SPS Mapping of Exclusion Areas For Rectenna Sites DOE 1978

HCP/R-4024-10 Satellite Power System (SPS) Mapping of Exclusion Areas For Rectenna Sites October 1978 Prepared for: U.S. Department of Energy Office of Energy Research Satellite Power System Project Office Under Contract No. EG-77-C-01-4024 DOE/NASA Satellite Power System Concept Development and Evaluation Program

Satellite Power System (SPS) Mapping of Exclusion Areas For Rectenna Sites HCP/R-4024-10 Dist. Category UC-11,41,60,63, 63a,b,c,e,64,66e,95f,97c October 1978 Prepared by: James B. Blackburn, Jr. Bill A. Bavinger PRC Energy Analysis Company McLean, Virginia 22102 Prepared for: U.S. Department of Energy Office of Energy Research Satellite Power System Project Office Washington, D.C. 20545 Under Contract No. EG-77-C-01-4024 DOE/NASA Satellite Power System Concept Development and Evaluation Program

Acknowledgements Project Staff James B. Blackburn, Jr. - Principal Investigator School of Architecture Bill A. Bavinger - Co-Principal Investigator School of Architecture Arthur A. Few - Co-Principal Investigator - Space Physics Department Peter G. Rowe - School of Architecture J. V. Leeds - Electrical Engineering Department Research Assistants Abdullah Al-Homomeidy - School of Architecture Francis Essien - School of Architecture John Gaulden - School of Architecture Emilio Giralt - School of Architecture Kevin Newman - School of Architecture Sara Procianoy - School of Architecture Mottaz Salama - School of Architecture Perry Winston - School of Architecture Laboratory Assistants Jabir Al-Hilali - School of Architecture Tim Barry - School of Architecture Owen Richards - School of Architecture William Turner - School of Architecture Michael Welsh - School of Architecture Secretarial Christina Salas - Space Physics Department Special Acknowledgements Special acknowledgement is extended to Allan D. Kotin, Economic Consultant, Los Angeles, California, with whom we worked in developing many aspects of this report. His work is reported in a separate document on Resource Requirements For the Satellite Power System. Thanks are also extended to Wayne Smith and John Freeman of the Rice University Space Solar Power Research Program. Also, a special thanks to Joan Neagli, of Neagli's Secretarial Service for typing this report. The authors wish to acknowledge the following people for the time taken to review the draft of this paper and prepare written comments on it: T. Stephen Cheston of the Georgetown University Graduate School, Paul E. Custer of the National Science Foundation and Dr. James W. Moyer of Southern California Edison Co.

Digitized by the Space Studies Institute ssi.org

Executive Summary This report sets forth the results of a study by Rice University to determine those areas of the United States that were not available as potential sites for receiving antennas that are an integral part of the Satellite Power System (SPS) concept. Under the current SPS program, 60 satellite-rectenna pairs would be developed. Each pair would produce 5 gigawatts of power and the rectenna would require the dedication of approximately 50,000 acres of land per site. Therefore, 60 sites of 50,000 acres each, totalling approximately 3 million acres, will be required. This study's approach to finding where, or even if, 60 such sites existed was to determine those areas of the United States where the rectenna could not be sited. 36 variables with the potential to exclude the rectenna were mapped and coded into the Rice University Computer System. Some of these variables absolutely exclude a rectenna from locating within the area of its spatial influence, and other variables potentially exclude the rectenna. These maps of variables were assembled from existing data and were mapped on a grid system of the United States. Each grid square was 26 km on a side. The analysis of the information was completed by utilizing overlay or sieve analysis. Under this approach, variables were laid over other variables and the composite of this union of variables would represent areas where the rectenna could not be located. This report shows, in Section IV, 11 summary maps that indicate the land areas excluded as rectenna sites under various combinations of variables. The areas in "white" are not excluded as sites and are considered as "eligible" areas. It is important to note that the only interpretation to be given to these eligible areas is that they were not ruled out as sites. The areas should be studied in more detail to determine where rectennas could be located within these subset areas. These various summary maps go from being less rectrictive to being more restrictive with respect to sites. Under Summary Map 1, approximately 50% of the United States was excluded as potential sites, with Summary Map 8 excluding 73% of the United States and with Summary Map 9 excluding 83%. Each summary map is accompanied by a detailed statistical analysis which describes the "eligible" areas on a state by state basis with respect to other variables not directly utilized in the creation of the summary maps. Due to the complex nature of siting studies such as this, the Rice University team feels that this report is certainly not definitive with respect to siting. However, the methodology utilized appears appropriate to the problem of siting. Future work will be required prior to definitive sites being identified, and a major attempt should be made to coordinate additional work with existing Federal governmental data management systems.

