Journal of Applied Sciences Research, 2(11): 1014-1018, 2006 © 2006, INSInet Publication

Perspectives on the Effects of Harmattan on Radio Frequency Waves

D.D. Dajab

Department of Electrical Engineering, Ahmadu Bello University, Zaria, Nigeria.

Abstract: Harmattan dust clusters at its highest intensities can reduce visibility to less than a hundred meters.Similar like absorb and scatter electromagnetic energy leading to attenuation in its signal strength.In this paper spatial harmattan particle densities as approximated from visibility data are utilized to determine the specific particle attenuation possible on radio frequency signals.The attenuation variations obtained indicates a significant leap in the values during harmattan and non-harmattan periods.The predominant propagation mechanisms likely during the harmattan and how the loss in power is incurred in radio frequency signals are also discussed.

Key words:

INTRODUCTION MATERIALS AND METHODS

The harmattan is a condition in the in The records of daily meteorological observations, which dust particles (precipitates) are blown up into the obtained from the Meteorological Department of the air by defined as air in horizontal motion relative to Federal Airports Authority of Nigeria (FAAN9), the earth surface1 and pushed southwards from the Zaria Aerodrome,and the meteorological section of the desert by the northeast winds. Harmattan intensity may be soil science department of the Institute for Agriculcural so great that visibility at ground level is reduced to less Research(IAR10) serve as the inputs to the foregoing than a hundred meters by the dust clusters. Harmattan analyses.The data cover the periods 2001 to 2002. occurs in Nigeria during the dry , that is, between There are three ways of evaluating the performance the months of November and March and with its of radio communication systems: by mathematical micronsize particles, harmattan dust clusters resemble that analysis, by computer simulation and by measurement. of and the space they cover can be considered as a The characteristics of these techniques can be summarized dielectric – since the clusters consists predominantly of as follows: quartz layer which non-coherently scatters RF signals 2. Propagation studies are required to identify the likely Mathematical Analysis: This offers quick results with a channel characteristics that would be encountered in any very good insight into what is actually happening with the operating environment in order to form the channel model process being evaluated, for example, it enables us to see of the physical layer of the communication system. the relationship between the various parameters in the Established wireless systems in the frequency range system and the effect of the variation of one parameter. between 500 MHz and 5 GHz have resulted in the However, in most cases it is necessary to make many development of different approaches to the prediction of simplifying assumptions to enable such an analysis be wave propagation. Within this frequency range are used. This is inevitably the case when communication is different telecommunication applications and services3. taking place over propagation channels with multipath Studies have shown 4,5,6 that precipitations such as those and fading. of rain, harmattan dust etc absorb and scatter electromagnetic energy which leads to attenuation in its Computer Simulation: An increasing trend in the field signal strength. Improved models are desired that explore of computer simulation of complex systems of all types the realm of overall macroscopic signal propagation in a (that is, link or physical level and network models) is local environment and that adequately captures the time- away from dedicated programs written in conventional varying behaviour of representative communication high level languages (for example, FORTRAN, C, C++) channels7. Wireless communication links are generally towards the use of high-level simulation packages, described as time-varying frequency-selective fading wherein simulation models are built up from blocks that multipath channels. The topology, link performance and are usually graphically represented and joined together in quality of service (QoS) delivered to the user is a drawing package-like environment. High level packages 8 characterized as highly time varying . of this nature are available for a wide range of application

Corresponding Author: D.D. Dajab, Department of Electrical Engineering, Ahmadu Bello University, Zaria E.mail: [email protected]

1014 J. Appl. Sci. Res., 2(11): 1014-1018, 2006

areas. The area of suitable application of a package Available meteorological data on harmattan dust, depends on features such as data supported in the however, are currently presented in terms of visibility in libraries. units of distance and the question is how to relate these records of ground level visibility to the actual particle rate Measurement: This is more time-consuming than (or density) of harmattan dust in the air. It is known1 that computer simulation if hardware design or redesign is a pressure increase and a temperature decrease are required and likely to prove very expensive. However, it responsible for the movement of winds, which serve as a does provide real-time operation and the most accurate means of transporting heat, dust particles, moisture etc. estimate of performance if used in conjunction with a Here a spatial particle density ρ [g/m3] may be a channel simulator or by trialing in the real-world 11 meaningful concept, given that the dust particle carrying environment .Of these mathematical analyses method capacity of the winds depend on the relative . will be applied in this paper. The spatial harmattan particle density may be Generally the problem of estimating the attenuation approximated15 by the empirical relation. of RF waves caused by the various forms of (rain, , fog and harmattan dust etc) is quite difficult. In practice it has been convenient to express the (VV−ρ) ρ= fm i p g/m3 (4) attenuation due to rain as a function of the precipitation i h rate 12,5,13. In the case of harmattan, no known attempt at quantifying or measuring its precipitation rate is available except the measure of how it distracts visibility (with Where Vfm - the best visibility figure for the year respect to distance). To fully understand the mechanism (period) [m] of harmattan attenuation therefore it is essential to Vi - observed (measured) visibility profile [m] 3 estimate the rate of the harmattan precipitations. ρP - mean particle density [g/m ] By analogy to the attenuation of microwaves by h - humidity profile (in decimals) [g/m3] rainfall13, the specific attenuation due to harmattan dust particles may be predicted using eqt(1) though further We will assume a state in which the harmattan experiments are required to establish the validity of this dust particles consists of a subset of the complete set interpretation. of fixed sizes (which have small details compared to the wavelength) and with a certain distribution per b arain = aρ dB/km (1) cubic meters (which may be ascertained experimentally) and then we may obtain using the adapted eqn (1) Where ρ- the rainfall rate [mm / hr] b aparticle = aρ , the specific particle attenuation variation a, b – regression coefficients which are a function of (dB/m). the frequency f, of the temperature T, of the drop size distribution and to a lesser extent, of the RESULTS AND DISCUSSIONS wave polarization The problem of calculating the values of a and b for The transmission condition is the degree of mutual a given meteorological data is given by the relations14. visibility between the transmitter and receiver antennas; the conditions that may be encountered in practice are ⎛⎞⎛⎞⎛⎞nnn line-of-sight (LOS), partial LOS and non-LOS 16. nxyx⎜⎟⎜⎟⎜⎟∑∑∑ii− ii y i b = ⎝⎠⎝⎠⎝⎠i1=== i1 i1 The periods selected for the purpose of this study nn2 capture the markedly differentiated of the year ⎛⎞⎛⎞2 (2) nx⎜⎟⎜⎟∑∑ii− x (i.e. May-June and Nov-Dec). From the temperature ⎝⎠⎝⎠i1== i1 curves shown in Fig. 1, it can be observed that the onset of the harmattan corresponds to the November-December nn period during which temperatures are lowest, while May- ∑∑ybxii− June is a period corresponding to the onset of the rainy a = i1== i1 (3) (wet) season when the temperatures are moderately n higher. where – is the number of pairs of observation The low humidity which occurs at the onset of x- is the temperature observations harmattan from November makes the air to be very dry y- particle size distributions (consequently all surrounding ground materials) and the

1015 J. Appl. Sci. Res., 2(11): 1014-1018, 2006

25 May 0.035 (a) 20 0.03 icro dB)

m 0.025 15 0.02 perature

m 10 2002 0.015 e T 2003 ttenuation (

A 0.01 5 0.005

0 Specific 0 1 2 3 4 5 6 7 8 9 10 11 12 0 5 10 15 20 25 Month Hour of the day Fig. 1: Annual temperature profile December 800

) (b)

B 700

100 d

90 m 600 ( n 80 500 atio

u 400

70 n 300 60 200 ic Atte

50 May f June 100 40 Nov peci

S 0 30 Dec 0 5 10 15 20 25 -100 20 Hour of the Day Fig. 3: Specific attenuation(a) May (b) December. 10

0 5 0 5 10 15 20 25 themselves but to scattering which depends on relative Hours values of the wavelength and the size of the particles. Fig. 2: Daily average ambient relative humidity Thus, the total pathloss will be expressed as

low temperatures too cause the North East winds to ⎛⎞d blow across the surface of the earth raising with it Lppo10particle() d=+ L ( d ) 10n log⎜⎟ +α d [] dB an increasing content of dust particles into the air. ⎝⎠do (5) Generally, it can be observed that the appearance of

harmattan is always associated with a drop in the Where d0 = 0.99 distance of the reference far field ambient relative humidity, while its absence is reflected from the transmitter

