The Communications Research Centre Canada (CRC) in Ottawa employed a multifrequency profiling radiometer (Radiometrics TP/WVP-3000) to model the impact of clouds and water vapour on Earth-Space links. This instrument uses 12 channels, five in the K-band between 22.235 and 30 GHz for water vapour profiling, and seven in the V-band between 51.25 and 58.8 GHz for temperature profiling.

The profiling radiometer can retrieve vertical profiles of temperature, humidity and cloud liquid water every 30 seconds or every minute from the surface up to 10 km in height under clear sky and cloudy conditions. It is also capable of assessing the columnar cloud liquid water content (also known as the liquid water path L) and the water vapour path continuously, unlike what can be achieved with conventional radiosondes. The profiler also features an infrared thermometer that detects the cloud base temperature: the cloud base height of the lowest layer can thus be estimated using the contemporaneous temperature profile.

Statistics on retrieved liquid water path, cloud height and thickness for liquid-phase and mixed-phase Stratus and Stratocumulus clouds observed over CRC in 2005 and 2006 during the daytime will be presented in this paper. According to the land cloud climatology of Hahn and Warren et al. (2006), these clouds are one of the most frequently occurring liquid-bearing types over a 5-degree by 5-degree grid box that includes the Ottawa area.

Cloud height and thickness will be assessed using CRC’s cloud detection algorithm [Bouchard, 2005]. It is based on the Chernykh and Eskridge (CE) method for detecting cloud layers and amounts that was originally designed for radiosonde profiles [Chernykh and Eskridge, 1996]. Our algorithm uses the CE method with retrieved vertical profiles of temperature and relative humidity, but also data not measured by radiosondes: the cloud base height estimated by the infrared thermometer, and the retrieved L. This extra set of conditions can be summarized as follows: a cloud contains liquid water whenever L exceeds a certain threshold value and its cloud base height is lower than some maximum value. This cloud detection algorithm has been shown to significantly reduce the number of false positives compared to the application of the original CE method [Bouchard, 2005].

These results will be compared to the 6-year climatology of midlatitude low-level continental clouds from the United States ARM Southern Great Plains Central Facility, Oklahoma, reported in Dong et al. (2005).

References:

P. Bouchard, "Improved algorithm for detecting cloud layers and amounts using retrievals from a surface-based multifrequency profiling radiometer," in Remote Sensing of clouds and the Atmosphere X, edited by Klaus Schafer, Adolfo T. Comeron, James R. Slusser, Richard H. Picard, Michel R. Carleer, Nicolaos Sifakis, Proc. SPIE Vol. 5979, Article 597903 (2005). Chernykh, I.V and R.E. Eskridge, 1996: "Determination of cloud amount and level from radiosonde soundings." Journal of Applied Meteorology, vol. 35, no. 8, pp. 1362-1369. Dong X., P. Minnis and B. Xi, 2005: "A climatology of midlatitude stratus clouds from ARM SGP site. Part I: Macrophysical, microphysical and radiative properties." J. Climate, vol. 18, pp. 1391-1410. Hahn, C.J. and S.G. Warren, 2006: Climatic atlas of clouds over land and ocean. Available on the Web at: http://www.atmos.washington.edu/~ignatius/CloudMap/.

The flexibility given by a reconfigurable antenna can be exploited in various ways: redirecting capacity, tailoring the shape so as to match the borders of the served region, maximizing the total throughput etc. Another possibility, particularly valuable at Ka-band owing to the sensitivity to atmospheric attenuation, is that of minimizing the number of non-served users in a broadcasting system. This is achieved by redirecting the power flux so as to concentrate more power towards regions affected by rain. The efficient utilization of the limited power available on board allows the achievement of the same system availability of a fixed antenna but with lower power levels.

Modern satellite-antenna technology could reach a spatial resolution on ground which is comparable to that of the atmospheric weather features to be mitigated. In the case of Europe this goal could be obtained by a reflector antenna of about 2m diameter illuminated by a focal array composed by about 100 feeds.

This paper presents an assessment of the performance of a broadcasting system operating in Ka-band based on simulations utilizing high resolution weather spatial fields, provided by the European Centre for Medium-Range Weather Forecast (ECMWF) and concurrent Meteosat images. The simulation covers a significant period of time (some years) so as to provide a statistically sound estimate of the adaptive gain.

A balance of the adaptive gain vs. the limitations of reconfigurable beam-forming network will be attempted so as to assess the cost- and risk-effectiveness of such technique for systems in the near future.

This work is supported by the ESA/ESTEC Contract n. 17877/04/NL/JA.

Nowadays, there are strong needs from end users to manipulate multimedia information and therefore high data rates that require wider bandwidths. High frequency bands such as Ka-band (20-30 GHz) or Q/V-band (40-50 GHz), in which wider bandwidths are still available, are therefore currently envisaged by satelmite operators to establish new satellite communication systems.