TABLE OF CONTENTS Page I. Introduction ................................................... 1 II. Methodology ................................................... 5 A. Mapping of Variables ...................................... 5 B. Analysis of Data.......................................... 1. Data Encoding Storage Access and Display ........... ^2 2. Data Analysis.......................................... 15 III. Discussion of The Variables................................... 17 A. Land and Water............................................ 17 B. Federal Lands............................................ 19 1. National Recreation Areas............................. 19 2. Indian Reservations.................................... 21 3. Military Reservations................................. 22 4. Other Federal Lands. .................................. 22 C. National Forests .......................................... 23 D. Population ................................................. 25 1. Standard Metropolitan Statistical Areas............... 27 2. Population Density Greater Than 50 Persons per Square Mile............................... 29 3. Adjusted Population Distribution .................... 30 E. Wetlands................................................... 31 1. Marshes................................................ 33 2. Wetlands.............................................. 33 F. Topography Unacceptable................................... 33 G. South Slopes.............................................. 36 H. Navigable Waterways........................................ 38 I. Interstate Highways........................................ 38 J. Endangered Species Habitats............................... 41 K. "Prime Agricultural Lands" ............................... 43 L. Flyways of Migratory Waterfowl ........................... 47 M. Seismic Hazards............................................ 49 N. 40 Degree Latitude........................................ 51 0. Winds Greater Than 50 Knots............................... 53 P. Number of Days With Hail................................. 55 Q. Number of Days With Thunderstorms........................ 55 R. Sheet Rainfall............................... 58 S. Acid Rain................................................... 58 T. States..................................................... 61 U. Conclusion................................................. 61 IV. Analysis of The Data.......................................... 63 Summary Map 1........................................................................................................... 66 Summary Map 2................................................... 69 Summary Map 3................................... 72 Summary Map 4................................................... 75

IV. Analysis of The Data - (continued) Summary Map 5................................................. 78 Summary Map 6................................................. 81 Summary Map 7................................................. 84 Summary Map 8................................................. 87 Summary Map 9................................................. 90 Summary Map 10................................................. 93 Summary Map 11................................................. 96 V. Conclusions and Suggestions For Future Work................... 99 Appendix A..................................................... 103 Appendix B..................................................... 107 References..................................................... Ill

MAPPING OF EXCLUSION AREAS FOR RECTENNA SITES I. INTRODUCTION In determining the overall feasibility of the Satellite Power System (SPS), many important issues must be analyzed in great detail. One major area of inquiry concerns where the receiving antennas for microwaves beamed from space can be located within the continental United States. As set forth in the reference design, these receiving antennas will require sites of approximately 50,000 acres each, with 60 such sites being required across the United States. These 60 sites, therefore, would require the dedication of approximately 3,000,000 acres of land for the receiving antennas exclusive of land required for transmission facilities, access roads and other activities related to the land use. The major purpose of this research effort was to determine where, or even if, 60 such sites existed. The approach utilized in this study was one of excluding land areas from consideration rather than seeking sites which had desirable characteristics. In other words, certain land areas cannot be considered as being eligible for rectenna sites since they already are dedicated land areas (as with existing cities and urban areas) or because of certain environmentally related characteristics that preclude other uses. If this set of variables can be determined and mapped, the land areas that were not mapped would emerge as "eligible" areas because no critical (or exclusion) variables were present in these areas. In conjunction with Allan Kotin, a set of important locational variables was compiled. These variables are included in the white paper on Resources by Allan Kotin, and an extensive list of references and a review of pertinent literature is also included in the Kotin Report. Even though the Kotin Report contains an extensive review of

pertinent aspects of the Satellite Powe/ System and the Rectenna, a short summary of certain characteristics is needed to place this locational work in a system context. Under the SPS, a large satellite in geosynchronous orbit at the equator beams microwaves to a receiving antenna on the earth's surface. The satellite and rectenna are sized for 5 gigawatts D.C. power output. The satellite consists of a flat solar array with a transmitting antenna (1 km diameter) on one end. The receiving antenna is elliptical n shape and is 13 km on tne north-south axis and 9 km on the east-west axis. This design is based upon a reference latitude of approximately 34 degrees north. The satellite's position in geosynchronous orbit means that the circular microwave beam will project an ellipse on the earth's surface anywhere but at the equator. Therefore, the north-south dimension of the rectenna will increase as a site moves north from the 34 degrees north reference posi tion. The receiving antenna is composed of a large number of 10 meter X 10 meter receiving panels. These panels are elevated in certain designs and are on the land surface in other designs. These panels, whether on the land surface or elevated, will cover approximately 25,000 acres. The power density at the center of the rectenna will be approximately 23 milliwatts per square centimeter with the power density diminishing to 1 milliwatt per square centimeter at the edge of the rectenna. Although the United States standard for microwave exposure is 10 mW/cn/, standards in other countries such as the Soviet Union are much more restrictive. For this reason, the reference site in the Kotin study and in this Rice University study contains a 2 km buffer zone surrounding the rectenna. This makes the configuration of the site a 17 km X 13 kn ellipse. The use of a 2 km buffer zone lowers the microwave power density to 0.1 mW/cm^ at the edge of the buffer. This level is 10 times higher than the Russian standard for non-occupational exposure. The current implementation plan calls for 60 such satel1ite/rectenna

