Read Radome Effects on Weather Radar Systems text version

RADOME INFLUENCE ON WEATHER RADAR SYSTEMS PRINCIPLES AND CALIBRATION ISSUES Alexander Manz, Gematronik GmbH PO Box 210351, 41429 Neuss, Germany +49 2137 782-0, [email protected] 1. Introduction

Advanced modern dual-polarized weather radar systems have very high demands on performance and accuracy of gain, sidelobe levels and other electromagnetic performance parameters of the antenna system. Gain and beamwidth are used in the radar constant. Therefore their accuracies are an integral part of the measurement accuracy. The antenna system is enclosed by a radome. The radome itself will change the antenna gain and beamwidth. Therefore it is necessary to take the radome into consideration when the system calibration accuracy is calculated. In general the radome is considered to be a constant source of losses, but on a closer look the radome performance reveals a function depending on variables like antenna position and rain rate. A well designed radome can support the high performance and accuracy which is expected by a sophisticated high-gain, low-sidelobe antenna, but neglecting the radome can also cause significant measurement errors especially when rain increases the radome losses by several dB. The first step into radome calibration is to understand the radome effects. Radomes are affecting the antenna diagram through their scattering fields which can be added in the farfield. Scattering fields are determined by four major contributing effects : 1) transmission through the radome wall, 2) scattering by radome joints, 3) geometrical distribution of the joints on the radome sphere, 4) transmission through an additional water layer As a second step this presentation will discuss possible ways to include the radome effects into the weather radar calibration. 2. 2.1. Principles of Radome Effects Radome Wall Diffraction

The uniform radome wall represents 90%-96% of a radome surface. Therefore it has to be designed carefully. Any A-sandwich consists typically of three layers as demonstrated in Figure 1 : a) inside skin b) foam core c) outside skin (with hydrophobic coating)

electromagnetic ray Outside Skin Foam Core Inside Skin Hydrophobic Coating

Figure 1 : Principal Sandwich Cross Section The transmission of electromagnetic waves through any sandwich can be calculated as the diffraction of electromagnetic energy at boundaries between areas of different dielectric properties. This is a well researched theory and described many times like in [1] or [2]. Based on the knowledge of the electromagnetic properties of the materials like dielectric constant and loss factor it is possible to develop a transmission matrix T for every single boundary depending on the incidence angle, frequency and polarization of the electromagnetic wave and to calculate the transmission loss. Figure 2 shows the one-way transmission loss for an electromagnetic wave over frequency at 0°, 20° and 40° incidence angle and linear TE polarization for a typical A-sandwich designed

for C-band weather radar applications. Measurements have shown a good agreement between theory and reality.

Transmission in dB 0 TE-wave @ 0° 0.1 TE-wave @ 20° 0.2 TE-wave @ 40° 0.3 0.4 0.5 Frequency in GHz 0 1 2 3 4 5 6 7 8 9 10

Figure 2 : Transmission of A-Sandwich The accurate calculation of the radome wall loss has to include an integration of the radome wall transmission over all possible incidence angles weighted by the antenna illumination function. The radome designer has to find the optimum transmission over the integrated wall transmission loss. The scattering by a carefully designed radome wall has only significance for the total transmission loss of the radome. Any other parameter, like boresight error, sidelobe level change etc., is dominated by the scattering of the joint, because away from boresight the scattering fields due to the radome wall are 20 to 30 dB below the scattering fields of the joint and therefore negligible. 2.2. Joint Scattering

The sandwich radome panels are connected through nuts and bolts at the joints. Joint areas are typically built up by several layers of skin material to ensure the structural strength of the radome in this region. Electromagnetically, joint areas are the significant scatterers, if the wall thickness is optimized. Figure 3 shows the typical cross section of a sandwich joint.