by an increase. From the curves of Fig3. the lowest Lp(d0) is free space pathloss humidity periods relate to the highest harmattan n is the propagation pathloss exponent dust densities. The estimated density of harmattan dust for The study of the mechanism of harmattan attenuation the period of study are as presented in Table 1 of radio frequency signals will involve the examination of (Assuming the mean particle density to be its dielectric constant structure using equipment such as equivalent to the mean particle density in air borne refractometers which goes beyond the scope of contaminated water = 0.889 ± 0.01 g/m3 )17. our present consideration. The curves in Fig 3 clearly indicate the possibility of It is a fact that whereas rain drops are falling with a increases in attenuation due to particle attenuation as the velocity, most harmattan particles are suspended (or harmattan intensity deepens. The attenuation ranges from moving) in space and exhibit complex form of motion. 4.4 µdB up to 33 µdB and from 2.13 mdB up to 718 mdB The harmattan dust particles may be said to exhibit two for May and December respectively(assuming a distance major forms of interaction, that is, electromagnetic and of 15m). There is remarkable leap as observed in the gravitational. The interaction between particles which are magnitude of the particle attenuation between the two charged are referred to as electromagnetic interactions, on periods. The power absorption and scattering by these the other hand, gravitational interactions are charge particles (harmattan dust, fog etc) increase rapidly with independent (rather mass dependent)18. Theoretical frequency and with density of the precipitation. The treatment of the electromagnetic interactions is the subject greatest losses are not due to absorption in the particles of quantum electrodynamics.

1016 J. Appl. Sci. Res., 2(11): 1014-1018, 2006

Table 1: Harmattan dust density in air (ρi) Month May 2002 June 2002 November 2002 December 2002 ------oC 3 o 3 o 3 o 3 hours T( ) ρi(g/m )T(C) ρi(g/m )T(C) ρi(g/m )T(C) ρi(g/m ) 0 22.5 40.28 21.2 1.02 17.8 114.29 13.5 205.15 ------3 22.8 35.06 20.6 1.01 15 114.29 10.8 133.35 ------6 22.5 35.81 20.1 1.95 14.3 108.22 8.9 106.68 ------9 25 45.23 24.4 1.19 23.5 248.9 18.9 322.25 ------12 30.7 55.56 26.8 1.46 28.5 311.13 23.7 622.25 ------15 30.5 57.88 27.1 1.56 27.5 286.44 25.5 429.67 ------18 22 50.32 21 1.34 23.5 161.13 18.5 368.28 ------21 22.7 42.19 21.5 1.08 20.2 135.68 15 234.36

Considering, however, that wavelengths that are long compared to the particle size, C The velocity of the electromagnetic waves is much the attenuation due to scatter is larger than the attenuation greater than the velocity of the harmattan particles or due to absorption5. When the wavelength is large (that is 3 x 108 m/s >> about 5m/s) and compared with the size of the particle, the wave is C There is a strong inhomogeneity in harmattan dust scattered in arbitrary or all directions, but when the particles, wavelength is small compared with the particle size then most of the scattering takes place within a narrow cone It can be assumed that the dust particles may be surrounding the forward direction of propagation of the approximated as static and the consideration of its incident radiation4. The wavelength of the Global System interactions are best reduced to the interaction of particles for Mobile telephony(GSM) at 900 MHz is 33cm which with their nearest neighbours to give a good is fairly large relative to the sizes of the harmattan dust approximation to the total interaction. In such an particles. In this circumstance, the mechanism of imperfectly conducting medium(dielectric space) the scattering is predominant. As a result plane waves become dominant harmattan propagation mechanisms are the inhomogenous type and the amplitude of the wave is absorption and scattering. no longer constant over planes of constant phase. This inhomogenous (scattered) plane wave has the following Absorption Mechanism: The field strength of the properties: transmission is given19 by; C The direction of propagation of the wave is along the ω ⎛⎞dn − pd jtω−⎜⎟ X-axis. ii) The amplitude of the field components c ⎝⎠c EEeezm= V/m (6) (scattered) decrease exponentially in the + z direction20. Equation (6) shows that propagation in an imperfect and the scattering of energy in arbitrary directions gives medium is accompanied by a reduction in the wave rise to a net loss of energy in the direction towards the amplitude in an exponential fashion. The quantity receiver5. ω δ= pm−1 (7) Conclusion: Attenuation or signal fading can be caused c by transmission path length, obstructions in the signal is called the absorption coefficient and the velocity of path (such as harmattan dust clusters) and multipath waves in the imperfect medium is v = c/n m/s effects, thereby leading to a drop in the signal power. The harmattan dust filled space has the dielectric When more than one path are available for radio signal particles subjected to polarization by the electric field. propagation, multipath occurs. The phenomenon of The inducement of dipoles, in the particles making up the reflection, diffraction and scattering etc, all give rise to volume of space absorb electrical energy as in (6) and the additional radio propagation paths beyond the direct “line waves suffers attenuation as it proceeds. of sight” path between the transmitter and the receiver. The other impacts of these impairments include: path loss, Scattering Mechanism: Attenuation due to absorption is temporal and spatial variation of path loss and larger than attenuation due to scatter for wavelengths that polarization mismatch. As a result of multipath the are small compared with the particle size. For received field at a terminal is made up of a combination