pairs to be constructed. Construction will connence in 1996 with the first system operating by the year 2000. From the year 2000 to the year 2030, two satellite/rectenna pairs will become operational each year, totalling 60 pairs. The total land area estimated to be required for the 60 rectennas (with a 2 km buffer zone) is approximately 13,300 km^ (over 5100 square miles), slightly less than 0.2% of the total land area of the continental United States. Therefore, the following factors are important from the perspective of seeking 60 sites. First, some 55,000 acres will be required for each rectenna. Of this, approximately 50% will be cleared, with the remaining 50% being left uncleared but with restricted access due to microwave levels. Second, microwaves levels will be measureable beyond the rectenna with the buffer zone, and such levels may interfere with radio and other types of communications and navigation equipment. The extent of this radio frequency interference is not known, but the rectenna and the microwave beam per se are considered as a problem from a radio communications standpoint in this siting study. Third, the possibility of multiple use beneath the receiving antenna has been raised, but this siting study assumes that the land area directly beneath the receiving antenna will be lost from functional, if not physical, standpoint. In this study, no attempt was made to quantify additional land areas required for access roads, construction buildings and transmission lines. In the sections which follow, the siting methodology is discussed first. Then, a detailed description and analysis of the data used will be presented. Section IV presents the results of certain analytical efforts and Section V lists our conclusions and suggestions for future work.

II. METHODOLOGY The methodology utilized in this analysis consisted of three distinct steps. The first step involved the compilation and mapping of 36 data items. The second step involved the encoding of these 36 variables into Rice University 's computer system. At this stage, the 48 states were also coded and entered as data items. The third step involved sequentially overlaying variables to produce "synthesis" maps, representing compilations across specified variables. In this manner, a declining number of eligible areas were identified in each map. Further the impacts of the addition of certain new variables can be viewed directly. For each of these synthesis maps, a tabular summary was compiled which offered additional information concerning the grid cells that emerge as "eligible" A. MAPPING OF VARIABLES As mentioned earlier, the Rice University team worked with Allan Kotin in determining a list of important locational variables. Once these variables were identified, the mapping exercise was initiated. Those variables that are mapped and discussed subsequently in this paper are: Land and Water - Figure 4 Federal Lands - Figure 5 National Recreation Areas Indian Reservations Military Reservations Other Federal Lands National Forests - Figure 6 Population - Figure 7 Standard Metropolitan Statistical Areas Population on Density Greater Than 50 persons/sq. mile Adjusted Population Density Marsh Vegetation - Figure 8 Wetlands - Figure 9 Topography Unacceptable - Figure 10 Open Mountains Hills Mountains

Topography Unacceptable, South Slopes - Figure 11 Open Mountains Hills Mountains Navigable Waterways - Figure 12 Interstate highways - Figure 13 Endangered Species' Habitats - Figure 14 Land In Cultivation - Figure 15 Irrigated Land Cropland Land Suitable for Cultivation - Figure 16 Greater Then 67% suitable 50% to 67% suitable Flyways of Migratory Waterfowl - Figure 17 Seismic Hazards - Figure 18 Major Damage Potential Moderate Damage Potential 40 Degree Latitude - Figure 19 Windstorms - Figure 20 2% Probability of Winds Greater than 50 Knots 1% Probability of Winds Greater than 50 Knots Hail - Figure 21 Thunderstorms - Figure 22 Sheet Rainfall - Figure 23 Acid Rainfall - Figure 24 PH Between 4.0 and 5.0 PH Less than 4.0 Once the data was gathered for each of the above variables, the information was entered onto a map of the United States that was divided into grid cells. These grid cells were used for coding purposes, and the translation of the information to this form was essential to the completion of the project. Prior to entering the information on the gridded map of the United States, a decision was made concerning the size of the grid cell. The size of the grid cell represents a compromise between the time and resources available for the task and the desire to obtain as much spatial resolution as possible. The result was the choice of a grid square approximately 26 km or 16.2 miles on a side. The total land area within the grid square is approximately 170,000 acres. In Figure 1, the relationship of the rectenna site to the grid square can be seen. The rectenna occupies approximately 30% of a single grid square. While it is arguable that a greater resolution