However, these high frequency bands suffer from some limitations such as the lack of maturity of the technology, high intra-system interference (co-channel) due to the use of multi-beam antennas and strong propagation impairments such as: attenuation due to oxygen, water vapour, clouds and especially rain, scintillation in both clear sky and rain conditions, as well as depolarisation due to rain and ice.

In order to provide a sufficient availability on the end-user point of view, to optimise system capacity and to offer a quality of service similar to terrestrial networks, new Satcom systems have to be adaptive in order to match traffic and/or propagation conditions, thanks to the use of Fade Mitigation Techniques with Adaptive Resource Management. In order to design these techniques and their associated access protocols and to minimise intra-system interference, complex system simulation is needed.

To carry out the optimisation process, traffic and channel models are required. As far as propagation models are concerned, the available statistical methods, that allow distributions of propagation effects to be predicted are not sufficient anymore and a fine description of the temporal and of the spatial behaviour of the propagation channel is needed.

The objective of this paper is to describe recent activities carried out in this field for ESA, CNES and in the framework of European projects. In a first part, the temporal behaviour of the propagation channel is briefly addressed thanks to the development and validation of time series synthesisers. In a second part, which is the main core of the paper, it is shown how space variation of propagation impairments can be described with the development of generators of rain attenuation fields from 2 studies carried out by Politecnico di Milano for ESA and by ONERA for CNES. Then, the "hot topic" of the space-time variability of the propagation channel is dealt with, thanks to current research activity carried out by ONERA and Politecnico di Milano.

These activities have been or are carried out in the framework of: - ESA study n° 16532/02/NL/EC "Adaptive Coding Modulation Techniques for ka /Q Band Systems",

- ESA study n°16865/03/NL/EC: "Development of propagation models for telecom satellite systems",

- ESA study n°17760/03/NL/JA: "Characterisation and modelling of propagation effects in 20-50 GHz band",

- CNES study n°02/CNES/0775/00: "Modelling of the space variation of rain attenuation"

- CNES study n°DCT/RF/TT-2005.935: "Modelling of the space-time variation of propagation impairments in 20-30 GHz frequency band"

- COST 280 action: "Propagation impairment mitigation for millimetre-wave radio systems",

- The EU Network of Excellence: SatNEx.

System simulation is of current interest in the study of adaptive fade mitigation techniques. There is considerable activity in the development of attenuation time-series generators that are capable of reproducing typical fade events. One approach is to use measured or modelled data from weather radars or numerical weather prediction models. One of the limiting factors of such approaches is the limited spatial and temporal resolution on which the data is available. For example, in order to simulate a particular fade mitigation technique there may be a requirement for 1s attenuation time-series. It is not at all clear that a representative 1s timer-series can be derived from a series of 5-minute interval data.

It is generally accepted that rainfall demonstrates fractal scaling. That is, rain processes exhibit a degree of self-similarity on all spatial and temporal scales. One way to alleviate the problems of finite resolution in fields such as rainfall rate is to artificially introduce synthetic small-scale structures. This can be achieved by exploiting the fractal scaling properties of rain, so as to introduce small-scale structures while preserving the original statistics. This technique produces a representative small-scale structure overlaid on the original larger scale structure. Although the additional small-scale structure is determined stochastically, the resulting fields display the correct properties making them suitable for simulation purposes.

In this paper we will present an overview of fractal downscaling applied to temporal and spatial rainfall fields. An analysis of fractal dimension and generalised structure functions will be presented to demonstrate the fractal properties of rain. The technique will be applied to numerical weather prediction estimates of rainfall and to data from the 3 GHz Chilbolton Advanced Meteorological Radar (CAMRa). It will be shown that the downscaling process conserves both the statistics and the power spectral density properties.

Next generation satellite systems and High Altitude Platform (HAP) systems designed to provide broadband services will work in Ka and V bands. At such frequency bands, a link may suffer severe degradation caused by tropospheric factors, namely: cloud and rain attenuation, scintillation and gaseous attenuation. To compensate for these effects, various Propagation Impairment Mitigation Techniques (PIMTs) can be implemented so that the system of interest is able to maintain a certain level of performance in any atmospheric condition. Among these PIMTs, antennas with reconfigurable patterns that produce multiple beams, when installed in satellite systems with on-board-processing, may be used to optimize the allocation of the available power to given sections of the coverage area. Thus, the power directed towards sections affected by bad weather (clouds, rain) can be increased, making use of spare on-board resources. These are Resource Sharing Systems (RSS) techniques whose development is strongly dependent on the thorough knowledge of the spatial properties of various atmospheric phenomena.

