## Effects of Vegetation

An empirical model for representing the shadowing effect induced by trees along roads has been made available for the 2-30 GHz frequency range by the ITU-R (ITU-R P.681 4.1). In this model, the influence of the trees is represented by the percentage of optical shadowing resulting from their presence for a 45 degree elevation angle along the direction of the source of the signal.

In the case where the environment is known with an adequate level of precision (for example, if geographical base contour data are available), ray models can be employed for the determination of the coverage area of a transmitter at any frequency.

The attenuation due to vegetation appears to increase with the frequency. However, while the ITU-R provides a few attenuation models in the UHF band, data available at frequencies higher than 10 GHz are few and scarce. This lack of results has been repeatedly stressed in the last recent years (Grindrod 1997; Seville 1997). Further, the existing results present quite different values for this parameter, depending on the type of vegetation under consideration.

An ancient ITU-R model gave the following equation for the excess attenuation due to vegetation:

where f is the frequency, d is the distance travelled inside vegetation and a, p, and y are three parameters of the model (CCIR 1986).

A comparison between this equation and experimental measurements with the coefficients indicated by the CCIR yields a 22 dB standard deviation error (Seville 1997). Adjusting the coefficients permits to bring down the error to 11 dB. Taking into account the region simultaneously illuminated by the emitting and receiving antennas, the standard deviation error can be further reduced down to 8 dB (Seville 1997). In the aforementioned study, the author points out the great diversity of results, which depend on the type of vegetation: for instance, whereas for a spruce the transmitted wave attenuation is on an average equal to 10 dB, it is equal to 22 dB for a ficus at 40 GHz.

An experimental and theoretical study based on the energy radiative transfer theory (Schwering et al. 1988) was conducted at the 9.6, 28.8 and 57.6 GHz frequency ranges. This study reveals that the vegetation attenuation expressed in dB as a function of the distance d inside the vegetation increases linearly and rapidly, with rates ranging from 1.3 to 2 dB/m when d is relatively small, i.e. smaller than 30 metres. As d increases, the attenuation due to vegetation passes through a transition phase beyond which it still linearly increases, albeit at the slower rate of 0.05 dB/m. The explanation advanced is that this phenomenon results from the combination of the rapidly attenuated path with the more slowly attenuated diffused paths. If the diffused paths are predominant, the depolarisation of the waves can reach a high degree.

In some propagation models, the vegetation in urban areas is regarded as a shadowing region which prevents the propagation of electromagnetic waves (Cor-reia 1996). At the 60 GHz frequency, the authors indicate values ranging from 6 to 8 dB for the attenuation due to vegetation (Correia et al. 1994).

A few qualitative results have been presented in a study realised at frequencies near 28 GHz by the US-WEST. According to this study, the attenuation due to vegetation may reach the order of several tens of dB. The excess attenuation due to vegetation and shadowing effects in suburban areas compared to free-space attenuation was found to have an average value, of interest from a statistical point of view, equal to 4 dB/km with a 10 dB standard deviation at the 28 GHz frequency (US-WEST 1995).

The presence of vegetation is also the cause of the fast fading of the received field, correlated with the speed of the wind. In narrow band, an adjustment of this fast fading can be performed using a Rice-type law (US-WEST 1995). The study conducted by US-WEST shows that in this case the Rice parameter is correlated to a reasonable extent with the additional attenuation due to vegetation.

Measurements of the attenuation of the signal have also been carried out at the 60 GHz frequency in forest areas along roadside trees whose branches extend over the road (Grindrod et al. 1997). The transmitter is therefore only partially in line-of-sight. Up to a distance equal to 20 metres, the attenuation due to the foliage hanging over the road is no higher than 5 dB. Beyond this distance, strong variations of signal can be observed, which are explained by the authors by the interferences arising between the direct path attenuated by the leaves and the diffused paths. When the transmitter is placed directly behind a group of conifers, the attenuation reaches values ranging from 30 to 50 dB, corresponding to distances inside vegetation ranging from 60 to 80 metres.

In addition, the crossing of vegetation causes a depolarisation of the waves in direct proportion to the additional attenuation that it induces. For an initial isolation value of 24 dB between the two horizontal and vertical polarisations in freespace propagation, experiments carried out by the US-WEST have shown that the isolation of the two polarisations decreases by 1 dB when the additional attenuation due to vegetation rises by 3 dB. It suggests that the depolarisation effect due to vegetation is much more important than the depolarisation due to rain.

For further detail on the evaluation of attenuation due to vegetation, the reader is referred to Appendix J on vegetation attenuation.

## Post a comment