would have been oesirable, it is the opinion of the Rice University research team that this resolution is sufficient to be meaningful from a locational standpoint. Therefore, all of the maps of these variables will be presented in a 26 km grid cell format. Figure 2 shows the gridded map of the United States. The mapping of the information at the grid cell level required a determination of the presence or absence of the variable from grid cells across the United States. Certain variables were coded as being present if any portion of the variable was indicated as being present within the grid cell whereas other variables were mapped as being present only if approximately 50% or more of the cell contained the variable. The discussion in Section III offers an explanation of the coding procedure on a variable by variable basis. Due to time constraints and/or data limitations, certain variables considered to be important from a locational standpoint were not mapped in this study. Those variables of concern that were not treated include: Local or State Owned Land (State and Local Parks) Poor Soils High Groundwater Table Highways Other Than Interstate Highways Airports and Air Approach Corridors Major Air Corridors Railroads Dust Storm Areas Wildlife Habitats (Other Than Designated Endangered Species) Very Poor Air Quality Near Major/Numerous RF Sources As will be explained in Section III, "Discussion of the Data", many of these variables are represented to some degree through other variables that were mapped. For instance, the mapping of Standard Metropolitan Statistical Areas should include much of the land area that is dedicated to major airports and major air approach corridors, and would include major/numerous RF sources. Nonetheless, it is important to note that the above variables were not mapped independently.

B. ANALYSIS OF THE DATA The analysis of the date consisted of two different yet complementary procedures. The first procedure concerned the decision on a variable by variable basis that (1) the variable was an absolute exclusion variable or (2) the variable was a potential exclusion variable. Due to the preliminary stage of the reference design, the analysis of the importance of most variables was from the perspective of the dedication of land areas for other uses. Generally, those variables that represented land uses that could not be preempted by the rectenna were identified as absolute exclusion variables. The remaining variables were treated as potential exclusion variables. Further, there are two types of potential exclusion variables. These are (1) variables that represent an environmental or resource constraint that may not be addressed through design modifications and (2) those variables that exhibit the capacity to exclude the reference system but that can be addressed through design modifications. At this time, it is difficult to speculate upon which of the non-design related potential exclusion variables will emerge as critical locational criteria. For example, many variables in this category are identified because of unique 1egal/institutional problems associated with their use (i.e., the use of Indian lands for sites) and others are identified due to the uncertainty of the microwave effects (i.e., the effects upon migratory waterfowl). Due to the disparity between such variables, the analytical approach is designed with an ability to aggregate and disaggregate the variabl es. The variables indicated as being design variables appear easier to assess. In many respects, these variables will cause design modifications if the rectenna is to be located in areas where these variables occur. In turn, addressing these variables will require additional dollars to be expended and modifications in the cost expectations relative to the rectenna will result. Therefore, these variables appear

more applicable in the context of understanding the full costs of building the rectenna. These environmental variations were not addressed in the reference system explicitly, although they are implicitly addressed in the range of costs to be expected (lowest cost and highest cost per satellite and rectenna pair). Additionally, certain of the variables mapped in this study are potential inclusion variables. In other words, there are aspects of these variables that may make location within their spatial dominion desirable. For example, "other federal lands" (excluding national forests) would be potentially good sites since 25% or more of the mapped area is under federal control. These other federal lands are relatively inexpensive; controversial regarding the aggregation of a 50,000 acre contiguous parcel may be less than would be the case where private property would have to acquired for the entirety of the site. Therefore, there are many nuances of the information presented in Section III. For this reason, a major attempt was made in this study to fully document the approach and assumptions so that additional variables can be considered and the impact of individual variables can be traced. In other words, a critical element of this approach is to determine those variables that "drive" the locational decision. A further result may be the identification of options that have not been considered heretofore. 1. Data Encoding, Storage, Access and Display: Prior to analysis per se, the data utilized for analytical purposes was entered into the Rice Architecture Geographic Information System (RAGIS). The basic elements of RAGIS are shown in Figure 3. One of the central features of the information system is the use of a nost language to support and control its operations. The computer language Speakeasy (developed at Argonne National Laboratory, see Cohen and Pieper, 1977) was used since it stresses the use of English syntax, conversational input-output modes and on-line interactions. Speakeasy is an extensible language that comes with broad general


operating capabilities but also allows users to include functions and operations peculair to their classes of problems. These special functions or operations may in fact take the form of algorithms written and compiled in other system supported languages, such as FORTRAN, and simply linked into Speakeasy's processor. These linked load modules are called linkules. To the user linkules are a vocabulary of English language key words that allow the associated programs to be called and executed by name. During the development of RAGIS, an extensive number of linkules specifically related to napping and spatial analysis were established. In fact, RAGIS has become a subsystem within Speakeasy consisting of more 250 programs and special operations for geographic information processing. Data is encoded through the use of a CRT (Cathode Ray Tube). A data encoding linkule establishes a uniform grid of cells (64 x 48) as a two dimensional array across the CRT screen. Maps or other special features to be encoded are reproduced at the appropriate scale as black and white film positives. These film positives were taken from the 21 maps of the United States, with six individual 64 x 48 grid "cards" being required for each map to be encoded. The film positive is placed on the screen and the pattern of each feature is then visually coded in a raster-like fashion as a dot pattern. In this study, a feature was coded as either present or absent within the appropriate grid cell. Checking the encoded data is accomplished on-line at the CRT screen by simply displaying each dot pattern and checking it against the film positive overlay. The coded data is stored in the computer memory as a logical bit stream with each image forming a distinct binary pattern. This form of storage utilizes the computer's memory switches and results in considerable savings for basic storage, retrieval and display operations. Each data pattern is assigned a unique alphanumeric code or name and becomes like any other word in the system from the user's standpoint. Access is achieved by simply calling for the pattern by name and having

