Large‐scale modeling of rain fields from a rain cell deterministic model

Abstract

A methodology to simulate two-dimensional rain rate fields at large scale (1000 Â 1000 km 2 , the scale of a satellite telecommunication beam or a terrestrial fixed broadband wireless access network) is proposed. It relies on a rain rate field cellular decomposition. At small scale ($20 Â 20 km 2), the rain field is split up into its macroscopic components, the rain cells, described by the Hybrid Cell (HYCELL) cellular model. At midscale ($150 Â 150 km 2), the rain field results from the conglomeration of rain cells modeled by HYCELL. To account for the rain cell spatial distribution at midscale, the latter is modeled by a doubly aggregative isotropic random walk, the optimal parameterization of which is derived from radar observations at midscale. The extension of the simulation area from the midscale to the large scale (1000 Â 1000 km 2) requires the modeling of the weather frontal area. The latter is first modeled by a Gaussian field with anisotropic covariance function. The Gaussian field is then turned into a binary field, giving the large-scale locations over which it is raining. This transformation requires the definition of the rain occupation rate over large-scale areas. Its probability distribution is determined from observations by the French operational radar network ARAMIS. The coupling with the rain field modeling at midscale is immediate whenever the large-scale field is split up into midscale subareas. The rain field thus generated accounts for the local CDF at each point, defining a structure spatially correlated at small scale, midscale, and large scale. It is then suggested that this approach be used by system designers to evaluate diversity gain, terrestrial path attenuation, or slant path attenuation for different azimuth and elevation angle directions.

A large scale, high resolution channel model for propagation impairment techniques design and optimization

“…It extends the range of validity of the work presented in [1] and [4] by the use of outputs from a reanalysis database. This allows generating long-term rain attenuation time series correlated in space at the scale of the coverage of a SatCom system.…”

“…As the range of validity in time and in space of those simulators is limited to some hours and some hundredth of kilometers, they may not be appropriate to compute such outputs. The aim of this paper is to extend the range of validity of the stochastic rainfall field simulator described in [4] and adapted for rain attenuation in [1], using rain amount time series from general circulation models. In particular inputs from the low-resolution reanalysis database ERA-40 from the ECMWF (European Center for Mediumrange Weather Forecast) are considered.…”

“…However this approach is not sufficient as RRM (Radio Resource Management) algorithms need to be designed to manage FMT all over the coverage. For that purpose, realistic attenuation inputs correlated not only in space as in [1] but also in time at the scale of a spot-beam or of the whole satellite coverage [3] have to be provided to system simulators. As neither experimental nor meteorological data can be directly used for that purpose, spatio-temporal channel models have been developed ( [4], [5]).…”

A feedback suppression algorithm for reliable satellite multicast based on spatial–temporal prediction of the satellite channel

“…For the performance evaluation of the proposed protocol, a simulation model using Matlab has been developed. Similar propagation models have been implemented and proposed in [23][24][25]. Packets are sent to all receivers using the RFC-3453 forward error correction scheme.…”

Characterization and modelling of time diversity statistics for satellite communications from 12 to 50 GHz

“…The proposed model relies on the ergodicity property of rain fields at mid-scale [12][13][14], which assumes that the variability of rain fields over several tens of kilometres is similar in space and in time. Therefore, a similar model as the one proposed by Paraboni and Barbaliscia [7] for site diversity is considered to model time diversity.…”