minimum amplitude is called a notch, and its number and position are functions of soil moisture and vegetation height [1, 2]. Small et al. [3] have been using GNSS multipath to qualitatively estimate the growth status of vegetation from the Plate Boundary Observatory (PBO) GNSS network and found that NDVI (Normalized Vegetation Index) was negatively correlated with multipath effect amplitude. Since the fewer GNSS observation stations are set up in forest areas, this method is only applicable to farmland, grassland, and shrubland.

Dry snow remains in sub-zero environments for a long time and does not melt. Dry snow has a low density, so the L-band microwave signal can penetrate hundreds of meters of snow depth. Dry snow areas occur in Greenland and the inland areas of the Antarctic ice sheet, where the snow thickness can reach several kilometers. The L-band signal can penetrate the accumulated snow layer, corresponding to the snow accumulated over the past thousands of years [1–3]. Understanding the climate of these areas is very important for studying the possible impacts of climate change on glaciers melting and dynamics. Meanwhile, studying and prediction of the Antarctic glaciers continuous changes require better snow accumulation distribution maps, and snow accumulation distribution maps are also important for monitoring sea level changes [4, 5]. A feasible future space-based GNSS-R mission can achieve dense sampling in these areas [6].

Seasonal and permanent frozen soils comprise about 35% of the total land area of the Earth, predominantly distributed in high-latitude and high-altitude regions. These soils undergo seasonal changes in freeze-thaw states with significantly impacting the environments where human live. The freeze-thaw cycles influence energy exchanges between solid Earth and its atmosphere and affect surface runoff and carbon cycling processes. As water transits between solid and liquid phases in the soil, it significantly affects surface radiation energy conversion, evapotranspiration and surface runoff intensity. This process plays a crucial role in determining both surface energy balance and water cycle dynamics, making freeze-thaw states sensitive indicators for climate change. To effectively monitor these changes, it is essential to observe the spatio-temporal distribution of soil freeze-thaw conditions and related physical parameters. Traditional remote sensing methods using visible light and thermal infrared are often hindered by adverse weather conditions. In contrast, microwave remote sensing techniques —both active (radar) and passive (radiometer)—provide reliable monitoring regardless of time or weather. Satellite data enhance the spatial resolution of soil moisture and surface freeze-thaw monitoring. However, its temporal resolution (global repeat coverage every 3 days) may not fully meet the scientific requirements for freeze-thaw monitoring.

In recent years, a number of ocean remote sensing experiments have been conducted with achieving numerous progress. In October 2000, the Hurricane Hunter aircraft of the National Oceanic and Atmospheric Administration (NOAA) in the United States with carrying GNSS-R receiver flew into Hurricane Michael from the coast of South Carolina. By analyzing the GNSS signals reflected from the sea surface during the tropical cyclone, wind speed results were obtained [1]. The UK-DMC satellite of the United Kingdom successfully inferred the Earth’s surface physical coefficients such as sea surface roughness using its onboard GNSS-R equipment [2]. GNSS reflection signals from calm sea areas can also provide highly accurate altimetric results [3]. Currently, sea surface monitoring using GNSS reflected signals from the sea surface is a hot topic [4, 5]. For instance, sea surface roughness and ocean wind were monitored using GNSS signals. However, detailed analysis such as electromagnetic field scattering theory, energy and Doppler coefficient recovery methods [6], and the characteristics of L-band slope probability density functions still require further study.