EARTHQUAKE INVESTIGATION in the United States THE PENNSYLVANIA STATU UNIVERSITY I [ON SPECIAL PUBLICATION NO. 282 Revised (1964) Edition U.S. DEPARTMENT OF COMMERCE Coast and Geodetic Survey U.S. DEPARTMENT OF COMMERCE John T. Connor, Secretary COAST AND GEODETIC SURVEY H. Arnold Karo, Director SPECIAL PUBLICATION NO. 282 Revised (1964) Edition Earthquake Investigation in the United States U.S. GOVERNMENT PRINTING OFFICE WASHINGTON : 1965 For sale by the Superintendent of Documents, U.S. Government Printing Oflfio Washington, D.C., 20402 - Price 20 cents Prepared originally by Frank Neumann, Geophysicist, Coast and Geodetic Survey, and Published 1952 Revised 1953, 1958, 1962, and 1964 CONTENTS Page Introduction 1 Seismological organizations 1 World earthquakes 3 Earthquakes in the United States 5 Eastern region 6 Mississippi Valley region 7 Western mountain region 9 Pacific coast and Nevada 10 Alaska 13 Hawaii 14 Puerto Rico and Panama Canal Zone 15 Seismic sea waves 15 Nature of earthquakes and seismic waves 17 Locating earthquakes from instrumental data 17 Seismic sea wave warning service. 21 Seismographs 23 Descriptive reports on earthquakes 28 Engineering seismology 29 What to do when an earthquake strikes 33 Microseisms 35 Appendix 37 Table 1. — Active teleseismic stations of the United States 37 Table 2. — Lives lost in major United States earthquakes. 43 Table 3. — Damage caused by major United States earth- quakes 44 Seismological publications of the Coast and Geodetic Survey 45 in ILLUSTRATIONS Page Figure 1. — General trend of the earthquake belts of the world. 4 Figure 2. — Coast and Geodetic Survey epicenter map of the United States 8 Figure 3. — Earthquake waves recorded on typical seismogram and paths of propagation through the earth 19 Figure 4. — Travel-time curves with idealized seismograms super- imposed 20 Figure 5. — Geophysicists of the Coast and Geodetic Survey illus- trating technique of locating an earthquake on a 32-inch globe_ 22 Figure 6. — Modern seismograph as used in the World-Wide Net- work of Standardized Stations 24 Figure 7. — An assembly for visible-recording seismograph 25 Figure 8. — College, Alaska seismogram showing record of an earthquake on February 22, 1958, in the Andreanof Islands, Aleutian Islands 27 Figure 9. — Coast and Geodetic Survey type accelerograph 29 Figure 10. — Olympia accelerogram of Puget Sound, Wash. earthquake of April 13, 1949 32 Figure 11. — Seismogram showing microseismic storm of Decem- ber 1 and 2, 1956, at Sitka, Alaska 35 IV Earthquake Investigation in the United States Introduction To the layman, earthquakes are generally considered a very unusual type of natural phenomenon — an indication that something occasionally goes wrong deep down in the earth. To the seismologist who studies earthquakes from their scientific viewpoint they represent a process of nature as unceasing and routine as the winds that blow over the oceans. It is the purpose of this booklet to explain the more important facts of earthquake phenomena and to outline the part played by the Federal Government and various seismological organizations in dealing with the scientific and economic aspects of the earthquake problem. The booklet supersedes previous publications bearing the same title and published in revised form in 1929, 1952, 1953, 1958, and 1962. Seismological Organizations The United States Coast and Geodetic Survey is author- ized by law to make investigations in seismology, but other organizations in the United States are also engaged in such work. One of these is the Jesuit Seismological Association that coordinates the work of 27 affiliated seismological stations spread over most of the country. The California Institute of Technology operates 17 stations in southern California for the intensive study of earthquakes in that area, and the University of California operates a somewhat similar network of 20 stations in the northern California and western Nevada area. In addition, more than a score of stations are operated independently by universities in connection with their geological and geophysical programs. One of the seismological functions of the Coast and Geodetic Survey is to serve as a clearing house for much of the sta- 2 COAST AND GEODETIC SURVEY tistical and other information collected by these various groups, to publish summaries of this information, and to fill in serious voids in the instrumental program. In this collective effort to locate earthquakes accurately in the United States and elsewhere, the Survey operates 14 stations of its own and actively cooperates in the maintenance of 11 others, mostly university stations. The United States Geological Survey operates seismograph stations in Hawaii to study local earthquakes due to volcanic activity. It will subsequently be shown that, far from being ex- clusively of academic interest, the results obtained from this concerted effort are of practical use to many profes- sional groups — to structural engineers and insurance actuaries who want to know exactly where, and to what degree, earth- quake dangers exist; to structural engineers who want to know the magnitude and nature of the earthquake forces they must design structures to resist; to geologists who en- deavor to associate seismic activity with the faults they map in the course of their field surveys; and to meteorologists who want to know just how the minute vibrations of the earth's surface that are unceasingly registered on seismo- graphs are related to storms and other meteorological