Figure 3 : Cross Section of Sandwich Joint Every single joint scattering is calculated by multiplication of the angle dependent Induced Field Ratio (IFR) and the normalized scattering farfield weighted by the antenna aperture illumination. The complete scattering farfield of all illuminated radome joints is the sum of all single contributions. IFR is defined as the ratio of the scattered fields of the joint to the radiation fields of a rectangular uniformly illuminated aperture with the dimensions of the joint. The IFR can be measured with nearfield measurements as theoretically derived by [3] with the measurement set-up described in [4]. The results are two complex numbers for the incident electrical field polarisation parallel and perpendicular to the joint. Theoretical conclusions and empirical observations show that the IFR parallel to the joint is generally larger than perpendicular. For any joint m the angle dependent IFR can be calculated as IFRm = IFR|| cos (m) + IFR sin (m) (with m inclination angle of the joint to the polarization)

2 2


The pattern of a uniformly illuminated rectangular aperture can be described with sin(x)/x. This has proven to be a good model for the normalized scattering fields. The assumption of uniform illumination can be assured by subdividing the joint into small enough pieces. The farfield of the joint m in spherical coordinates with the origin at the center is written in [5] as

Im = W mlm

sin(am ) sin(bm ) jk(xcossin+ysin+zcoscos) e am bm


with W m , lm as width and length of joint m kWm am= sin(`) cos(`+m) and 2


klm sin(`) sin(`+m) 2

From equation (2) one can see, that joints scatter most of the energy typically perpendicular to their own length because W m is in general much smaller than lm. In consequence, horizontal joints increase mainly the elevation sidelobes of the antenna as vertical joints affect the azimuth cuts. This will be graphically shown in the following paragraph. Figure 4 shows the scattering farfield of a typical sandwich radome joint with the nominal width of 10.16cm (4") and an angle of 15° between the plane wave front and the normal vector of the joint.


90° 120° 150°

incident wave

60° 30°



210° 240° 270° 300°


Figure 4 : Farfield of Joint at 15° to incident Wave with E parallel to Joint 2.3. Joint distribution

Due to its size the radome has to be subdivided into panels for production and transportation reasons. Out of an infinite number of possible solutions most of them can be assigned to one the following types : Igloo Orange Peel Quasi-Random The radome is subdivided into small regular pieces with either vertical or horizontal joints. The radome is subdivided into relatively large vertical pieces. with mostly vertical joints. The radome is subdivided into a number of irregular pieces without preferred joint direction.

Figure 5 shows typical examples of the radome types.

Quasi Random

Orange Peel Figure 5 : Different Types of Sandwich Radomes


The complete radome scattering summarizing all illuminated joints using equations (1) and (2) is Es(,) =


Am(xm,ym,zm) Im(,,xm,ym,zm,m) IFRm(m) , m



Am xm,ym,zm

antenna llumination of joint m location of joint m

spherical coordinates, origin at antenna vertex inclination angle to antenna polarization

The scattering energy of parallel joints is concentrated like an antenna array. The unwanted scattering energy is even higher if the polarization is parallel to the joints because of higher IFR values. Figure 6 shows the calculated scattering farfields of the three types of radomes for linear horizontal polarization assuming the same joint IFR and same c-band reflector antenna. Any difference is only due to the different geometries. The scattering farfields have been calculated for angles from ­4° to +4° degrees from boresight in elevation and azimuth.

Figure 6 : Scattering Farfields (random ­ orange peel - igloo) The scattering of an igloo-type radome has very strong peaks in azimuth and elevation cuts as expected. This has a high impact on the antenna sidelobes directed to the ground and therefore increases problems with ground clutter. The orange-peel-type has relatively low scattering for horizontal polarization as here the main joints are perpendicular to the incident electromagnetic field. Most of the energy is going to the azimuth direction and not into the ground sidelobes of the antenna. The quasi-random-type radome shows no preferred scattering direction because of the randomized location of the illuminated joints. The calculated scattering farfield has to be added in phase to the antenna farfield to get a worst case of the radome influence. By comparing unperturbed antenna to the antenna perturbed by radome scattering one can derive the radome performance parameters, i.e. the reduction of the mainlobe is the radome transmission loss, boresight error is the movement of the 3dB-points and sidelobe level change from the difference in both radiation patterns. An igloo-type radome increases the antenna sidelobes in azimuth as well as in elevation raising eventually up by several dB unacceptable for any system performance. The orange-peel-type radome shows no extreme increase in elevation sidelobes especially for horizontal polarization and may be a trade-off solution between price and performance. The only technically recommendablr solution for dual polarization weather radar is the random-type radome with low scattering at both polarizations and all pattern cuts. 2.4. Rain Effects and Hydrophobicity