it displayed or integrated into a computational sequence in the appropriate manner. In addition to various spatial display characteristics, RAGIS also employs a relational data base management subsystem Culled Rspeak (Schlicting, 1977) which complements other existing analytical capabilities such as multi-variate clustering and factoring. The major importance of the relational data base management capabilities relates to the ability it offers in understanding or perceiving the results of the analysis vis a vis the raw data. One use of the relational capabilities will be to analyze the results of a certain analytical exercise with respect to other variables that were not utilized in the analysis directly. In other words, through the development of tables of data, one can gain many insights about analytical results independently from the information gained directly through the analysis. 2. Data Analysis In the determination of areas that are "eligible" for rectenna sites, the Rice University approach first identifies areas from which the rectenna would be excluded. The areas remaining after the exclusion area had been determined would be the "eligible" areas. This relatively simple concept is achieved through the use of overlay or sieve analysis. Generally, this technique requires that a list of environmental features be prepared and arranged so that features are ranked in order of assumed decreasing (or increasing) order of importance. In the case of this determination of exclusion areas, the absolute exclusion variables would be considered first with the potential exclusion variables and the design variables considered subsequently. These variables are displayed on transparent maps, and by overlaying these maps, the areas of composite shading becomes apparent. Unlike other uses whereby the darkness of the shading indicates the degree of developability (or non-developability), the approach utilized in this study weights all absolute exclusion variables equally. Therefore, the new set resulting from the union of mapped variable set A with

mapped variable set B would be exclusion area 1. In this respect, the methodology utilized for this study differs from traditional studies such as those by Lewis (1962), Alexander and Manheim (1962) and McHarg (1969) because these planners were concerned with the intersection of the mapped variable sets and with interpretations of development suitability based upon the number of variables intersecting. Although the previously cited authors used map overlays rather than computer generated overlays, the applicability of a computer system to this type of analysis sould be obvious. Attempts ;o computerize the approach are fairly numerous (Ward and Grant, 1970; Krauskopf and Bunde, 1972; Rowe and DeLeon, 1973), and several recent attempts to innovate the basic technique are well summarized by Hopkins (1976). Therefore, the approach utilized for data analysis will be as follows. First, an initial overlay map will be composed by containing five variable sets to determine their cumulative coverage. This initial overlay will result in overlay Map 1, which will become a new variable. This resultant map will have shaded areas (exclusion areas) and white areas (eligible areas). A statistical profile will then be generated (utilizing the Rspeak capabilities) which will describe the number of eligible sites (grid cells) by state, and this table will also contain information about some of the design variables such as the number of eligible sites above 40 degrees north latitude, the number of sites subject to acid rain, etc. Then, Exclusion Map 1 will be added to variable 6 to form Exclusion Map 2. The same statistical profile will then be used to describe the eligible areas. Then, additional variables will be added until several exclusion maps, each being more restrictive, will be compiled. In this manner, the effects of certain variables will be clearly identified. Finally, these exclusion maps will be overlayed with at least one and possibility two informational maps that will place the resulting

sites in a better perspective from the standpoint of (1) the existing power distribution system and (2) the existing interstate highway system. In this manner, a strong description of eligible sites will exist. III. DISCUSSION OF THE VARIABLES As discussed in Section IIA, 21 maps consisting of 36 environmental variables were prepared during the course of this study. Prior to presenting the results of analyses performed using this data, the rationale for use of the information and an assessment of reliability of the data must be presented. In the sections that follow, each map is discussed from the perspective of (1) the rationale for the use of the information; (2) the source and reliability of the data and (3) the spatial coverage of the variable. A. LAND AND WATER - FIGURE 4 Rationale: Two reasons exist for mapping land and water areas. First, a base map was needed for coding purposes that established a uniform treatment of grid cells at the interface of land and water. As shown in Figure 4, the decision was made to code a cell as land if a portion of the cell included land area. This coding decision established a protocol for treating cells in subsequent mapping efforts. The second reason for coding land and water areas was to identify the degree to which water sites need to be considered for rectenna sites. At this time, water sites are considered as potential exclusion areas because the reference SPS system does not include offshore construction specifications. At this point in the analysis, it is impossible to assess if offshore sites are needed. However, the expectation is that onshore sites may be difficult to locate within the eastern half of the United States. Therefore, potential water sites are identified to a distance of 32 miles offshore (2 grid squares). All of the Great Lakes are also