phe- nomena. In the academic field the study of seismic wave propagation has furnished the most authentic information yet obtainable on the physical nature of the earth's interior. The study of earthquake waves has also furnished the basic concepts used in locating oil-bearing structures by seismic exploration methods. The program of the Coast and Geodetic Survey is a rep- resentative amount of the seismological work done in this country and includes projects of a highly specialized nature. Besides operating seismological stations to locate earth- quakes it collects statistical information on all types of earth- quake phenomena, including damage, prepares earthquake catalogs and epicenter maps, and conducts various types of investigations directed toward a better understanding of earthquake phenomena. One of the most important phases of the Survey's seismo- logical program is the investigation of destructive earth- quake motion, a program that is of basic importance to the engineer who must design structures to successfully resist EARTHQUAKE INVESTIGATION IN THE UNITED STATES 3 earthquake forces. The 700 persons killed in the great California earthquake of 1906 and the billion dollar (present- day value) property loss caused from the fires that followed will always stand as a warning to those who feel that the earthquake menace can be ignored. Years of study have shown that the problem of the design engineer is technically difficult because earthquake forces are vibrational or dy- namic in character and cannot be treated the same as static, or steady, forces. Much has been accomplished, however, and the Survey has played an important role in this ac- complishment through furnishing authentic information on destructive ground and building motions. While the work of other seismological groups is devoted in part to the study of regional seismicity, the scope of their programs is extremely broad and includes such projects as world seismicity investigations, study of continental and oceanic structure by seismic exploration methods, the rela- tion between microseisms and hurricanes and scores of other projects, many of which are closely related to other fields of science. An interesting phase of seismological research is its inter- national aspect. The fact that a strong earthquake in any country is registered on seismographs all over the world has resulted in a world-wide exchange of data and cooperative effort that is matched in few other fields. This international effort has resulted in a great accumulation of technical data that has not only made possible an authentic history of world earthquakes over the past 60 years, but exhaustive analyses of the data have given the scientific world the most accurate picture it has of the physical structure of the interior earth. World Earthquakes It is estimated that more than a million earthquakes occur throughout the world annually. They range from minor temblors that are barely perceptible locally to catastrophic shocks. Fortunately, most earthquakes originate beneath the sea where they cause little concern except when seismic sea waves are generated. Such waves occasionally cause damage and loss of life near the earthquake origins and also thousands of miles away. Approximately 700 of the mil- COAST AND GEODETIC SURVEY y'J : » . *y. ' ;-V;V:, -is:; *© « ^ ■3 ^ ^ £ 05 « o i&a EARTHQUAKE INVESTIGATION IN THE UNITED STATES 5 lion annual shocks may be classed as strong, capable of causing considerable damage in the areas where they occur. The Coast and Geodetic Survey, with the cooperation of many foreign and domestic agencies, locates most of these within a few hours or days. The annual epicenter maps that pinpoint these locations are illuminating exhibits that show where most earthquakes occur. (See Figure 1, page 4.) One is struck not only by the existence of certain well- defined seismic belts stretching over large areas of the world, but also by the persistence with which the over-all pattern is repeated year after year. The rim of the Pacific Ocean outlines the world's greatest seismic belt. This belt includes the Pacific coast and western mountain region of the United States and a large part of Alaska. A second major belt branches off to the west, ex- tending across the southern portion of Eurasia, through the Mediterranean area, to the Atlantic Ocean. There are also many minor belts such as that looping eastward from the Pacific through Mexico, the West Indies (including the Windward Islands), and those countries bordering the south- ern shores of the Caribbean Sea. Earthquakes in all of these belts are taken for granted but great earthquakes also oc- casionally occur outside these zones. In this country south- eastern Missouri and Charleston, S.C. are examples of such areas, but in such cases many years usually elapse between destructive shocks. This brief outline shows that not only are the great cities of our Pacific coast vulnerable to de- structive earthquakes, but the menace also exists in latent form in many areas ordinarily thought to be immune. Earthquakes in the United States The earthquake history of the United States is outlined in Coast and Geodetic Survey Serial Publication 41-1, Earth- quake History oj the United States, Parts I and II. Part I covers continental United States (exclusive of California and western Nevada) and Alaska. Part II covers the stronger earthquakes of California and western Nevada. Both vol- umes are available for purchase from the Superintendent of Documents, Government