United States Office of Radiation Programs ORP/EAD 78-5 Environmental Protection Las Vegas Facility June 1978 Agency PO. Box 15027 Las Vefeas NV89114 Radiation *xEPA Population Exposure to VHP and UHF Broadcast Radiation in the United States ------- Office of Radiation Programs Technical Publications Nonionizing Radiation Publications of the Office of Radiation Programs are available from the National Technical Information Service (NTIS), Springfield, VA 22161. Current prices should be obtained directly from NTIS using the indicated NTIS Order number. Single copies of some of the publications listed below may also be available without charge from the Office of Radiation Programs (AW-461), 401 M St., SW Washington, DC 20460. EPA ORP/SID 72-3 EPA/ORP 73-2 EPA-520/2-73-001 / EPA-520/1-74-005 EPA-520/2-74-008 ORP/EAD 75-1 ORP/EAD-76-1 ORP/EAD-76-2 EPA-520/2-76-008 ORP/EAD-77-2 ORP/EAD-77-3 Reference Data for Radiofrequency Emission Hazard Analysis (NTIS Order No. PB 220 471) Environmental Exposure to Nonionizing Radiation, (Available NTIS only, Order No. PB 220 851) Nonionizing Measurement Capabilities: State and Federal Agencies (Available NTIS only, Order No. PB 226 778/AS) RF Pulse Spectral Measurements in the Vicinity of Several ATC Radars (NTIS Order No. PB 235 733) An Evaluation of Satellite Communication Systems as Sources of Environmental Micro- wave Radiation (NTIS Order No. PB 257 138/AS) An Analysis of Broadcast Radiation Levels in Hawaii (NTIS Order No. PB 261 316/AS) Radiation Characteristics of Traffic Radar Systems (NTIS Order No. PB 257 077/AS) A Measurement of RF Field Intensities in the Immediate Vicinity of an FM Broadcast Station Antenna (NTIS Order No. PB 257 698/AS) An Examination of Electric Fields Under EHV Overhead Power Transmission Lines (NTIS Order No. PB 270 613/AS) An Investigation of Broadcast Radiation Intensities at Mt. Wilson, California (NTIS Order No. PB 275 040/AS) An Analysis of Radar Exposure in the San Francisco Area (NTIS Order No. PB 273 188/AS) ------- POPULATION EXPOSURE TO VHF AND UHF BROADCAST RADIATION IN THE UNITED STATES Richard A. Tell Edwin D. Mantiply June 1978 U.S. Environmental Protection Agency Office of Radiation Programs Electromagnetic Radiation Analysis Branch P.O. Box 15027 Las Vegas, Nevada 89114 USA ------- DISCLAIMER This report has been reviewed by the Office of Radiation Programs, U.S. Environmental Protection Agency, and approved for publication. Mention of trade names or commercial products does not constitute endorsement or recommendation for their use. 11 ------- PREFACE The Office of Radiation Programs of the U.S. Environmental Protection Agency carries out a National program designed to evaluate population exposure to ionizing and nonionizing radiation, and to promote development of controls necessary to protect the public health and safety. This report presents the latest estimates of population exposure to radiofrequency radiation determined by this agency. Readers of this report are encouraged to inform the Office of Radiation Programs of any omissions or errors. Comments or requests for further information are also invited. Floyd L. Galpin, Director Environmental Analysis Division Office of Radiation Programs 111 ------- ABSTRACT The U.S. Environmental Protection Agency has been collecting broadcast signal field intensity data for over two years to estimate population exposure to this form of nonionizing radiation. Measurement data have been obtained at 373 locations distributed throughout 12 large cities and collectively represent approximately 11,000 measurements of VHF and UHF signal field intensities. The VHF and UHF broadcast service is the main source of ambient radiofrequency exposure in the United States. A computer algorithm has been developed which uses these measurement data to estimate the broadcast exposure at some 39,000 census enumeration districts within the metropolitan boundaries of these 12 cities. The results of the computations provide information on the fraction of the population that is potentially exposed to various intensities of radiofrequency radiation. Special emphasis has been placed on determining the uncertainty inherent to the exposure estimation procedure and details are provided on these techniques. A median exposure level (that level to which half of the population is exposed greater than) of 0.005 yW/cm2 time averaged power density has been determined for the population of the 12 cities studied, the cumulative population of which represents 18 percent of the total United States population. The data also suggest that approximately 1 percent of the population studied, or about 380,000, are potentially exposed to levels greater than 1 yW/cm*, the suggested safety guide for the population in the USSR. Alternative techniques of using the measurement data to estimate population exposure are examined and future extensions of this work are discussed. IV ------- TABLE OF CONTENTS Page ABSTRACT . iv LIST OF FIGURES vi LIST OF TABLES viii BACKGROUND 1 METHOD OF MEASUREMENTS 3 APPROACH USED TO DETERMINE POPULATION EXPOSURE 5 MODELING METHOD 9 POPULATION EXPOSURE RESULTS 12 DIRECT ESTIMATION METHOD 23 CONCLUSIONS 25 FUTURE WORK 27 REFERENCES 28 ------- LIST OF FIGURES Page Figure 1. Figure 2. Figure 3. Figure 4. Figure 5. Figure 6. Figure 7. Figure 8. Figure 9. Figure 10. Figure 11. Figure 12. Figure 13. Figure 14. Measured FM radio broadcast