shown, although an international border does bisect the mapped waters. Reliability of the Data: The base map from which the land and water areas were drawn was Richard Edes Harrison's map titled "Shaded Relief", wnich was published by the United States Geological Survey in the National Atlas of the United States. The map was drawn at a scale of 1:7,500,000, and 1" was equal to approximately 118 miles. The coding decision with respect to land areas leads to an over representation of the land area of the United States. Therefore, the sum of the grid cells identified as land would represent a slightly larger land area than is actually to be found in the Continental United States. The only water areas mapped were coastal waters and the Great Lakes. Therefore, water areas are underrepvesented both with respect to coastal boundaries and with respect to smaller lakes within the borders of the United States. Nonetheless, the information from which the map was drawn is considered to be highly reliable. Spatial Coverage: Al shown in Figure 4, land consists of 11699 grid cells. Because this map will be used as a reference map for coding purposes, the total number of grid cells available for coding of information is 11699. B. FEDERAL LANDS - FIGURE 5 In Figure 5, a map of the lands under federal ownership is presented. Four distinct variables are displayed in the map. These are (a) National Recreation Areas; (b) Indian Reservations, (c) Military Reservations, and (d) Other Federal Lands. This map was assembled directly from a map which had all four variables of concern, and Figure 5 was compiled by first scoring grid cells over National Recreation Areas, then scoring Indian Reservations, then military reservations and then other federal 1ands. 1. National Recreation Areas Rationale: Certain federal lands have been dedicated to


recreational or wildlife preservation uses. These lands are preserved as part of the heritage of the United States and their conversion to other uses is prevented by federal law. Although congressional action removing such prohibitions is possible, such action is unlikely. Therefore, these areas have been mapped and are considered as absolute exclusion areas. Included in the National Recreation Areas category mapped in Figure 5 are (1) National Parks, (2) National Monuments, (3) Federal Wildlife Reguges, (4) National Seashores and (5) National Recreation Areas. Reliability of the Data: The areas coded as National Recreation Areas were taken from a United States Geological Survey map titled "Federal Lands". This map was published in the National Atlas of the United States and compiled by the USGS as of January 1, 1968. Therefore, this data is considered to be highly reliable as of that date, but land areas added to these categories since 1968 are not included. This variable was coded as being present in a cell if the variable occurred in any portion of that grid cell. Spatial Coverage: The areas identified as National Recreation Areas are present in 424 grid cells. 2. Indian Reservations Rationale: Indian reservations are federal lands that are administered by tribes living on these reservations with the Department of Interior performing a guardianship function. This institutional situation gives the Indian tribes substantial control over the use of land within reservations, and there is a strong possibility that these tribes will not allow a rectenna to be constructed on their lands. Given this jurisdictional situation, Indian reservations were mapped and are considered as potential exclusion areas. Reliability of the Data: The areas coded as Indian Reservations were also taken from the USGS "Federal Lands" map cited previously. The reservations mapped are considered accurate, but certain smaller

reservations, such as the Alabama-Coushatta reservation north of Houston, Texas, are absent from this map, indicating that a size threshold was used by the USGS in assembling the "Federal Lands" map. Therefore, certain other smaller reservations may not be mapped. This variable was also coded as being present if the reservation appeared in the grid cel 1. Spatial Coverage: The areas identified as Indian Reservations are present in 558 grid cells. 3. Military Reservations Rationale: Two reasons exist for mapping military reservations. First, certain military reservations may not be desirable sites because sensitive radio and telecommunications equipment could be subject to radio frequency interference from the rectenna. Second, certain military reservations may have substantial acreages that are removed from population centers. These latter areas may provide excellent sites whereas the former would be undesirable sites. Therefore, military reservations are mapped and are considered as potential exclusion and potential inclusion areas. Reliability of the Data: The information appearing on Map 4 was also obtained from the previously cited "Federal Lands" map. This data is considered reliable, but the utility of the generic classification (military reservation) is questionable. A more detailed investigation of military reservations will be necessary prior to determining the proper interpretation to ascribe to this generic land use type. Again, the variable was coded as being present if the land use were in a cell. Spatial Coverage: The areas identified as military reservations are present in 175 grid cells. 4. Other Federal Lands Rationale: Other Federal lands were mapped for consideration as an inclusion variable. These lands are either wholly or partially