Printing Office, Washington, D.C. 20402, at 60 cents and 40 cents, respectively. Information 768-393 0—65 2 6 COAST AND GEODETIC SURVEY on recent earthquakes will be found in the annual issues of the United States Earthquakes series also published by the Coast and Geodetic Survey. The following is a resume of the earthquake history of the United States based on infor- mation appearing in these publications. (See Tables 2 and 3, pages 43 and 44, respectively, for lives lost and resultant damage in major United States earthquakes.) Eastern Region The earliest recorded earthquake of destructive intensity felt in the United States occurred in the valley of the St. Lawrence River in 1663, centering apparently between the present sites of Montreal and Quebec. Great landslides kept the river muddy for a month. A less severe earthquake occurring in the same general region in 1870 was felt as far south as Virginia, but damage in the epicentral area was comparatively light. Another shock centering 300 miles west of this area, near Timiskam- ing, Canada, shook most of the northeastern part of the United States on November 1, 1935. On September 5, 1944, a shock centering close to Massena, N.Y. and Cornwall, Canada, caused $2 million damage in these towns. Various parts of New England have experienced moderate to severe earthquakes, some of sufficient intensity to damage houses, especially chimneys. The most severe was that of November 18, 1755, probably centering near Cambridge, Mass. Many chimneys were thrown down or twisted and roofs were wrecked by the fall of chimneys. In 1791 there were strong shocks near East Haddam, Conn., and for years the area was noted for subterranean noises of seismic origin. On November 18, 1929, all of New England was shaken by a strong submarine earthquake off the Grand Banks. A coastal area extending from Richmond, Va. to Portland, Maine, and inland as far as northwestern Pennsylvania, was shaken on August 10, 1884, by an earthquake that apparently originated near the edge of the continental shelf off the mouth of the Hudson River. On Long Island, walls and plaster were cracked at many places. The lake region of northeastern New York has experienced several shocks strong enough to damage chimneys. On August 12, 1929, EARTHQUAKE INVESTIGATION IN THE UNITED STATES 7 Attica, in the western part of the State, experienced a shock that knocked down 250 chimneys. In New Jersey, Pennsylvania, and the states to the south, occasional strong but nondestructive shocks are believed to be the result of the settling of the coastal plain sediments on the underlying basement rock that structurally repre- sents the base of the Appalachian Mountain range. In 1897 one of the strongest of this type occurred in Giles County, Va. where old brick houses and chimneys were cracked. It was felt as far as Indiana and Georgia. The greatest shock in the east was the Charleston, S.C. earthquake of August 31, 1886. It was felt as far as Chicago and Boston. In Charleston, 60 persons were killed and many buildings were ruined or severely damaged. In the surrounding country, bridges were wrecked, railroad tracks were twisted, and sand and water were ejected from numerous craterlets. Mississippi Valley Region Although this region is not generally considered seismic, in 1811 it was the scene of the greatest earthquake this country ever experienced — an earthquake rated as one of the great earthquakes of known history. It centered near New Madrid in southeastern Missouri and was felt over two-thirds of the United States. With this area of the country now so thickly settled, one can only conjecture what might happen if a disturbance similar to that of 1811 were repeated. Evidence exists that there was a similar earthquake in the same region about 100 years prior to the 1811 shock. A lasting result of the 1811 disturbance was the lowering of a large part of the countryside in southeast- ern Missouri and northeastern Arkansas now known as the "sunken country. " In time of flood the sunken country, with its many lakes and bayous, has served as a temporary reservoir for the flood waters of the Mississippi River, but in recent years other measures have been taken to solve the flood problem. Severe shocks were reported from the general area in 1843 and 1895 but hardly a year has passed without some COAST AND GEODETIC SURVEY EARTHQUAKE INVESTIGATION IN THE UNITED STATES 9 disturbances being reported. The 1843 shock was felt strongly in St. Louis, Mo.; Frankfort, Ky.; Cincinnati, Ohio; and Nashville, Tenn. In 1895 a lake was formed over 4 acres of ground that sank near Charleston, Mo., and many chimneys were demolished. Near Bertrand, sand was ejected from the ground and the water table was apparently raised . The shock was felt in 23 States. Southwestern Indiana is a center of moderate activity. The central portion of western Ohio has experienced five strong shocks since 1875, most of which cracked some walls and toppled a small number of chimneys. Other centers of moderate activity are found scattered through Illinois and generally west of the Missouri River in eastern Nebraska and Kansas. Western Mountain Region Excluding Nevada, the principal seismic zone in this area extends in a broad band from Helena, Mont., southward to northern Arizona. Two destructive earthquakes centered close to Helena in 1935. In 1925 violent shocks in the region of Manhattan and Three Forks, Mont, cracked many buildings, toppled chimneys, and caused landslides that blocked the main line of the CM. and