**1. Introduction**

It is well known that Global Navigation Satellite Systems signals (which include for example the U.S. GPS and its modernization, the Russian GLONASS, the future European Galileo, the Chinese COMPASS), commonly processed for navigation purposes, can also be used to characterize media where they propagate in. In the last decade, GNSS atmospheric and Earth's surface remote sensing become more and more important, thanks to technical improvements applied to the processing of such "free-of-charge", everywhere available and weather insensitive signals.

For example, remote sensing of wet part of troposphere is possible "extracting" the atmospheric delays from GNSS observations. These delays are associated to water vapour and are accumulated by the signal along its propagation path. In the double difference phase observation adjustment (a standard GNSS signal pre-processing) it is possible and quite easy to estimate the wet contribution to atmospheric total delay mapped into the zenith direction, the so-called Zenith Wet Delay. From one side the estimate of propagation delays is essential to improve the accuracy of the height determination in the geodetic positioning framework (Kleijer, 2004). From the remote sensing point of view, Zenith Wet Delay may be then transformed into the so-called Integrated Precipitable Water Vapour (IPWV). Therefore, the knowledge of the temporal behaviour of IPWV above a GPS receiver network allows meteorologists to know the evolution of total water vapour content in atmosphere, which is one of the variable operatively used in Numerical Weather Prediction Models. These aspects are described in section 2.

A second important application allows to add vertical variability information to the atmospheric parameter distribution with respect to the previous one, which represents an "integrated" quantity. The amplitude and phase variations experienced by GNSS signal crossing the atmospheric "limb" and received on-board a Low Earth Orbit satellite, can be used to infer temperature and water vapor profiles, thanks to the GNSS Radio Occultation technique (Melbourne et al., 1994; Ware et al., 1996; Kursinski et al., 1997; Hajj, 2002). Even if aspects related to such very important Remote Sensing technique are not treated in the present chapter (a comprehensive tutorial can be found in Liou et al (2010), while review of results

GNSS Signals: A Powerful Source for Atmosphere and Earth's Surface Monitoring 173

Interferometric Synthetic Aperture Radar (InSAR). Several techniques are well established to derive the vertically Integrated Precipitable Water Vapor (IPWV)1, in particular using ground-based and spaced-based radiometers, radiosonde observations and GNSS receivers. Radiosonde observations produce an accurate measurement of the water vapour profile, but the temporal and spatial resolution is rather poor. Radiosondes are typically launched every 6 to 12 hours, which may cause significant variations in water vapour to go undetected.

Ground-based microwave radiometers show problems during periods of rain fall and spacebased radiometer observations can be degraded in the presence of clouds. This prevents reliable measurements during periods where changes in water vapour could be quite great.

The technique to estimate IPWV by means of GNSS receivers is based on measurements of the tropospheric delay time of navigation signals. Therefore the delay, regarded as a nuisance parameter by geodesists, can be directly related to the amount of water vapour in the atmosphere, and hence is a product of considerable value for meteorologists. Furthermore, water vapour estimation with ground-based GNSS receivers is not affected by

So, GNSS is a valuable complement to radiosondes and radiometers, taking into account that GNSS IPWV estimates come from an existing GNSS infrastructure and frequently from

The use of GNSS receivers to estimate IPWV is based on measurements of the delay affecting the navigation signals during their propagation in troposphere (neutral atmosphere) from the GNSS satellites to the receivers on ground. The dispersive ionospheric effect can be removed with a good level of accuracy by a linear combination of dual

Such a technique is founded on the non-dispersive refractive characteristics of the neutral atmosphere, governed by its composition. The water vapour molecules in atmosphere are polar in nature possessing a permanent dipole moment. All the other gases are non-polar molecules and a dipole moment is induced among these gases when microwave propagates through atmosphere. These molecules reorient themselves according to the polarity of propagating wave. In the retrieval technique to be described the atmosphere is considered

Consequently, the neutral delay due to the troposphere can be decomposed into the hydrostatic delay associated with the induced dipole moment of the atmosphere constituents and the wet delay associated with the permanent dipole moment of water vapour (Askne & Nordius, 1987; Brunner & Welsch, 1993; Treuhaft & Lanyi, 1987). The zenith hydrostatic delay (ZHD) has a typical magnitude of about 2.4 m at sea level, and it grows with increasing zenith angle reaching about 9.3 m for elevation angle of 15°. With

1 Consider the total amount of atmospheric water vapour contained in a vertical column of unit cross section: if this water vapour were to condensate and precipitate, the equivalent height of the liquid water within the column is the Integrated Precipitable Water Vapour, usually measured in cm or in g/cm2

Besides these limitations, all systems involve considerable costs.

quite dense receiver networks.

frequency data.

rain fall and clouds, and can therefore be considered an all-weather system.

**2.1 Description of observables, theoretical basis and retrieval technique** 

as the sum of a dry component (mainly due to O2) and a wet component.

obtainable can be found in Anthes et al. (2008) and Luntama et al. (2008)) a mention is due. When signals cross in this way the atmosphere, they are delayed and their path is bent: therefore, the signal can be received also below the terrestrial limb, when the satellites are not yet in view. GNSS Radio Occultation is based on the inversion of the excess-phase (carrier phase in excess with respect the one experienced considering vacuum propagation) and amplitude evolution measured on the received signal when it is "occulted" with respect to the transmitter. Applying Geometric Optics algorithms or Wave Optics algorithm and Fourier operators to such observables, time evolutions of two important parameters identifying each trajectory followed by the signal can be derived: its total bending and its impact parameter, which is the distance of the trajectory asymptotes from the Earth's mass centre. Such quantities are in turn related to the integral of the atmospheric refraction index vertical profile, in a mathematical formulation that is invertible in a closed form. Result of the inversion is a veryaccurate and high-resolved (up to about 100 m) atmospheric refractivity vertical profile, from which the corresponding temperature and humidity profiles can be inferred.

The second technique described in this chapter adds a further spatial variability characterization possibility with respect to that given by IPWV and Radio Occultation. It deals with the three-dimensional reconstruction of atmospheric refractivity and, thus, water vapour density, applying tomographic techniques to phase delays measurements collected by small (but dense) networks of GPS receivers. Because of volume dimensions, inhomogeneity spatial distribution and geometric constraints, all the weak points of tomography emerge in characterizing neutral atmospheric parameter distributions using GNSS signals. Results and comments are given in section 3.

The last application we will describe (section 4) is the most recent and maybe the most challenging one. It foresees the use of GNSS signals reflected off from lands and oceans for characterizing the Earth's surface at L-band frequencies. The signal is received under bistatic geometry since the received signal power is that which is forward scattered from the Earth's surface towards the GNSS-R (GNSS-Reflectometry) receiver. The reflected signal contains many differences with respect to the direct one, in terms of delay, Doppler shift, power strength and polarization. Once the reflected signal is received, it is processed using hardware or software correlators. The reflecting surface features are dipped inside the shape, the magnitude and the maxima location (which is related to the propagation delay) of the obtained correlation function. Among the possible remote sensing applications we list: ocean altimetry (from delay); wind speed and ocean scatterometry (from shape and spreading), ice topography and monitoring (from delay and magnitude); soil moisture (from magnitude).