field intensity spectrum in Portland, Oregon Accumulative fraction of population in Boston exposed £ log S(yW/cm2) Accumulative fraction of population in Atlanta exposed <_ log S(yW/cm2) Accumulative fraction of population in Miami exposed <_ log S(yW/cm2) Accumulative fraction of population in Philadelphia exposed £ log S(yW/cm2) Accumulative fraction of population in New York exposed £ log S(yW/cm2) Accumulative fraction of population in Chicago exposed <, log S(yW/cm2) Accumulative fraction of population in Washington exposed <. log S(yW/cm2) Accumulative fraction of population in Las Vegas exposed <. log S(yW/cm2) Accumulative fraction of population in San Diego exposed <. log S(yW/cm2) Accumulative fraction of population in Portland exposed ^ log S(yW/cm2) Accumulative fraction of population in Houston exposed <. log S(yW/cm2) Accumulative fraction of population in Los Angeles exposed <. log S(yW/cm2) Accumulative fraction of population in 12 cities exposed <. log S(yW/cm2) 15 15 16 16 17 17 18 18 19 19 20 20 21 VI ------- Figure 15. Figure 16. LIST OF FIGURES (Continued) Distribution of uncertainties in exposure calculations Site exposure and population exposure in Los Angeles Paqe 21 24 vii ------- Table 1. Table 2. Table 3. Table 4. LIST OF TABLES Measurement system uncertainties in VHF and UHF broadcast bands Summary of information relevant to environmental RF and MW field studies Population exposure results in 12 cities Summary of exposure test program results 22 22 viii ------- BACKGROUND The United States (US) Environmental Protection Agency (EPA) is presently gathering information pertinent to the development of guidance to Federal agencies within the US concerning limitations on radiofrequency (RF) and microwave (MW) exposure of the general population. This information consists of both detailed descriptions of the biological effects of RF and MW energy in experimental test animals and man, and normally encountered environmental exposure levels throughout the country. This report provides detailed information on the results of our environmental measure- ments program and presents our most current estimates of population exposure based on these measurement data. It is pertinent to describe the general approach used by the USEPA in collecting these data; in the first instance, numerous and widely distributed measurement points, generally selected on the basis of population distributions, located throughout many US high density metropolitan areas have been used to determine ambient exposure levels of RF and MW energy. These measurement data are then used in conjunction with a computer automated algorithm which contains census data to provide estimates of the fraction of the studied population exposed to various intensities of RF and MW radiation. Via this method, good estimates of exposure of most of the population are obtainable. In the second instance, many field intensity measure- ments are conducted without regard to population distributions but rather from the viewpoint of determining the maximum or highest intensities of exposure that are possible to be found in the environment. The principle purpose of this report is to provide the results of our efforts in the first instance, but to the extent that the secondly described measurement approach provides relevant exposure data, we will discuss these "specific source" types of measurements. 1 ------- Previous discussions of USEPA activities in this area are available (Janes, et al., 1977a; Janes, et al., 1977b; Athey, et al., 1978). This report contains new and more extensive data and results for US cities and uses an improved propagation modeling technique for generating estimates of population exposure. Additionally, a technique is discussed which provides insight to the consideration of the accuracy with which exposure estimates are obtained. ------- METHOD OF MEASUREMENTS Detailed discussions of the development of a specially instrumented mobile electromagnetic radiation analysis van used in the collection of the environmental exposure data are available elsewhere (Tell, et al., 1976a). The instrumentation approach involves spectrum analysis techniques coupled with on-line computer assisted data acquisition for purposes of recording, correcting, and processing of the acquired spectral intensity data. A series of calibrated antenna systems appropriate to the frequency bands of primary consideration are used to provide signal input to the spectrum analyzer. Appropriate account is taken for the polarization of the impinging waves in certain bands by the use of orthogonal dipolar antenna systems. The mini-computer system provides various features including signal averaging whereby fluctuating signal amplitudes are processed to obtain time-averaged values of field intensity, and the capability to retain instantaneous peak signal intensity excursions during the overall observation period. Extensive efforts resulted in our ability to specify the measurement system uncertainties as outlined in Table 1. It is noted that the mobile measurement system has been designed to principally operate in the bands assigned to domestic broadcasting within the US; this was done because of the generally higher environmental levels of RF and MW energy being the result of the broadcast service. Several changes in the mobile measurement system are currently underway which include a new super-broadband