St. P. Railroad for many days. In 1959 an eight-state area and British Columbia felt the Hebgen Lake, Mont, earthquakes in which 28 people were killed by the great Madison Canyon slide. Surface damage extended 50 miles westward from Yellow- stone National Park and the estimated timber and road damage exceeded $11 million. In 1934 a strong shock centering near Kosmo, Utah, caused many craterlets to form, cracked walls and chimneys locally, and caused some plaster to fall in Salt Lake City. At Elsinore, 150 miles south of Salt Lake City, a series of strong shocks in 1921 damaged half the buildings, knocked down many chimneys, and wrecked a school building. In New Mexico a long series of disturbances near Socorro reached a climax in 1906 when a brief series of strong shocks cracked some adobe walls and toppled others. The Panhandle area of Texas has experienced a number of widespread but nondestructive shocks. In 1931, southeast of El Paso in the Great Bend 10 COAST AND GEODETIC SURVEY area, a violent shock centering near Valentine badly damaged all types of buildings, knocked down many chimneys, and was felt over the entire State. Large quantities of water were emitted from cracks and craters that formed in the ground. Across the border in the State of Sonora, Mex., an earthquake in 1887 destroyed the church at Bavispe and damaged many other buildings. Great numbers of cracks emitted sand and water and millions of cubic feet of rock were thrown down the mountains. Pacific Coast and Nevada Two-thirds of the seismic activity of the country is cen- tered in this area, most of it in the coast ranges of Calif- ornia. Some of the outstanding historic earthquakes in- clude one in 1812 in the Santa Barbara-Ventura-northern Los Angeles County region; one in 1838 along the San An- dreas Fault near San Francisco; another in 1857 in the northwest corner of Los Angeles County; one in 1872 in Owens Valley in the Sierras; and finally, the earthquake of 1906 which resulted in the destruction of a large part of San Francisco by fire. Other great earthquakes in this area include 2 that originated along the Haywards Fault just east of the southern half of San Francisco Bay; 2 sub- marine shocks off the northern coast of California; 2 in Nevada; and 1 each in the Los Angeles area, in Owens Valley, in Lower California, and in Imperial Valley. These were not necessarily the most notable earthquakes because most of them expended their violence either in isolated mountain and desert areas or on the ocean floor. It is a fact, now seldom realized, that when the 1838 earthquake struck the Bay area, San Francisco was little more than a scattering of huts and crude structures of the early pioneers. The California earthquake of April 18, 1906 was the most notable of all earthquakes in this country. Seven hundred lives were lost, and San Francisco was practically razed by fires that followed shortly after the earthquake. The city burned unchecked for days due to the wrecking of the water supply system. The actual earthquake damage probably did not exceed $24 million (1906 dollar value), but the fire increased this perhaps twenty fold. This scene of disaster EARTHQUAKE INVESTIGATION IN THE UNITED STATES 11 was duplicated on a smaller scale in many other California cities and towns. The 1906 earthquake was the result of a readjustment of basement rocks along the San Andreas Fault, a great rift that extends from Imperial Valley, just north of the Mexican border, northwestward through the coast ranges to Golden Gate. A short distance north of Golden Gate it passes out to sea through an elongated bay — Tomales Bay — that was apparently formed by repeated movements along the fault over a long period of years. The maximum horizontal slipping in 1906 was 21 feet. Vertical slipping was only a small fraction of this. Along the fault, fences and roads were sheared and separated many feet by the irresistible force of the earth movement. Structures built on the fault were wrenched and ruined. The following is a brief summary of other notable earth- quakes in the California-Nevada area. In 1868 almost every building in Hayward was damaged by a slip along the Hay- wards Fault, the damage in San Francisco reaching $350,000 (present-day value approximately $1 million). In 1872, at Lone Pine in Owens Valley, all adobe houses were wrecked and 27 persons out of a population of 300 were killed. In 1892 practically all brick buildings in Vacaville were wrecked. In 1899 nearly all brick buildings in San Jacinto and Hemet were damaged, only 2 chimneys being left intact in Hemet, and six persons were killed. The same towns experienced another shock in 1918 that wrecked all poorly constructed buildings in the business areas causing more than $200,000 damage. In 1920 Inglewood experienced a highly localized disturbance in which many buildings were wrecked and few escaped damage. In 1925 Santa Barbara suffered several million dollars damage. Few buildings on the main street — State Street — were left undamaged. This disturbance which killed 13 persons was caused by a fault slip in the mountain ranges just north of the city. In 1933, on March 10, 115 persons were killed and $40 million damage was caused by an earthquake that struck close to Long Beach. This was not potentially a great earthquake but the zone of greatest energy release seems to have been only a few miles away— in or near the Signal Hill area. On May 18, 1940, an earthquake slightly greater than that in Long Beach centered near Imperial in Imperial Valley, 12 COAST AND GEODETIC SURVEY causing $6 million damage and killing several people. It was featured by a horizontal fault movement of 15 feet. On June 30, 1941, another shock near Santa Barbara resulted in $100,000 damage, and on November 14 of the same year the Torrance-Gardena area of Los Angeles County suffered $1 million damage. In 1949 a minor disturbance centering on Terminal Island in San Pedro Bay sheared 200 oil wells near the 1,700-foot level causing more than $9 million damage. In 1951, 1955 and 1961, earthquakes barely perceptible at the surface caused additional subsurface damage ($3, 3, and 4.5 million, respectively) to oil wells on Terminal Island and adjacent mainland. On July 21, 1952, the strongest earth- quake to hit California since 1906 shook the area south of Bakersfield. The main shock and the series of strong after- shocks that followed killed 14 people and caused an estimated $60 million damage to buildings, railroads, and crops. It was caused by a slip along the White Wolf Fault, one of the more obscure faults that parallels the Garlock Fault about 20 miles northwest of the latter. The hardest hit towns were Arvin and Tehachapi although more damage occurred in Bakersfield because of its greater size. Bakers- field was badly shaken not only by the main shock but also by a strong aftershock that centered very close to the city on August 22. On December 21,1954, Eureka and Areata were violently shaken and nearly every building suffered structural damage. The reported damage was $2 million in Eureka and $100,000 in Areata. One fatality occurred in Eureka. On October 23, 1955, a strong shock between Con- cord and Walnut Creek caused the death of one person and $1 million property damage. The San Francisco Bay area earthquake on March 22, 1957 caused landslides, broke windows, and cracked highway pavements and a reinforced concrete water reservoir. Over-all damage amounted to $1 million. A swarm of earthquakes from July through December 1954 near Fallon, Nev. in the Stillwater and Clan Alpine mountains ranged in intensities from VI through X, and caused great surface ruptures, with damage to canals and drainage systems, irrigation structures, and paved highways. Since the area was sparsely settled, it experienced little property damage. Of particular interest was the 5-15 feet EARTHQUAKE INVESTIGATION IN THE UNITED STATES 13 of vertical movement in the Dixie Valley and east of Fairfield Peak, 6-20 feet of vertical and 4-12 feet of horizontal movements. The Puget Sound area has always been known to be mod- erately seismic with a record of a few hard jolts at long in- tervals, but an increase in activity during recent years was climaxed by a strong shock on April 13, 1949. More than $25 million damage was caused in cities bordering the Sound and several persons were killed. A shock in 1946 caused about $250,000 damage in Seattle, and several others were only slightly less severe. An unusually strong and widely felt shock in 1936 centered in northeastern Oregon in an area that was not thickly populated. The greater violence and damage associated with Pacific coast and Rocky Mountain earthquakes as compared with those in the East is generally attributed to their shallower foci. Most of the major rock fractures in California appear to be only 10 or 15 miles deep whereas in other areas this may be doubled or tripled, thereby causing less violent mo- tion at the surface. Alaska Alaska is one of the important links in the great seismic belt that circumscribes the Pacific. In the Aleutians, many submarine shocks occur in and near the deeps that parallel the islands on the south side. This belt moves inward on the Alaskan Peninsula, widens out in the region of the Kenai Peninsula, and extends northward into central Alaska, the region of Fairbanks marking the northern terminus. The Anchorage area is frequently shaken by earthquakes that would cause severe damage in a more thickly populated region. Just east of this active area is a quiescent zone but as the Canadian border is approached another active zone begins. This extends from the region around the southern half of the border to the southern portion of southeastern Alaska and on into British Columbia. Most of the shocks center inland. In 1899 the Yakutat Bay area near the southern tip of the Canadian boundary experienced one of the notable earthquakes of the last century. The shore was raised over a considerable length, and at one point there was a vertical 768-393 O — 65 3 14 COAST AND GEODETIC SURVEY fault slip of 49 feet — one of the greatest fault movements known. On July 9, 1958, a major shock, the strongest in southeastern Alaska since the Yakutat earthquake of 1899, was felt as far south as Seattle, Wash, and eastward to Whitehorse, Yukon Territory. Three persons were killed on Khantaak Island in Yakutat Bay, and a fishing boat with two persons was missing after being caught by a huge wave in Lituya Bay. The wave in Lituya Bay was reported to have reached a height of 1,740 feet, either by avalanche, wave action, or a combination of the two. The March 28, 1964, Prince William Sound earthquake, the strongest ever re- corded in North America, caused severe damage in An- chorage and other communities. The earthquake and ac- companying tsunami killed 115 persons and caused approx- imately $400 million damage. The tsunami was destructive in the Gulf of Alaska, along the Canadian coast, and as far south as Crescent City, Calif. The earthquake also caused seiche action in the rivers and