1017 J. Appl. Sci. Res., 2(11): 1014-1018, 2006 of the direct rays and the reflected, diffracted, scattered 9. Weather Department, 2001-2002. Federal Airports rays, etc, which experience longer paths than the direct Authority of Nigeria (FAAN),Meteorological rays, are thus delayed and result in the received waveform Records. being a summation of the transmitted replicas of the 10. Soil Science department, 2001-2002. Institute for various weighted time shifts of the rays which could add Agricultural Research(IAR),Meteorological Records up distractively. 11. Alston, I.D., J.C. Currie and C.M. Dearlove, 1997. “Modelling and Simulation of Communication REFERENCE Systems” the GEC Journal of Technology, vol. 14, No. 2. pp: 75-84. 1. Lal, D.S., 1991. “”Climatology” Chaitanya 12. ITU Report 234 1963. “Propagation: Influence of Non-ionized regions of the Atmosphere on the Publishing House. th 2. Neyman, A.B., 1981. “Study of Short Wave Propagation of Wares” CCIIR X plenary Assembly, Reception in Zaria” PhD dissertation, Electrical Engr Vol. 2. Dept, ABU. 13. Filippo Giannetti, Marco Luise and Ruggero Reggiannini, 1999. “Mobile and Personal 3. Gerd Wolfle, Philip Wertz and Friedrich M. Communications in the 60GHz Band: A Survey” Landstorfer, 2003. “Performance, Accuracy and wireless personal communications Journal, Vol. 10, Generalization Capability of Indoor Propagation No. 2 pp: 207-243. Models in Different Type’s of Buildings” University 14. Ezekiel, M., 1941. Methods of Correlation Analysis, of Stuttgart, Pfaffenwaldring 47, D-70550 stuttgart, 2nd Edition, New York, John Wiley, chp 12. Germany. Email: [email protected]. 15. Dajab, D.D., 2005. Characterisation and Modelling 4. Glazier, E.V.D. and H.R.L. Lamont, 1958. of 900MHz Indoor Wireless Communication “Transmission and Propagation” Her majesty’s Channels in the Savannah Region,PhD Stationary Office, Vol. 5. Dissertation,Electrical Eng Dept,Ahmadu Bello 5. Ingvar Henne, Per Thorvaldsen, 1999. “Planning of University,Zaria nd Line-Sight Radio Relay systems” 2 Edition, NERA 16. Nikki Swartz, 2000. “Ready, Set, Wireless Office; Telecommunications. “wireless Review, vol. 17, No. 11, pp: 28-34. 6. ITU Report 227-1 1974. “Propagation in Non-ionized 17. Le Guer, Y., P. Reghem, I. Petit and B. Stutz, 2003. Media: General Methods of Measuring, Field “Experimental Study of Buoyant Particle Dispersion Strength and Related Parameters” CCIR XIIIth in Pipe Flow”. Journal of Chemical Engineering Assembly, Vol. 5, pp: 143-151. Research and Design, vol.81, No. A9, pp: 1136-1143. 7. James, C. Chen, and Jeffrey M. Gilbert, 2001. 18. Longo, M.J., 1973. “Fundamentals of Elementary “Measured Performance of 5.3GHz 802. 11a Particle Physics” McGraw Hill. Wireless LAN Systems” Atheros communications, 19. Dolu Khanov, M., 1971. “Propagation of Radio Inc, www.atheros.com waves” Mir Publishers, Moscow. 8. Alan Stevens, 2003. “Network: Five-Minute Guide 20. Clarke, R.H., and John Brown, 1980. “Diffraction to Wireless Networking” PC world, pp: 152, Theory and Antennas” Ellis Horwood Limited, www.pcw.co.uk. England.

1018