antenna system capable of a flat response over the 50-900 MHz region and a spectrum analysis system which will result in an enhanced capability for measurement of pulsed, radar field intensities. ------- Use of more portable instrumentation has been made in different studies of unique exposure situations, such as the main beam illumination of tall buildings and other locations not generally accessible by the mobile van system. Some of this instrumentation, the applicable studies involving its use, and discussions of accuracy limitations have been described in previous reports (Tell and Nelson, 1974a, 1974b; Tell and O'Brien, 1977; Tell, 1976; Tell, 1978) . TABLE 1. MEASUREMENT SYSTEM UNCERTAINTIES IN VHP AND UHF BROADCAST BANDS Band Frequency Range (MHz) RMS System Error(dB) Low VHP TV 54- 88 2.5 FM Radio 88-108 2.1 High VHP TV 174-216 2.3 UHF TV 470-806 2.0 ------- APPROACH USED TO DETERMINE POPULATION EXPOSURE The method used for our assessment of population exposure incorporates (a) identification of sites representative of the population distribution in a given metropolitan area, (b) measure- ment of the ambient field intensities existing at these representative sites, and (c) subsequent use of a model, to estimate the exposure that would have been measured at many other locations throughout the city. The results of this modeling phase are then analyzed to determine the fraction of the population potentially exposed to different intensities of RF and MW radiation. An important, underlying factor in our approach is the availability of detailed census data for the entire US suitable for machine processing. These census data, based on the 1970 census of the US, represent the number of persons residing in specific geographical cells called Census Enumeration Districts (CEDs) and the geographical coordinates of the centroid of each CED. A CED is a relatively small geographic area, consisting of, for example, a few city blocks within densely populated areas such as cities, but may be larger in rural regions wherein the population is more sparsely distributed. The entire US population is contained within some 257,000 such CEDs. We have developed a method for selecting environmental measurement sites which are representative of the population within a city. First, general boundaries are defined for a city which include essentially all of the metropolitan area population and all corresponding CEDs within these boundaries are then selected for subsequent processing from the overall census data base. In effect, each of these CEDs is assigned a weighting factor, according to the population within each CED. We then use 5 ------- a random process to select any desired number of these CEDs to use as measurement sites. Thus, we use a technique which incor- porates an equal likelihood of choosing any particular CED, except that those CEDs having a greater population are weighted in such a way as to increase their chance of being selected as a measurement site. Out of this process, we obtain those sites which are deemed to be most representative of the total city population. Field measurements are then accomplished at each of the selected sites, usually between 30 and 40, from which subsequent propagation models are generated. In addition to these sites, selected irrespective of RF and MW source locations, a few measurement sites are also included very near to selected trans- mitters to ensure a comprehensive approach to defining the full range of environmental levels. Field measurements are then performed at each selected site using the aforementioned mobile measurement van. This field activity is normally accomplished during an intensive two-week period of time. The actual measurement process is performed by situating the measurement van at a specific stationary location. No attempt is routinely made to evaluate standing wave phenomena in the vicinity of each measurement site and thus seek out either maximum or minimum field intensities which are characteristically present in such measurements. The extent to which such immediate location variability affects the resulting measurements is reflected in the scatter of the final data and is inherent in the variance with which we subsequently predict field intensities via a model. The results presented in this report are the product of USEPA field measurements conducted in 12 US cities which include in the order that they were studied Boston, Atlanta, Miami, Philadelphia, New York, Chicago, Washington, Las Vegas, San Diego, Portland, Houston, and Los Angeles. The total population studied in these 12 cities is 38,144,845 and includes 38,548 ------- CEDs yielding a mean population per CED of 990 persons. From these field studies, approximately 11,000 individual signal field intensities were determined from a total of 373 measurement sites. Figure 1 illustrates the type of field intensity data collected; in this case the spectral data show one of the measure- ments of the FM broadcast band obtained in Portland. Here each spectral peak observed is a single FM radio station signal. In this particular case the measurement site was very near to a multiple broadcast transmission center and the measured power density was 14 yW/cm2. Table 2 summarizes the relevant informa- tion pertaining to each city investigated. FM AVERAGE FIELD STRENGTH ... TOTAL rOUER OEMS!TV (FOft IMTCMftTIOM ! T MMC 0* M 0»> 14. 212« UM/CN/CN 7» »» IN MCQUKNCV <«WZ> 25 MANS SM OHTM POJMTS 1189 MOWS 4 HIM IT 317 MVS 1»7( M - ! KMZ fITC CO»E- « OPMftTOft COW- 7 1M Figure 1. Measured FM radio broadcast field intensity spectrum in Portland, Oregon ------- TABLE 2. SUMMARY OF INFORMATION RELEVANT TO ENVIRONMENTAL RF AND MW FIELD STUDIES oo Boston Atlanta Miami Philadelphia New York Chicago Washington Las Vegas San Diego Portland Houston Los Angeles #CEDs 2003 1249 1897 3606 11470 4646 2291 356 1113 1194 1127 7596 Population 1953665 1221431 1661012 3407059 12269374 4743905 2516917 264501 1071887 818040 1265933 6951121 Number of Stations # of Field Low High Strength Values FM VHF VHF UHF # of Total Sites 252 396 448 941 1426 1378 1107 632 956 816 810 1801 14 11 13 17 23 20 17 6 17 12 14 29 3 2 3 2 3 2 2 2 1 3 1 3 1 2 2 2 4 3 2 3 2 3 3 4 3 3 2 3 3 3 3 0 2 0 2 7 21 18 20 24 33 28 24 11 22 17 20 43 9 16 16 31 36 39 37 42 38 38 33 38 TOTAL 38548 38144845 10963 193 26 31 31 281 373 ------- MODELING METHOD Athey et al., (1978) described a method whereby the actual measurement data were used to modify a presumptive propagation model for calculation at all CED sites throughout a city. Athey"s report made use of a propagation model form which was obtained by analyzing measured field intensity data obtained in Miami which suggested a classically recognized decrease in electric field intensity with increases in distance between FM broadcast stations and measurement sites. This form for the model was then applied to data obtained in all VHP and UHF broadcast bands to determine exposure. In the present case, we have developed an enhanced method for predicting exposure at the various CEDs by taking into consideration the fact that each city and individual stations possess their own distinctive propagation characteristics. The method we have used includes the following features. For each station under consideration, the field intensity obtained for the station at each measurement site is used to obtain a linear, least squares fit of the data. This provides a functional form describing the way by which the electric field strength varies as a function of distance from the station. Since this model is generated from actual measurement data for each station, note that no specification of transmitter power or antenna height is necessary. If, by chance, because of poor data, i.e., high variability in measured values of field strength, the resulting computed slope of the least squares fit is positive, the slope is changed arbitrarily to be equal to zero. This in general is not a common problem, occurring in only 12 instances for the entire set of measurements reported. Next, the straight line model is used to calculate the field intensity which would be expected at each CED within the cities' bounds. From extensive tests we 9 ------- determined that maximum accuracy was usually obtained in the modeling procedure by using the predetermined slope of the line model but shifting this line model vertically to form a least squares fit with the measurement data obtained in the neighbor- hood of the calculational point Ca CED location). We observed that this shifting process was effective in reducing the uncer- tainty whenever the particular station was closer than 5 km to the CED. Thus we incorporated this feature of appropriately shifting the line model to best fit the measurement data obtained at the two nearest measurement sites. Tests revealed a non- significant reduction in uncertainty by shifting the model to best fit more than the two nearest sites. The effect of this process is to lend weight to the local measurement data in improving estimates primarily of high intensity exposures. It was found that the shifting technique produced little, if any, apparent improvement in other than the higher exposure levels. If it occured in the calculational process that a CED was identi- fied as being closer than 100 meters from a nearby station, then the actual distance was arbitrarily changed to correspond to 100 m. This was accomplished to protect against the erroneous computation of very high exposure levels when the CED - station distance was very short. An important feature in the development of our work was the construction of a test program which could be used to estimate the uncertainty associated with the modeling method. In lieu of performing additional measurements to examine the accuracy of the method, we elected to make use of the metropolitan area measure- ments themselves in a special way. The process consists of starting at one specific measurement site where data has been obtained and then creating the least squares line model for each station based on the measurements obtained at all other measure- ment sites, but not including the site under test. The exact calculational process described above is then used, always rejecting any data obtained at the test site, to arrive at the 10 ------- estimated field strength for each station. Then, a direct comparison is made between the predicted field and the field strength actually measured at the site. This is accomplished for each station involved and in addition to individual signal field strength differences, a comparison is made between the predicted total power density of exposure and that actually measured and being the result of exposure from all signals present at the site. The process is then repeated at each other measurement site to obtain an indication of the goodness of the modeling procedure. Once the process has been completed for all measure- ment sites in a city, the results are assessed statistically by determining the mean deviation between actual and predicted field strengths and the mean deviation between actual and predicted total power densities of all signals. These results are then used as an indicator of the quality of the more comprehensive calculations performed at all CEDs within a city. Undoubtedly, the variances of the deviations apparent in this process are partly due to the immediate location variability discussed previously. Longely (1976) has discussed this subject in detail. Repeated application of the test program, using different criteria for shifting, provided the insight by which the final modeling criteria were determined. Extensive computer time was spent before arriving at the optimum criteria. 11 ------- POPULATION EXPOSURE RESULTS The aforementioned modeling method was applied to the measurement data obtained in each of the 12 cities. Exposure levels were computed at each CED location and the resulting exposure was assumed to apply to all of the population associated with each CED. After calculation of the exposures the number of persons associated with various ranges of intensities were determined; in particular, approximately one-third decadic power density ranges were used to classify exposure, i.e., 0.001, 0.002, 0.005, 0.010, 0.020, 0.050, 0.100 yW/cm2, etc. The final results of the analysis are presented in terms of the accumulative fraction of the population which are potentially exposed equal to or less than these different one-third decadic power density intervals. Results for each of the cities under study are presented in Figures 2-13 wherein the exposure level is plotted logarith mically and the population fraction follows a near normal distri- bution. Figure 14 provides the results for all cities taken together. Each figure provides the population exposure determined for each band separately and for all measured bands together. The results suggest that the exposure levels are approximately normally distributed and reveal the interesting finding that of the exposure contributed by the various VHF and UHF broadcast bands, the FM radio broadcast band is clearly discernable as being most responsible for overall exposure, particularly at the highest exposure levels. This finding supports the earlier proposition offered by Tell and Janes (19751 implicating FM radio broadcast transmissions as generally dominant in creating the highest ground levels of RF fields. Despite the lower effective radiated powers authorized for FM broadcasting compared to other VHF and UHF television emissions, a combination of relatively low 12 ------- tower heights and broad vertical antenna radiation patterns for FM transmission conspire to produce these relatively high fields. It is also interesting to note the relatively low contribution provided by the UHF TV band in as much that UHF television stations in the US carry the maximum power authorizations. In our experience we have found it informative to discuss these results using two different indices. The first is the median exposure level, i.e., that power density at which 50 percent of the population are exposed less than and 50 percent are exposed greater than. The second is the measure of the fraction of the population potentially exposed above 1 yW/cm2. The data for total band exposure presented in Figures 2-14 have been summarized from the point of view of these two indices in Table 3. The most significant results are for the accumulative population of all the cities in which a median exposure of 0.005 yW/cm2 was determined while something less than 1 percent , of the population are apparently exposed at intensities greater than 1 yW/cm2. It is worthy to reemphasize that these data apply only to the domestic broadcast service in the US and cannot account for population mobility. Though the population data base itself is dated, we feel that the results are probably representa- tive for the actual present distribution of population. The results of the test program designed to estimate the uncertainty associated with exposure calculations are presented in summary form for the 12 cities in Table 4. The tabulated data refer to the average of all individual field strength deviations and power density deviations at all measurement sites within each city. The observed high deviation in power density calculations in Boston undoubtedly reflects the few measurement sites used in that study. 13 ------- In order to assess the uncertainty in our overall estimates of population exposure for all cities studied to date, Figure 15 was prepared which provides the frequency of occurrence of deviations between measured and calculated values of exposure at all 373 sites visited. Figure 15 shows that the distribution of these uncertainties is approximately chi-squared in nature suggesting that the population of power densities from which these determinations were obtained is normally distributed, this being in consort with the general appearance of Figure 14. The most significant point of Figure 15 is that the most likely uncertainty appears to be about 3dB while 70 percent of all our exposure calculations are within 8dB. 14 ------- flCCUMULflTIVE FRACTION OF POPULBTION EXPOSED flS fl FUNCTION OF POMER OCNSITV 99 93 . 