coastal waterways along the Gulf coast of the United States. Hawaii Seismic activity in the islands centers on the island of Hawaii which is the largest as well as the farthest east in the group. Hawaii is best known for the spectacular erup- tions of Kilauea Volcano and the lava flows that frequently break from the slopes of 15,000-foot-high Mauna Loa. While there are an abundance of strong tremors directly associated with volcanic activity some of the heavier shocks that are felt at times over all the islands as far as Honolulu seem to be of tectonic origin. These epicenters trend gen- erally in a northeast-southwest direction and may originate on shore or beneath the sea. An earthquake in 1868 was extremely violent and destructive, considering the sparsely settled nature of the island. Hilo, on the northeast coast, frequently suffers severe earthquake damage. Strong sub- marine shocks off the southwest coast occasionally cause considerable damage nearby while those to the north and east may be felt with nearly equal violence on several islands. EARTHQUAKE INVESTIGATION IN THE UNITED STATES 15 Puerto Rico and Panama Canal Zone Puerto Rico is in the loop or belt of seismic activity that extends eastward through the West Indies from the Pacific. Many strong shocks in this area have their origins in ocean deeps. A violent shock in the deep off the northeastern coast of the Dominican Republic caused widespread destruc- tion in that country on August 4, 1946, and was strongly felt all over Puerto Rico. Violent shocks have also origi- nated near the islands just east of Puerto Rico. The most notable disturbance in recent years was that of October 11, 1918, originating in Mona Passage at the northwest corner of the island. The sea wave generated killed 116 persons and caused widespread damage in Mayaguez and nearby towns. Of the earlier shocks, that of 1867 was the strongest. While some moderately strong shocks have originated on the island itself, the greatest menace has been from those of submarine origin. Strong earthquakes do not originate in the Canal Zone nor in the portions of Panama immediately adjacent to it, but these areas are occasionally shaken by earthquakes originating in western Panama or northwestern Colombia. The main seismic belt of the Pacific bypasses the Canal Zone to the southwest, but submarine shocks from this area are sometimes felt strongly there. Seismic sea waves Submarine earthquakes occasionally generate seismic sea waves that travel thousands of miles over the oceans and cause great damage to shore property when they either pile up and break, or simply flood shoreline areas. Many ships have been carried ashore by such waves and thousands of lives have been lost, particularly in South America and Japan. On April 1, 1946, sea waves generated by a submarine earth- quake in the Aleutian Islands caused $25 million damage in the Hawaiian Islands. Hilo was hardest hit. A lighthouse high on a cliff a few miles from the epicenter off Unimak Island was washed away and its five occupants drowned. At sea, in deep water, destructive seismic sea waves are sel- dom detected by ships because the distances between crests 16 COAST AND GEODETIC SURVEY may be of the order of 100 miles and the wave height probably only 1 or 2 feet. It took the seismic sea waves of 1946 about 4% hours to travel from the Aleutian Islands, where the earthquake originated, to Hawaii. On March 9, 1957 another major earthquake, magnitude 8-8^, in the Andreanof Islands, Aleutian Islands, generated disastrous sea waves that caused extensive damage, about $3 million, on Oahu and Kauai, Hawaii, and on Adak and other nearby islands. A 40-foot wave surged into the area where the Scotch Cap Lighthouse was completely destroyed in 1946. Unusually large waves were reported by many of the islands of the Pacific, along the west coast of North and South America, and the Japanese Islands, but little or no damage occurred. The major tremor was accompanied by a swarm of more than 300 aftershocks with 14 having magni- tudes above 6%. The tsunami which originated off the coast of Chile on May 22, 1960, devastated the adjacent coastline, inflicting heavy loss of life and property damage. In Hawaii, 61 lives were lost at Hilo and there was about $22 million damage, although 6 hours of advanced warning was given by the Coast and Geodetic Survey's Seismic Sea Wave Warning System; in Chile, the earthquakes and tsunami caused more than 2,000 deaths and damage of $550 million; in Japan, 138 dead or missing and about $50 million in damages; in the Philippine Islands, 32 dead and missing; and along the west coast of the United States about $500,000 property damage. Unusually large waves were also reported on other Pacific islands. On March 28, 1964, an earthquake with a magni- tude of 8% in Prince William Sound, Alaska, generated destructive sea waves of 35 feet or more on shores bordering the Gulf of Alaska and inflicted heavy damages to Seward, Kodiak, Cordova, Whittier, Homer, and other communities. Damages were also heavy along the Canadian coast and at Crescent City, Calif. Of the 115 deaths which resulted, nearly all were caused by the tsunami. A seismic sea wave is due to a sudden rising or dropping of the ocean floor. In some areas like the northeastern coast of Japan, where destructive sea waves from nearby submarine shocks are frequent and destructive, the coastal populations race for the hills as soon as strong shocks are felt. This saves thousands of lives. EARTHQUAKE INVESTIGATION