under the jurisdiction of the federal government, and they may be utilized for multiple purposes. Included in this category are national forests, as well as lands with greater than 25% ownership by the Federal government. While these lands may not be indiscrim- ately used, they are potentially available as sites for rectennas. Reliability of the Data: The areas coded as other federal lands also were obtained from the U.S.G.S.'s map titled "Federal Lands". The major shortcoming of this information is that much of the land area coded as other federal lands is only partially under the control of the federal government. While this partial ownership is helpful in aggregating 50,000 + acres of land, the mapped information is misleading if one assumes all of these lands are under federal control. This variable was also coded if present in a cell. Spatial Coverage: Other federal lands are coded in 3606 grid cells. C. NATIONAL FORESTS - FIGURE 6 Rationale: Although multiple use of National Forests is allowed under Federal law, the conversion of portions of these forests into sites for receiving antennas would be opposed by environmental groups and the U.S. Department of Agriculture. Therefore, these national forests have been separated from the "Other Federal Lands" category and are mapped separately. From an analytical standpoint, these lands are considered as potential exclusion areas, although as a practical matter, these areas should not be considered as being available for rectenna sites unless no other alternative sites exist in the region of the United States being analyzed. Reliability of the Data: The areas coded as National Forests also were taken from the USGS map of federal lands. This data is considered as being highly reliable and the variable was mapped if it was present within the grid cell. It should be noted that National Grasslands were


not mapped due to time constraints. National grasslands do not have the spatial coverage of the national forests but they should have been included on this national forest map. Therefore, the information contained in Figure 6 should have been expanded to include national grasslands. Spatial Coverage: The areas coded as national forests are present in 1323 grid cells. D. POPULATION - FIGURE 7 Populated areas offer several constraints with respect to rectenna siting. First, populated areas represent dedicated land uses and the displacement of large numbers of people is considered to be highly undesirable. Second, land prices in more densely populated areas will be substantially higher than in other areas of the United States. Third, microwave exposure levels adjacent to the rectenna will be higher than the background non-occupational standard used by the Societ Union. Although much research will have to be conducted before definitive statements can be made about microwave effects, this study's approach was to avoid locations immediately adjacent to urbanized areas. Given a desire to avoid populated areas, the next question to be addressed concerns the definition of populated areas. Three variables were utilized in the map shown in Figure 7. These are (1) Standard Metropolitan Statistical Areas, (2) population density greater than 50 persons per square mile and (3) adjuster population distribution. The rationale for utilizing these variables is presented in the following sections. Two general problems must be raised at this stage. First, the source for population information is the United States census, with the last census having been conducted in 1970. Therefore, population data taken from the census is 8 years old. Because this information is dated, the issue of future growth is raised. Although it is difficult


to project growth to the year 2000, certain trends are now evident that should be considered prior to a discussion of the variables mapped for exclusion purposes in this study. One of the best and most succint statements is found in the Council on Environmental Quality's seventh annual report. In this document, the Council states: "There are three important patterns of population distribution evident in the United States in the 1970's, each with its own implications for the future. The dominant pattern of population settlement continues to be the growth of major metropolitan areas, a trend that accelerated in the post World War II period, but which has now slowed appreciably. The second pattern is a considerable regional shift of population from the north central and northeastern sections of the country to the southern and western regions. The third pattern is a more recently observed phenomenom: the relatively rapid growth of population in non-metropolitan areas. We have chosen to highlight this pattern in the Annual Report because it is a growing trend that runs contrary to the basic pattern of growth throughout most of our history. For the first time, population in non-metropol itan areas is increasing faster than that in metropolitan areas." This non-metropolitan growth trend has many implications for rectenna siting. Unfortunately, it is difficult to speculate at this time about the spatial characteristics of this trend. In this study, the goal was to identify areas where rectennas could not be located. Therefore, the variables mapped and described below should be considered as a conservative indication of populated areas. Those areas that are not mapped should not be considered as automatically being without population. Instead, these areas did not have a sufficient density to indicate that rectennas could not be located there. In other words, the "white" areas do not indicate a locational carte blanche, and these areas should be studied in greater detail to determine the actual distribution of people within these areas. 1. Standard Metropolitan Statistical Areas (SMSA's). Rationale: A substantial portion of the United States has been

urbanized and these settlement patterns represent dedicated land uses. An initial indication of the location of these urban settlements may be gained by mapping Standard Metropolitan Statistical Areas (SMSA's). SMSA's are the major metropolitan areas of the United States and are also areas where future population growth is likely to occur. The research team determined that these areas would exhibit settlement patterns precluding the aggregation of 50,000 + acres of land. Further, if such an agglomeration were indeed possible, the cost of the land would be prohibitive. These SMSA's are shown on Figure 7. Another reason for identifying SMSA's as exclusion areas is that many other activities that represent siting constraints are present in these areas. For example, most of the major airports of the United States are found in SMSA's (as well as other densely populated areas), and the approach corridors most likely will be present within SMSA's. Additionally, substantial concern has been voiced concerning the potential radio frequency interference effects of the rectenna. Since the majority of sources that could be disturbed by radio frequency interference are located in major metropolitan areas, the mapping of SMSA's (and other urbanized areas) begins to address the RFI issue. Reliability of the Data: The areas shown as SMSA's in Figure 7 are from a map prepared by the Geography Division of the Department of the Census. The definition of SMSA's was developed by the United States Office of Federal Statistical Policy and Standards as of December, 1977. The data is considered highly reliable. However, alterations were made to certain SMSA's prior to their being mapped on Figure 7. These alterations were made because of the basis for determining the spatial coverage of SMSA's. Generally, SMSA's are delineated along county boundaries. For most areas of the United States, the approach is sensible because counties are relatively small. However, the western United States (notably California, Arizona, Nevada, Utah and Oregon) has extremely large counties. To map the entirety of these