Both spatial distributions of cloud cover and rainfall rate in Spain have been investigated in the framework of COST Action 280. The study includes the analysis of cloud cover for individual sites and the spatial correlation properties for pairs of sites for both atmospheric factors. For the cloud cover study, 5 years of synoptic meteorological data, collected every 6 hours in 33 sites were employed, while data from about 50 sites (over 10 years of data per site which combine into more than 700 pairs of sites) were available for the rainfall rate analysis. The sites are located in areas of continental, Mediterranean and maritime climates with inter-site distances ranging from a few tens of km to 900 km.

Cloud cover amount probabilities for single sites derived from local synoptic records, as calculated in the proposed study, may be useful to produce and test new cloud attenuation models needed for the design of satellite and high altitude platform systems, as well as for the development of various PIMTs.

In order to quantify the spatial properties, the simultaneous occurrence of cloud cover, or rainfall rate, in pairs of sites has been investigated using the correlation coefficient and the statistical dependence index. The type of cloud or the rainfall rate threshold has been considered for the analysis of the variations of both parameters with distance.

The cross-correlation function of rainfall rate has been obtained for pair of sites, obtaining results that are dependent on the relative location of the sites with respect to the displacement of weather fronts.

Most of the previous statistical results have been incorporated into maps by means of contour lines. The orographic characteristics of the region, the proximity to the sea and the dominant movements of the rainy fronts determine the shape of the contour lines for both phenomena (rain and cloud cover).

The procedure employed to calculate the various statistics from the available data and to derive the corresponding maps could be useful for other meteorological factors and other types of data, and also in other regions of the Globe.

An accurate evaluation of the impairments on microwave signal propagation due to atmospheric gases and clouds represents a crucial aspect in both radio propagation problems and remote sensing studies.

Synthetic datasets of propagation parameters are often generated for this purpose from radiosonde observations and analyses from numerical weather prediction models. Since radiosondes do not measure cloud water density, models are necessary to estimate cloud liquid and ice density profiles for radiative transfer algorithms that simulate electromagnetic parameters such as atmospheric attenuation and brightness temperatures. Clouds are identified from humidity profiles when the relative humidity exceeds a suitable threshold function (which depends on the type of radiosonde humidity sensor being used).

In this work, we analyze two cloud models that are currently in use in propagation and remote sensing simulations in the presence of non-precipitating clouds: the model proposed by Decker et al., 1978, and the model proposed by Salonen and Uppala, 1991.

We also suggest an improvement of the Salonen`s humidity threshold (Basili et al., 2006) for detecting the presence of clouds, and implement it in the Salonen's cloud water model.

The performances of these models are assessed at the Atmospheric Radiation Measurement (ARM) Program’s Southern Great Plaints (SGP) site in Oklahoma, USA, by using data from Vaisala RS90 radiosondes, a ceilometer, a dual-channel microwave radiometer at 23.8 and 31.4 GHz (MWR), and an ARM Active Remotely Sensed Cloud Location (ARSCL) value-added product derived from the millimeter cloud radar, the ceilometer and the micropulse lidar. First, cloud base heights detected by the models are compared with those provided by a ceilometer. Then, cloud top and thickness identified from the radiosoundings are compared with the ARSCL data. Finally, the capability of the models to estimate cloud liquid and ice content is evaluated by comparing simulated brightness temperatures from the radiosondes, in the absence of scattering, with those measured by the radiometer.
A similar comparison is also performed at the site of Pomezia in Italy, by using data from Vaisala RS90 radiosondes, a ceilometer, a dual-channel radiometer at 23.8 and 31.7 GHz, and a 13-GHz radiometer.

A novel approach for computing cloud liquid density from an atmospheric vertical profile is also analyzed. A new cloud liquid model has been derived by Pulvirenti et al., 2005, based on the outputs of numerical simulations of a cloudy event occurred in the Mediterranean basin, and performed by means of a meso-scale forecast system with explicit cloud microphysics. The model assessment was carried out through a statistical comparison between the histograms of brightness temperature simulations from radiosondes and European Centre for Medium-Range Weather Forecasts (ECMWF) analyses with those from the Special sensor Microwave/Imager (SSM/I) over the Mediterranean Sea. In this work, we validate it at the site of Pomezia, since the model has been developed for a mid-latitude environment. Nevertheless, the methodology can be easily extended to any geographical zone where such meso-scale simulations are available

At the core of almost every radio attenuation model there are the statistics of rainfall. Although vital for the understanding of the attenuation phenomenon, the behaviour of rainfall could not be a more difficult entity to model. Being variable both in three-dimensional space and in time, rainfall, represented by its measurable parameter the rainfall rate, has to be treated as a multidimensional random process. The complexity of such a model makes it inapplicable for practical modeling purposes, such as those sought by the ITU and presented in many of its Recommendations. Simplifications are required in order to produce a workable model for the prediction of attenuation on microwave links due to precipitation.