IN THE UNITED STATES 17 Nature of earthquakes and seismic waves Locating Earthquakes From Instrumental Data Strong earthquakes are usually due to the rupturing of great masses of rock many miles beneath the surface of the earth. This generally takes the form of slipping or sliding along a rupture plane, called a fault. The repeated occur- rence of earthquakes along the same fault over a long period of years is not unusual. Judging from the ground displace- ments observed after great earthquakes these fault slips may be as much as 50 feet as previously mentioned in connection with the Yakutat Bay, Alaska earthquake of 1899. In some seismic areas such as Japan, geodetic surveys often show that great blocks of rock, tens of miles in dimension, undergo tilting. The ultimate cause of this continual adjustment of crustal rock is thought to be due to a slow but relentless flowing or creeping of the underlying ultrabasic rock that has persisted through geological ages. These rocks behave as plastic material which, like a mass of paraffin, will change shape without fracture if the distorting forces are applied slowly and steadily enough, but will crack if they are applied too quickly. The cause of this rock flow is largely a matter of conjecture. It could be due to cooling of the earth, con- vection currents set up by radioactivity in the deeper rocks, changing loads on the surface due to the transfer of material through erosion and deposition, or other causes. The great majority of so-called shallow earthquakes origi- nate at depths of 10 or 20 miles but some are deeper, extend- ing to a maximum depth of 450 miles. Earthquakes due to volcanic activity may seem violent locally but they are never very deep or felt at great distances. Compared with tec- tonic earthquakes that are frequently registered on instru- ments all over the world, they are very superficial. Earthquakes cannot be predicted with respect to time and place except in terms that have little or no meaning. When the history of any seismic area is studied the pattern of earthquake occurrence is found to be so complex and erratic that it is quite impossible to derive from the data a formula that could be used to predict future disturbances. The gravitational pull of the planets is not sufficient to cause 18 COAST AND GEODETIC SURVEY earthquakes nor has it been demonstrated that such stresses are sufficient to always serve as trigger forces in setting off earthquakes that are just about ready to occur. In some places, large areas of terrain move slowly along with the creep or flow of the deep basement rock and when the pattern of this movement shows that some sections of the rock are " stuck' J and obviously building up stresses that will even- tually cause fracture, it is clear that the stage is set for an earthquake in that area. Such a condition exists along the San Andreas Fault in California where precise geodetic sur- veys have shown that since 1906, the terrain on one side of the fault has moved about 10 feet relative to the terrain on the other side. One must assume that sooner or later the stresses building up in between will again fracture the rocks along the fault; that the crustal rocks and overlying terrain will again slide along the fault or surface of fracture; and that in a matter of seconds a new state of stress equilibrium in the rock will be established. One might logically predict earthquakes in such areas but the time of occurrence will always be in doubt until the complete histories of a number of earthquake cycles are thoroughly known. In Japan, some successful predictions of strong shocks have been made through the discovery that certain changes in terrain and certain earthquake sequences form a pattern that seems to repeat itself after each great shock in the area under study. When great masses of rock are fractured, violent vibra- tions are radiated in all directions through the rock and overlying soil. It is the force of these vibrations that causes damage and frightens people in earthquake areas. On the lighter geological formations that overlay the ruptured basement rocks the force of the vibrations may be increased 5-, 10-, or 20-fold, increasing in severity with the lightness of the rock or soil and the height of the local water table. This accounts largely for the lack of uniformity in the damage experienced in a community shaken by an earthquake, but much, also, depends upon the sturdiness of individual buildings. In a great earthquake these seismic vibrations or waves penetrate the entire structure of the earth and travel all over its surface. While great earthquakes are seldom felt farther than a thousand miles from their source, sensitive EARTHQUAKE INVESTIGATION IN THE UNITED STATES 19 y 1 |; • I PP ZHZ PPP ; ■ j — SKStrPS I: ; ' ; ; SSZtSSS j ; ; ' Figure 3. — Earthquake waves recorded on typical seismogram and paths oj propagation through the earth. seismographs have registered these unfelt vibrations in all parts of the world for more than 50 years. Such seismic waves are extremely complex but a few basic facts will serve to explain how they are propagated through the earth and how the distance to an earthquake can be determined from a seismograph record. 