9- . 8-- 7- 3 - 2- .1-- 91 BOSTON h- -3 ~^Z 37 TOTAL FH WHO LOM VHP TV o HIQN VMF TV * OMf TV -f- f -5 -4 LOQRR1THM OF POMER DENSITV IN UM/CH' Figure 2. Accumulative fraction of population in Boston exposed <. log S(yW/cm2) ACCUNULBTIVE FRACTION OF POPULATION EXPOSED flS R FUNCTION OF POWER DENSITV 1 * ". 99^- ? » 99 .9 . 6 . 7- :|: . 4- . J- . a- . i- 89 81 ATLANTA * . * * 1 . ' - * r ° . * » » .. " . * " ' A . Torm. LOU VHP TV o HIOH VHP TV * UMF TV 1 1 1 1 , 1 1 -3 -4 -2-2-1 8 1 2 LOGARITHM OF POWER DENSITV IN UH/CM2 Figure 3. Accumulative fraction of population in Altanta exposed <. log S(yW/cm2) 15 ------- ACCUMULATIVE FRACTION OF POPULATION EXPOSED- AS A FUNCTION OF POWER DENS I TV 99 ei MlftlK mm. - m «noio LOU VMF TV ° HIOH VMF TV * UHF TV -5 -4 -3 -2 -1 0 LOGARITHM OF POWER DENSITV IN UWVCH* Figure 4. Accumulative fraction of population in Miami exposed <. log S(yW/cm ) ACCUMULATIVE FRACTION OF POPULATION EXPOSED AS A FUNCTION OF POUER DENS1TV 99 95 .9- . «-- 7- 3- .2- . 1- 85 01 * TOTW. * rn RADIO LOW VH» TV MIOH VHP TV * IMf TV 1 1 1 1 \ 1 -3 -4 -3-2-1 8 1 2 LOGARITHM OF POWER DENSITY IN UW/CM2 Figure 5. Accumulative fraction of population in Philadelphia exposed ^ log S(yW/cm2) 16 ------- ACCUMULATIVE FRACTION OF POPULATION EXPOSED AS fl FUNCTION OF POWER DENSITV .99 . 93 . 9 . 9- I . 3 . 2- . 1- 95 NEW VO«K TOTBL rn «noiO LOU VHP TV o NIQM VHP TV * UHF TV -3 -4-3-2-1 8 12 LOGARITHM OF POWER DENSI TV IN UH/CM2 Figure 6. Accumulative fraction of population in New York exposed £ log S(yW/cm2) ftCCUMULRTIVE FRftCTIOH OF POPULATION EXPOSED AS A FUNCTION OF POWER DENSITV 99 95 o o » 74- es torn FH WPIO LOU VHT TV ° HtOH VMF TV UMf TV -3 -4 -3-2-1 0 i 2 LOGARITHM OF POWER DENSITV IN UW/CH2 Figure 7. Accumulative fraction of population in Chicago exposed £ log S (yW/cm2) 17 ------- ACCUMULATIVE FRACTION OF POPULATION EXPOSED AS A FUNCTION OF POWER DENSITY 99 95 . 9- 7-- 3 . 2- . 1- 05 01 MftfHlffOTOM 8 TOTAL FH RADIO LOW VHF TV o HIQH VHF TV » u«r TV -5 -4 -3 -2 -1 0 LOQARXTHH OF POWER DENSITY IN UM/CIT Figure 8. Accumulative fraction of population in Washington exposed ^. log S(yW/cm2) ACCUHULflTIVE FRACTION OF POPULATION EXPOSED AS A FUNCTION OF POWER DENSITY 99 95 . 9-- .8 7- 3 2- . ! 85 91 LHS TOTHL LQN VHF TV VMf TV TV -f- -3 -4-3-2-1 0 1 2 LOQARITHH OF POWER DENSITY IN UW/CK2 Figure 9. Accumulative fraction of population in Las Vegas exposed <. log S(yW/cm2) 18 ------- ACCUMULATIVE FRACTION OF POPULATION EXPOSED AS A FUNCTION OF POWER DENSITV 99 95 . 9 .8-- 7-- 3 . 2-- . 1 8s 81 sun orcoo F» RADIO LOW VMf TV o Ml OH VHF TV » WKF TV -3 -4 -3-2-1 8 1 2 LOGARITHM OF POWER OENSITV IN UW/CH2 Figure 10. Accumulative fraction of population in San Diego exposed £. log S(yW/cm2) ACCUNULflTIVE FRACTION OF POPULATION EXPOSED AS A FUNCTION OF POWER DENSITV . 99 .95 . 9 .8-- . 7-- . 3- .2- . 1- 85 81 PMTLANt TOTW. ' rn MOID LOU VHT TV o MIO« VMf TV UHF TV 4- -3 -4 -3-2-1 8 1 2 LOGARITHM OF POWER DENSITY IN UU/CH* Figure 11. Accumulative fraction of population in Portland exposed ± log S(yW/cm2) 19 ------- ACCUMULATIVE FRACTION OF POPULATION EXPOSED AS A FUNCTION OF POWER DENSITV 99 93 . 9 + . 8 it . 4- .3- .2- .!- 09 81 HOUSTON TOTflt * PM MO 10 LOW VMF TV ° MIOH VMF TV UMT TV -5 -4 -2-2-1 6 1 2 LOGARITHM OF POWER DENS!TV IN IW/CM2 Figure 12. Accumulative fraction of population in Houston exposed <. log S(yW/cm2) ACCUMULATIVE FRACTION OF POPULATION EXPOSED AS ft FUNCTION OF POWER DENSITY 99 95 n . 3- .2-- . 1 es ei LOS * * TOT«L FH LOH V«F TV oHIOH VMP TV *UHP TV -f- -3 -4 -3-2-1 8 1 2 LOGARITHM OF POWER DENSITY IN UU/CH* Figure 13. Accumulative fraction of population in Los Angeles exposed <. log S(yW/cm2) 20 ------- flCCUIIULRTIVE FRRCTION OF POPULflTION EXPOSED flS R FUNCTION OF POWER DEMSITV . 99 ,95 . 9 . 8 . 7 . 4- . 3 .2- . i 05 01 MSTOM N1ANI fHILADCLPHtn NCM VORK CMICAOO WASHINGTON LAS VC0AS SAN OICOO .PMUAN9 HOUSTON LOS flNKUS S « * o A TOTAL * FH flADIO LOU VHF TV oHIOH VHT TV *UM» TV ~5 ~4 ~i ~2 ""d. 0 i 2 LOQARITHH OF POWER OEHSITV IN UM/CH* Figure 14. Accumulative fraction of population in 12 cities exposed £ log S(yW/cm2) so.. 40.. (A 111 30 cc UJ m 20.. 10.. DISTRIBUTION OF UNCERTAINTIES IN EXPOSURE CALCULATIONS 373 SITES POWER DENSITY ERROR FOR POINTS NOT PLOTTED 33.2 33.7 34.4 38.5 42.7 49.1 66.5 . HIIIIIIIh HIh Hh^HA 01 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 UNCERTAINTY (dB) Figure 15. Distribution of uncertainties in exposure calculations 21 ------- TABLE 3. POPULATION EXPOSURE RESULTS IN 12 CITIES Boston Atlanta Miami Philadelphia New York Chicago Washington Las Vegas San Diego Portland Houston Los Angeles All cities Median Exposure (yW/cm2) 0.018 0.016 0.0070 0.0070 0.0022 0.0020 0.009 0.012 0.010 0.020 0.011 0.0048 0.0053 Percent of Population Exposed <1 juW/cm2 98.50 99.20 98.20 99.87 99.60 99.60 97.20 99.10 99.85 99.70 99.99 99.90 99.99 TABLE 4. SUMMARY OF EXPOSURE TEST PROGRAM RESULTS No. Mean Field Mean Power Density Sites Error (dB) Error (dB) Boston Atlanta Miami Philadelphia New York Chicago Washington Las Vegas San Diego Portland Houston Los Angeles 9 16 16 31 36 39 37 42 38 38 33 38 11.9 5.8 6.5 7.3 7.2 6.9 6.1 7.2 8.4 9.7 7.3 5.8 16.8 4.4 7.6 6.9 6.2 7.6 5.5 5.2 10.5 5.2 5.6 6.6 22 ------- DIRECT ESTIMATION METHOD Our choice of the population weighted random method for selection of CEDs as measurement sites was prompted by a desire to establish a consistent approach from city to city. In the beginning phases of the metropolitan area studies, measurement sites were not chosen on this basis but were decided upon by common sense and the apparent distribution of population as infered from city maps. An interesting observation from applica- tion of the computer selection method, however, is that if measurements are conducted at locations which are truly random in the population space, then a simple inspection of the measurement data according to sites should provide a direct assessment of population exposure in the general area. To illustrate this process, measurement sites corresponding to CEDs (most do) are sorted according to increasing power density and the accumulative fraction of sites are plotted against the logarithm of power densities on probability paper. Figure 16 provides an example of this method applied to data obtained in Los Angeles. From the data, which is seen to be almost perfectly log-normally distributed, one obtains a median exposure value of about 0.006 yW/cm2 which compares favorably with the most comprehensive method which necessitates many calculations at all CEDs in the area. Note that this method, after the initial site selection is completed, requires no further information on population. We have observed a generally good agreement between the two approaches in deter- mining population exposure, particularily near the median exposure values, and often utilize the direct method, in favor of its simplicity, to obtain preliminary estimates of results. 23 ------- SITE EXPOSURE AND POPULATION EXPOSURE IN LOS ANGELES 99.9. 99. 98. DC 0 « W QS UJO 9& to o UJ m sl80- is UJ Ul UJ t UJ Z -2 _J J ACCUMI POPL W 0 2- 1 0.5- H * > H * J * . ^^ / CALCULATED POPULATION I EXPOSURE . MEASURED CED SITE EXPOSURE > * _ 1 1 I 1 I 1 99.9 .99 .98 .95 .90 .80 50 .10 .5 2 1 0.5 -4 -3 -2-10 1 LOGARITHM OF POWER DENSITY IN uW/cm2 Figure 16. Site exposure and population exposure in Los Angeles 24 ------- CONCLUSIONS Results of the methods outlined here suggests that, of the population group studied, representing 18 percent of the total US population, a median exposure value of about 0.005 yW/cm2 time averaged power density exists and perhaps, more interestingly, less than 1 percent of the population are potentially exposed at levels above 1 yW/cm2. It is observed that the FM radio broadcast service is responsible for most of the continuous illumination of the general population. Indeed, that fraction of the population exposed beyond 1 yW/cm2 needs more careful definition and the absolute maximum intensities observed demand precise determina- tion, but it is interesting to note from our results that, even at this time, at least 99 percent of the population studied are not exposed to levels above the suggested level of safety estab- lished in the USSR of 1 yW/cm2 (Gordon, 1974). Additional data obtained by the USEPA, in special areas wherein mainbeam illumina- tion of tall buildings occur nearby various high power broadcast installations, has shown that it is difficult to find areas where intensities exceed 100 yW/cm2 (Tell, 1978). These data must be viewed from the standpoint of long term exposure and certainly, it is true that, on occasion, localized exposures may greatly exceed 1 yW/cm2. The authors recognize the case of limited time exposure of some individuals to microwave oven leakage, portable or mobile communication equipments, and various other sources of RF and MW exposure including pulsed sources, however, we feel that at this time, there do not exist adequate quantitative techniques for evaluating these more extreme exposure regimes in terms of their impact on our popula- tion exposure estimates provided in this report. It is our observation that these higher intensity situations must be 25 ------- addressed on the basis of the length of time spent in the field and will require an accentuated emphasis upon field measurements conducted from the viewpoint of determining absolute maximum exposure values that may be encountered such as inside building measurements. 26 ------- FUTURE WORK The evidence provided by the rather extensive environmental measurements program conducted by the USEPA within the US seems to overwhelmingly support the contention that most of the general population is not chronically exposed to high intensity (i.e., >100 yW/cm2) RF and MW radiation. Accordingly, future field measurement efforts will include to a greater extent examination of those unique kinds of exposure circumstances wherein relatively high intensity exposures are possible or expected. A more detailed investigation of environmental levels of pulsed RF and MW fields is currently being developed. Addition- ally, we are examining our data from the viewpoint of developing deterministic propagation models, provided transmitter effective . radiated power and antenna height, for different classes of transmitting stations. Our particular interest is in being able to more accurately model close-in exposure conditions, and in this connection we will be comparing our data and resulting propagation models with other existing models (Kalagian, 1976). 27 ------- REFERENCES Athey, T.W., R.A. Tell, N.N. Hankin, D.L. Lambdin, E.D. Mantiply, and D.E. Janes (1978): "Nonionizing Radiation Levels and Popula- tion Exposure in Urban Areas of the Eastern United States." Environmental Protection Agency Technical Report ORP/EAD-77-008, May. Gordan, Z.V. (ed.): Biological Effects of Radiofrequency Electromagnetic Fields, translated from Moscow 0 Biologicheskom Deystrun Elektromagnitnykh Poley Radiochastot in Russian No. 4, 1973. Available Through NTIS as JPRS Document 63321, October 30, 1974. Janes, D.E., R.A. Tell, T.W. Athey, and N.N. Hankin (1977a): "Radio-frequency Radiation Levels in Urban Areas," Special Supplement in biology to Radio Science, editors A.W. Guy and D.R. Justesen, SS-1 (in press) 1977. Janes, D.E., R.A. Tell, T.W. Athey, and N.N. Hankin (1977b): "Nonionizing Radiation Exposure in Urban Areas of the United States," Communication 304, accepted for presentation in Session No. S.07, Nonionizing Radiation, IVth International Congress of the International Radiation Protection Association, 1977. Kalagian, G.S. (1976): "Field Strength Calculation for TV and FM Broadcasting (Computer Program TVFMFS)." Federal Communications Commission Technical Report FCC/OCE RS 76-01, Washington, DC, January. Longley, A.G. (1976): Location Variability of Transmission Loss- Land Mobile and Broadcast Systems." Department of Commerce, Office of Telecommunications Report OT 76-87, May. Tell, R.A. (1976): "A Measurement of RF Field Intensities in the Immediate Vicinity of an FM Broadcast Station Antenna." Technical Note, ORP/EAD-76-2, U.S. Environmental Protection Agency, Silver Spring, MD, January (NTIS Order No. PB 257 698/AS)*. Tell, R.A. (1978): "Measurements of Radiofrequency Field Intensities in Buildings with Close Proximity to Broadcast Stations," Environ- mental Protection Agency Technical Note, April. Tell, R.A., and D.E. Janes (1975): "Broadcast Radiation: A Second Look." In Biological Effects of Electromagnetic Waves, ed. by C.C. Johnson and M.L.Shore,TSelected papers of the USNC- URSI 1975 annual meeting, Boulder, CO, October (.2 Volumes) , USDHEW Publication (FDA) 77-8011. 28 ------- Tell, R.A. and J.C. Nelson (1974a): "RF Pulse Spectral Measure- ments in the Vicinity of Several Air Traffic Control Radars." EPA Technical Report EPA-520/1-74-005, 45 pages, May. Tell, R.A., and J.C. Nelson (1974b): "Microwave Hazard Measure- ments Near Various Aircraft Radars." Radiation Data and Reports, Vol. 15, No. 4, pp. 161-179, April. Tell, R.A., N.N. Hankin, J.C. Nelson, T.W. Athey, and D.E. Janes (1976a): "An Automated Measurement System for Determining Environ- mental Radiofrequency Field Intensities II." In Proceedings of NBS symposium on Measurements for the Safe Use of Radiation. March 1976, NBS publication NBS SP456, (ed. S.P. Fivozinsky), pp. 203-213. Also presented at 1974 Meeting at USNC/URSI, October 14- 17, 1974, Boulder, CO. 29 ------- TECHNICAL REPORT DATA (Please read Instructions on the reverse before completing) 1. REPORT NO. 2. 3. RECIPIENT'S ACCESSION NO. 4. TITLE AND SUBTITLE POPULATION EXPOSURE TO VHF AND UHF BROADCAST RADIA- TION IN THE UNITED STATES 5. REPORT DATE June 1978 6. PERFORMING ORGANIZATION CODE 7. AUTHOR(S) Richard A. Tell and Edwin D. Mantiply 8. PERFORMING ORGANIZATION REPORT NO. 9. PERFORMING PRGANIZATION NAME AND ADDRESS U.S. Environmental Protection Agency Office of Radiation Programs Electromagnetic Radiation Analysis Branch P.O. Box 15027 Las Vegas, Nevada 89114 10. PROGRAM ELEMENT NO. 11. CONTRACT/GRANT NO. 12. SPONSORING AGENCY NAME AND ADDRESS 13. TYPE OF REPORT AND PERIOD COVERED Technical Note Same as above 14. SPONSORING AGENCY CODE 15. SUPPLEMENTARY NOTES The U.S. Environmental Protection Agency has been collecting broadcast signal kCJFi'L-tmtjj-Ly data £UL uvtji. twu yecLLtj tu est-uikitfc; population exposure to tills £01.111 16. of nonionizing radiation. Measurement data have been obtained at 373 locations distributed throughout 12 large cities and collectively represent approximately 11,000 measurements of VHF and UHF signal field intensities. The VHF and UHF broadcast service is the main source of ambient radiofrequency exposure in the United States. A computer algorithm has been developed which uses these measurement data to estimate the broadcast exposure at some 39,000 census enumeration districts within the metropolitan boundaries of these 12 cities. The results of the computa- tions provide information on the fraction of the population that is potentially exposed to various intensities of radiofrequency radiation. Special emphasis has been placed on determining the uncertainty inherent to the exposure estimation procedure and details are provided on these techniques. . A median exposure level (that level to which half of the population is exposed greater than) of 0.005 yW/on2 time averaged power density has been determined for the population of the 12 cities studied, the cumulative population of which represents 18 percent of the total United States population. The data also suggest that approximately 1 percent of the population studied, or about 380,000, are potentially exposed to levels greater than 1 yW/on , the suggested safety guide for the population in the USSR. Alternative techniques of using the measurement data to estimate population exposure are examined and future extensions of this work are discussed. 17. KEY WORDS AND DOCUMENT ANALYSIS DESCRIPTORS b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group 18. DISTRIBUTION STATEMENT Release to Public 19. SECURITY CLASS (This Report) Unclassified 21. NO. OF PAGES 20. SECURITY CLASS (This page) Unclassified 22. PRICE EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE ------- |