SMSA's as exclusion areas would be misleading since large portions of these counties are not urbanized. Therefore, certain alterations were made to these defined SMSA's. Alterations were made by examining a "dot" map of population distribution, prepared by the Department of the Census from 1970 Census data, and eliminating grid cells from the SMSA that were indicated as having less than 500 people. Spatial Coverage: Areas indicated as SMSA's on Figure 7 include 1871 grid cells across the United States. 2. Population Density Greater than 50 Persons per Square Mile Rationale: As discussed previously, populated areas other than SMSA's need to be represented. A second type of indication is a population density analysis, and areas that were identified as having a population density greater than 50 persons per square mile were mapped and considered as exclusion variables. Reliability of the Data: The areas mapped as having a population density greater than 50 persons per square mile were compiled from a map prepared by the Department of Census from 1970 census data. Therefore, the information represented by this variable is somewhat out of date. However, a more severe shortcoming of the data relates to the fact that the data was represented on a county by county basis. Although counties are commonly used for data representation purposes, the land area represented by many counties causes significant concentrations of people to be diluted when the data is mapped on a county by county basis. Therefore, while there is a great degree of confidence that the mapped information is an accurate representation of those counties with a population density greater than 50 persons per square mile, there are many areas of the United States that are not adequately represented through the use of this population variable. Spatial Coverage: The land area represented as having a population density greater than 50 persons per square mile include 1276 grid cells. It should be noted that this count does not include the areas previously mapped as SMSA's.

3. Adjusted Population Distribution Rationale: Due to the problem of county land areas diluting the population density per square mile measure used above, a third population variable was mapped as an exclusion area. This area is identified as "adjusted population distribution" and it represents a third approach to determine the exclusion area for population. Reliability of the Data: Unlike the two previous variables, adjusted population distribution represents a somewhat subjective approach to population density. Th..- data was developed by overlaying the grid of the'United States over the U.S. Department of the Census's dot map of the population of the United States. This dot map is based upon 1970 census data. It is important to note that certain counties of the United States are large enough that 10 to 20 grid cells fit within their boundaries. By overlaying the grid cells and the dot map, substantial areas were added to the populated areas map. Although replication of the process might lead to varying results, certain rules were followed. First, any grid cell with a city of 25,000 persons or more was added. Second, if there were two towns of 10,000 persons or more, the grid cell was scored. Third, if there were combinations of a town of 10,000 or more and a number of dispersed, smaller dots, the grid cell was scored. As stated earlier, this process was subjective and was based upon visual examination of the dot map. However, it is felt that those areas covered by the "adjusted population density" variable should not be considered as sites. The important point is that the combination of the three population variables represents a rather conservative assessment of the land areas unavailable as potential sites on the basis of population. Spatial Coverage: The land area mapped as "adjusted population distribution" consists of 419 grid cells.

E. WETLANDS - FIGURES 8 and 9 Rationale: Wetlands have been a focal point of environmental concern for many years now. The United States Army Corps of Engineers has jurisdiction over dredge and fill activities in most wetland areas of the United States under Section 404 of the Federal Water Pollution Control Act Amendments of 1972 and 1977, and the U.S. Environmental Protection Agency has issued guidelines that are intended to prevent the conversion of wetlands to other uses. Wetland areas are extremely important habitat areas, both for marine and avian species. Although it is possible under existing federal statutes to convert wetlands, it is clear that the intent of Congress is to protect wetland areas. Therefore, wetlands are considered as absolute exclusion areas, and are shown in Figures 8 and 9. Reliability of the Data: While the intent of Congress may be clear, the spatial distribution of wetlands is more obscure. Wetlands may be marshes. swamps or ponded areas within farmlands or forests. At this time, the United States Fish and Wildlife Service is attempting to compile a comprehensive inventory of wetland areas, but this study is not complete. For this reason many smaller wetland areas are not included in Figures 8 and 9. A 1955 Department of the Interior publication titled "Wetlands of The United States" identified important wetlands in a very generalized fashion. This publication determined that there were 22,400,000 acres of wetlands of primary importance to waterfowl and 52,000,000 acres of wetlands of lesser importance. However, these areas could not be mapped with a sufficient degree of accuracy. Therefore, a need exists for additional information before this issue can be adequately addressed. The two maps presented in Figures 8 and 9 include only larger wetland systems, and these figures should be considered as conservative indications of the spatial coverage of wetland areas.