In terrestrial links, the importance of the spatial structure of rainfall is paramount. Given the length of most microwave terrestrial paths, the radiowave, as it propagates through the lower troposphere, may cross many rain cells each with different rainfall rate profiles. The cumulative effect of such crossings on the received signal level is a spatial-temporal average of the attenuation due to the rain path.

To model this effect, a facilitator must be used. Many attenuation prediction models use the concept of effective path length, which corresponds to the propagation in uniform rain that would produce the same distribution of attenuation as observed in the actual path, and attempt to model it as a function of point rainfall, path length and additional parameters. Alternatively, an effective rainfall rate can be defined, that produces the same attenuation on the path as the actual, non-uniform rain field.

In this paper, the statistical characterization of the effective rainfall rate parameter is obtained based on the cumulative distribution of measured signal attenuation in converging links with different path lengths, spatial orientations and operation frequencies.

Results are based on rain attenuation data measured at 2 sites in tropical climates in Brazil. At the first site (São Paulo) the received signal from five links operating at 15 GHz and two links at 18 GHz was recorded. The second site (Brasília) provided data from the six links operating at 23 GHz and two links operating at 38 GHz. Without assuming any specific functional form for the distribution of measured attenuation, the cumulative distributions of effective rainfall rate that were obtained were fit to a variety of shapes, regardless of being statistical distributions or not.

In order to construct a model relating the obtained power law coefficients and the geoclimatical characteristics of the region in which the sites are located, namely path length, orientation and point rainfall rate, a multivariate regression analysis was performed, the results of which are presented and discussed.

Drop shape and size distributions in rain are two fundamental and important parameters which govern attenuation and depolarization effects along earth-space paths. Much work relating to the characterization of the two has been done in the past and it is generally recognized that they can be not only location dependent but also regime dependent, i.e. stratiform, convective, shallow warm rain, thunderstorm etc.

The 2-dimensional video disdrometer (2DVD) has been shown to be capable of providing accurate measurements of both size and shape distributions [1]. In the case of the latter, the axis ratio distributions were derived in terms of the equivalent drop diameter and shown to be different from the standard Pruppacher-Beard shapes for drop diameters less than 4 mm. The axis ratios were derived assuming equivalent oblate spheroidal shapes.

Here we report the actual probability contours of drop shapes derived directly from the two fast scanning cameras of the 2DVD. Three examples are given below corresponding to 3, 4 and 6 mm and compared with the Beard-Chuang theoretical equilibrium shape (static) model (black line). Deviation from the model is evident for diameters above 4 mm, the larger drops having more flattened bottom than oblate spheroids. Note the finite distributions of the contours represent drop oscillations.

Using these more accurate shapes and the measured size distributions in various locations, T-matrix calculations are performed to obtain the specific attenuation for horizontal and vertical polarizations at 20 GHz. The variation of cross-polar discrimination versus co-polar attenuation is re evaluated for the 20 GHz scenario considered in the earlier study [1]. Preliminary calculations do not show significant differences. However, differences are expected for very high rainrate events where the large drops become more significant in number.

Reference:
[1] M. Thurai, V.N. Bringi and A. Rocha, 2006: 'Specific attenuation and depolarization in rain from 2-Dimensional video disdrometer data', Submitted to IEE Proc. Microwaves, Antennas & Propagation.

Measuring precipitation from space is a long standing issue of operational meteorology, hydrology and climate. Since the first launch of meteorological satellites in the 60s methods for "inferring", rather than "measuring" rainfall intensity from space were conceived and calibrated from time to time. Products have greatly evolved in time and always offered a sufficient quality when averaged over suitable time and space scales, those of climate above all, but generally suffered from contradictory performances when coming down to instantaneous rainrates. Modern methods go back to the basic physical issues of precipitation formation and evolution since the cloud physics content of the algorithms is re-examined and better observational and modelling tools are now available.

The paper will address the current status of the international efforts in the field of precipitation measurements from space presenting:
- Status of physical algorithms.
- Status of missions including GPM, CloudSat, and Earth Care.
- Activities of international organizations, like the International Precipitation Working Group, the Global Precipitation Climatology Project and others.
- An overview of open problems and international research efforts, including generation of high resolution precipitation products for propagation applications.

This paper describes the contours of the Dutch CESAR monitoring and research site for atmospheric processes. CESAR Observatory, located in the heart of The Netherlands, consists of a multitude of remote sensing and in situ instruments, augmented by observations from space. The observatory is a key tool in the national Dutch program to monitor the physical properties of the complete atmospheric column with as much detail as possible. Specific research programs are developed to measure the hydrological cycle, land-atmosphere processes and the cloud-aerosol-radiation interaction. While the main motivation of the work at the site is given in by regional climate change, other applications are also envisaged: