**The ESA-ANISAP Study: Estimate of Tropospheric Scintillation Along LEO-LEO Links Through High Resolution Radiosonde Data**
*
Martini, Enrica ^{1}; Facheris, Luca^{2}; Freni, Angelo^{2}; Cuccoli, Fabrizio^{3}
^{1}Department of Information Engineering and Mathematical Sciences,University of Siena, ITALY; ^{2}Department of Information Engineering, University of Florence, ITALY; ^{3}CNIT-National RaSS Laboratory, ITALY
*

Due to turbulent induced random variations of the refractive index, a radiofrequency signal propagating in the troposphere undergoes rapid amplitude and phase fluctuations, known as scintillation. This kind of disturbance can significantly influence the performances of atmosphere sounding techniques based on the use of radiowaves propagating in a limb geometry, as it was pointed out by the ACE+ mission studies carried out within the framework of Earth Explorer Opportunity Missions supported by ESA. In this context, a correct modelling of scintillation effects plays a key role in the design of a successful radio link. A theoretical model for the analysis of turbulence-induced scintillations on LEO - LEO radio occultation links was developed in the framework of the ESA-AlMetLEO study, aiming at investigating the potential of the Normalized Differential Spectral Attenuation (NDSA) for the estimate of the total content of water vapor along the propagation path between two LEO satellites. The NDSA technique, proposed in 2002 by two of the authors, is based on the use of the normalized incremental ratio of the spectral attenuation and can provide a certain robustness against scintillation disturbances. The scintillation model was further developed with the inclusion of absorption effects in the successive study ESA-ACTLIMB. However, a serious limitation to the use of this model in the simulation of NDSA measurements was the fact that the scintillation parameters (in particular, the structure constant profiles) were not related to the local atmospheric turbulence encountered by the microwave link. To overcome this limitation, one of the objectives of the on-going ESA-ANISAP study is the development of a procedure enabling the calculation of the turbulence parameters from meteorological data. A second objective is the inclusion of such parameters in a parametric scintillation model of a radio link between two LEO satellites both for the counter-rotating case and for the co-rotating case. This work summarizes the developed procedure, which allows one to directly estimate the scintillation disturbance for a given LEO-LEO link configuration starting from high resolution radiosonde data. The refractive-index structure constant profiles are derived from high resolution radiosonde data by using the vertical gradient approach, which can be applied at all the altitudes of interest for the NDSA approach, as opposed to the structure function approach that is only suitable for the analysis of turbulence in the lower troposphere. The fact that turbulence in the free atmosphere is confined to vertically thin layers is accounted for by identifying the turbulent layers through the analysis of the Richardson number profiles. Then, the derived structure constant profiles are inserted into a parametric scintillation model to create a consistent scintillation disturbance estimate, to be used in NDSA simulations. To this end, the troposphere is described as a spherically symmetric turbulent medium. Rytov's first iteration solution for weak fluctuations is used to derive an expression for the variance of amplitude fluctuations of the wave as well as the correlation between the fluctuations at different frequencies. Indeed, these two quantities are the ones of major interest to assess the performances of the NDSA approach in the presence of scintillation, since only high correlation values between the fluctuations of signals at different frequencies allow NDSA measurements to outperform standard single-frequency approaches at all altitudes in the troposphere. The developed model represents a generalization of the one developed in previous studies mentioned in a companion paper, the two main upgrades being the possibility to handle an arbitrary layered structure constant profile and the inclusion of anisotropy. Anisotropic irregularities, which can contribute to amplitude scintillation in the higher troposphere, have been modelled by introducing an ellipsoidal spectrum model as a generalization of the conventional von Karman spectrum. An appealing feature of the resulting scintillation model is its flexibility, since it allows one to individually select the outer scale length and the anisotropy factor for each layer. The developed procedure has been tested on a large set of radiosonde data corresponding to different latitudes, seasons and times and results are presented in this work.