Besides the radome itself a water layer on the radome causes a significant attenuation of the radiation, leading to an underestimation of the precipitation intensity. Theoretical considerations of this effect (e.g. [9], [10] and [16]) suffer from the lack of knowledge of the distribution of water on the radome surface due to wind and other effects. Experimental results (e.g.[6], [8], [11], [12], [13] and [17]) are not easily generalised because of the very different properties of different surface materials and the not reproducible state of aging of the radomes. The wide range of antenna sizes and wavelength reduces the comparability of experimental results, too. From a purely scientic point of view, hydrophobicity is the only accesible measure to quantify and compare rain effects. The following subparagraph will describe the physics of hydrophobicity and the theoretical implications. Measurement results are following in the next subparagraph. The high dielectric constant and high loss tangent of water at microwave frequencies causes very significant attenuation of any signal passing through even a very thin film of water. Since a radome provides complete elimination of precipitation from all the antenna surfaces, the system performance during rain will, apart from

local propagation changes, be determined entirely by the quantity and form of the water temporarily present on the radome outer surface. Assuming that the water on the radome surface is in the form of a locally uniform thin film or sheet, the effect can be calculated based on the diffraction theory as mentionend in 2.1. It is possible to show that even for relatively modest rain rates, the additional attenuation due to the water film can amount to several dB, which is unacceptable for many applications. To provide low additional radome attenuation during rain, the properties of the radome surface should ideally promote two conditions: 1. 2. Formation of the water on the surface into separate "beads" rather than a continuous film. Rapid shedding of the water from the surface.

These two conditions are closely related, so that if the surface is such as to allow water beading to occur, then this will directly encourage rapid run off of the water from the surface. The form of the water beading on a typical high quality hydrophobic surface is shown below in Picture 7.

Picture 7 : Water Droplets on a Hydrophobic Coating For a given amount of water per unit area on the radome surface, the attenuation due to the water will be very much less when the water is in the form of separate beads or drops (as shown above in Picture 7) than when it is in a continuous film. This is due to the fact that the attenuation for the former case will be caused by relatively weak scattering from the water drops which will have dimensions of only a very small fraction of a wavelength. The quality of a given hydrophobic coating, in terms of its ability to form any surface water into separate beads rather than as a continuous film, is determined by the detailed surface chemistry of the coating, but can essentially be quantified by a single parameter. This parameter is the "surface free energy" (with low values indicating high hydrophobicity), however a more directly measurable quantity, which may be easily directly related to the surface free energy, is the "contact angle". As shown below in Picture 8, the contact angle is the angle between the tangent to the water drop surface at its contact point, to the surface. High values of contact angle indicate high hydrophobicity.

Picture 8 : Definition of Contact Angle

3. 3.1.

Calibration Considerations Transmission Loss and Beamwidth Change (except rain effects)

The radome influence has to be included in the radar constant at two points. Radome Losses (or as Reduction of the Antenna Gain) Beamwidth Change

Radome Losses and Beamwidth Change can be calculated as shown in the preceeding paragraphs. In general, both are treated as constant factors without any error margin. The values given by the radome manufactures are mostly estimations based on rough formulas (like in [4]) with input data based worst case assumptions. In order to be completely accurate it has to be noted that the radome performance is always a function of several independent parameters as there are : antenna pointing direction radome age radome surface condition actual rain rate wind direction

The radome perfomance could be given as a range of possible values and this range could be theoretically introduced into the error margin of the system calibration, but there are some good arguments against this procedure except for rain effects. a) The radome perfomance is given as a worst case value and still extremly small compared with other contributions to the radar constant. A typical radome loss would be 0.2dB to 0.3 dB and a typical beamwidth change is in the order of 1%. Variation of above mentioned variables would result into changes of the radome loss by several hundreths of a dB and changes in beamwidth by several thousandth of a degree. Such uncertainties are not significant. This is of course not valid for the influence of rain loss as this can be in the order of several dB. Even the sophisticated calculations as mentioned above are still only rough estimates and not exact numerical models of the real electromagnetic. The same arguments extends naturally to the calculation at different variables. Ageing and Surface Conditions are very much subject to specific local environmental conditions as there are pollution, sand, salt water, temperature or other influences which are very hard to quantify.



With respect to these points it is recommandable to stay with common practise and consider the radome as a constant modification of antenna gain and beamwidth. 3.2. Additional Loss due to rain

3.2.1. Experimental Data The following data is an example of rain influence measurements. The antenna of a Gematronik C-band radar at the research center of the University of Karlsruhe, Germany, is coverd by a radome manufactured by Ticon. It was errected in summer 1993 and has a diameter of 6.7 m. The surface is covered by NGA Gelcoat which should provide a certain degree of hydrophobicity. The type of radome is used at several other C-band radars, e.g. by the german and the swiss weather services. (For more information on the radar see [14]). To measure the attenuation by a water layer on the radome under conditions as reproducible as possible, a highly stable constant wave microwave source was used mounted on a meteorological measuring mast 900 m apart from the radar. This source is known to be stable within 0.1 dB for hours. The radome attenuation was measured under conditions of artificial rain (produced by the fire brigade of the research center) as well as natural rain. The first measurements were performed with the dirty radome as it was after five years of operation. Than the radome was cleaned and the measurements were repeated. The cleaning should decrease the wettability of the radome and therefore reduce the attenuation. The intensity of the natural rain was measured by a distrometer which was sited about 10 m south from the radome. The equivalent rain intensity for the artificial rain was estimated by the knowledge of the throughput of

the nozzle and verified by a measurement with the above mentioned distrometer. A more detailed description of the experiment is given by [15]. The result is presented in picture 9.

The crosses denote measurements with a dirty radome, the circles correspond to measurements with a cleaned radome. The two lines represent the information given by the manufacturer of the radome for a repellent and a wetting radome.

Picture 9 : Measurements of the one way attenuation by a water layer on the radome due to rain The equivalent rain intensity of the artificial rain was roughly 350 mm/h during the measurement on the dirty radome and a little more than 300 mm/h when the experiment was repeated with a cleaned radome. These are very high rain intensities, but they can be reached for a few minutes in very heavy rain events. The one-way attenuation was up to 5.4 dB in the first and up to 4.9 dB in the second case. Visual observations showed that the water formed a close film with ligaments perpendicular to the flow direction in both cases. It is obvious that for very heavy rain effects the water repellent layer will fail to avoid a closed film. As a consequence, there is no significant difference with the two experiments. The crosses and circles in Figure 9 denote the maximum rain intensities of natural rain and the corresponding attenuation measured with the dirty and clean radome, respectively. For comparison the information given by the manufacturer of the radome for a new repellent radome and a wetting one are given, too. 3.2.2. Calibration Implications As can be seen as a direct result there was no significant profit of the cleaning of the radome compared with the manufacturers information, allthough a comparison of the measurement with the dirty and the clean radome is difficult because of the very different rain intensities. It has to be stated, that the radome was quite clean in the direction of the microwave source even before the cleaning. There were other directions where it had been dirtier. The main conclusion from the experiment is rather the fact that rain influence can only be estimated with a high degree of uncertainty. The radome manufacturers provide results from extensive testing in laboratory conditions, but rain attentuation is always depending on the specific rain rate, the wind conditions and the radome surface conditions. There is no numerical model available for this purpose and ­ even, if it were available ­ there is no way to get accurate online input data for the environmental parameters. At the moment rain loss is still not included in the radar constant, because of the uncertainty. One possible solution would be to permanently install a transponder or test signal transmitter in order to measure differences in the radome attenuation online and use this number in order to adjust the radar constant. This approach can be realised with a reasonable effort and is currently under research in a combined effort between the University of Karlsruhe and Gematronik.

4. Conclusion Radome scattering is a very complex theory and appropriate radome design for weather radar has to include many parameters like radome wall, joints and joint distribution. The results can be used as constant parts of the weather radar constant while uncertainties are negligible. Rain effects on radomes are still a very critical point and the actual test results are not satisfactory. More testing is necessary to incorporate correct measures of the rain attenuation on the radome into the radar equation. 5.

[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17]


Stratton, J.A., ,,Electromagnetic Theory", McGraw-Hill, 1941 Rulf, B., Robertshaw, G.A., ,,Understanding Antennas for Radar, Communications and Avionics", Van Nostrand Reinhold, New York, 1987 Rush, W.V.T., Hansen, J.A., Klein, C.A., Mittra, R., ,,Forward scattering from square cylinders in the resonance region with application to aperture blockage", IEEE Trans. on A.&P., Vol. AP-24, pp. 182-189, March 1976 Skolnik, M.I., ,,Radar Handbook", 2nd Edition, McGraw-Hill, 1980 Shavit, R., Smolski, A.P., Michielssen, E., Mittra, R., ,,Scattering Analysis of High Performance Large Sandwich Radomes", IEEE Trans. on Ant. and Prop., Vol. 40, No. 2, Feb. 1992 Kai-Chiang Chang, ,,System Performance in Rain in a Radome-Enclosed Environment", International Journal of Infrared and Milimeter Waves, Vol. 7, No. 2, pp. 267-289, 1986 ,,Accelerated Weathering Test for Radome Panels", ESSCO Test Report TR94-16, March 1994 Anderson, Iain. Measurements of 20-GHz Transmission Through a Radome in Rain. IEEE Transactions on Antennas and Propagation, AP-23 (1975) pp 619-622. Blevis, B. C. Losses Due to Rain on Radomes and Antenna Reflecting Surfaces. IEEE Transactions on Antennas and Propagation, AP-13 (1965) pp 175-176. Blevis, B. C. Rain Effects on Radomes and Antenna Reflectors. Design and Construction of Large Steerable Aerials. IEE Conference Publications No. 21 (1965) pp 148-152. Cohen, A. and A. Smolski. The Effect of Rain on Satellite Communications Earth Terminal Rigid Radomes. Microwave Journal 9 (1966) pp. 111-121. Effenberger, John A., Richard R. Strickland and Edward B. Joy. The Effects of Rain on a Radome's Performance. Microwave Journal 29 (1986) pp. 261-274. Giger, A. J. 4-gc Transmission Degradation Due to Rain at the Andover, Maine, Satellite Station. The Bell System Technial Journal, 44 (1965) pp. 1528-1533. Gysi, Hermann. Orographic Influence on the Distribution of Accumulated Rainfall with Different Wind Directions. Atmospheric Research 47/48 (1998), pp. 615-633. Löffler-Mang, Martin and Hermann Gysi. Radome Attenuation of C-band Radar as a Function of Rain Characteristics. Annalen der Meteorologie 34 (1998). Setzer, David E. Computed Transmittion Through Rain at Microwave and Visible Frequencies. The Bell System Technical Journal (1970) pp 1873-1892. Weigand, R. M. Performance of a Water-Repellent Radome Coating in an Airport Surveillance Radar. Proceedings of the IEEE 61 (1973) pp. 1167-1168.


Radome Effects on Weather Radar Systems

8 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate


You might also be interested in

Radome Effects on Weather Radar Systems