20 COAST AND GEODETIC SURVEY Two types of waves travel at different speeds through the earth's interior and are known as interior waves. The faster one alternately compresses and dilates the rock as it travels forward; the slower one shakes the rock sidewise as it advances — like the vibration of a violin string. Seis- mological tables, based on many thousands of seismograph readings, show to the nearest second just how long it takes each of these wave groups to travel to points on the earth's surface at various great circle distances from an earthquake origin. Therefore, the difference in the arrival times of two such wave groups at a seismograph station corresponds to some particular epicentral distance that is shown in the seismological tables. These two waves are usually well defined on seismograph records and anyone who can recog- nize them can obtain the corresponding epicentral distance from the seismological tables. The largest waves recorded at distant stations, however, are usually waves that travel at nearly uniform speed along the surface of the earth only and are known as surface waves. Until recently, epicenters were located on large terrestrial globes by swinging arcs around several observatory locations, using as radii the epicentral distances determined from the station records. TRAVEL TIME CURVES TIME IN MINUTES Figure 4. — Travel-time curves with idealized seismograms superimposed. EARTHQUAKE INVESTIGATION IN THE UNITED STATES 21 Now they are being processed by modern electronic compu- ter techniques, thereby permitting the determination of many more earthquake locations in an equivalent amount of time. From about 250,000 earthquake messages sent annually by telegraph, radio, and other means to the Coast and Geo- detic Survey office in Rockville, Md., nearly 5,000 earth- quakes are located within a matter of a few hours or days after their occurrence depending upon the violence of the shocks and the number of reports received. The press and the public thus obtain immediate and authentic information on all of the stronger shocks. Seismologists throughout the world use the information for various types of scientific research and to insure correct interpretations of current seism ographic records. Besides registering earthquake vibrations, seismographs may also record ground vibrations from quarry blasts and other violent explosions. The distance at which such a dis- turbance may be recorded depends upon the strength of the blast, the distance to the recording station and the am- plifying power of the seismograph. Although records of blasts and local earthquakes are frequently recorded on seismographs, the average type of instrument does not furnish a good record of the high frequency waves gener- ated by them and interpretation is consequently difficult. The difficulty can be overcome by choosing a seismograph of a type that will give satisfactory recordings of high frequency waves. Seismic Sea Wave Warning Service One of the important services of the Coast and Geodetic Survey is the maintenance of a seismic sea wave warning program. The principal objective is to alert military instal- lations on Pacific islands and public officials in such areas as Hawaii, the west coast of the United States, and other Pacific countries and islands whenever seismographic records reveal the occurrence of an earthquake that might generate a destructive sea wave. Such a program would generally be impractical in areas near earthquake origins, but when 5 or 10 hours elapse between the time an earthquake occurs and the time sea waves might pile up on a distant 768-393 O— 65 4 22 COAST AND GEODETIC SURVEY Figure 5. — Geophy si cists oj the Coast and Geodetic Survey illustrating technique oj locating an earthquake on a 32-inch globe. shore, there is time, by working fast, to locate the earthquake, establish the existence of a sea wave, and issue warnings to coastal populations that might be endangered. In Hawaii, Alaska, Arizona, and Guam, the Survey oper- ates visible-recording seismographs that ring alarms when- ever an unusually strong shock is being registered. Ten other seismograph stations in various countries participate in this operation. Observers at 30 tide stations scattered throughout the Pacific immediately report unusual tidal dis- turbances to the montoring station near Honolulu. A high- priority communications service is maintained between reporting agencies through the combined facilities of the Defense Communication Agency, Federal Aviation Agency, and the National Aeronautics and Space Agency. With all EARTHQUAKE INVESTIGATION IN THE UNITED STATES 23 of these groups functioning, the Survey's central station near Honolulu is able to locate a submarine shock and verify the existence of a seismic sea wave within 2 or 3 hours. Seismographs Seismographs are basically nothing more than carefully constructed pendulums that appear to swing back and forth as the earth vibrates under them. Many seismograph pen- dulums take 10 or even 20 seconds to make a complete swing. The reason for this is that most waves from distant earth- quakes have long periods and long wave lengths and it requires a long-period pendulum to obtain a measurable difference between the moving pendulum and the moving ground. In nearby earthquakes, however, the ground vibra- tions are of short period and almost any kind of simple short- period pendulum would suffice. A long period is obtained by swinging the pendulum in a nearly horizontal plane about a nearly upright post, like a gate or door that is slightly tilted. It always takes a steady position in the direction in which the hinges are tilted. These are called "horizon- tal pendulums. " It requires two such pendulums, and another that measures vertical motion, to obtain the com- plete motion of the ground at a seismograph station. The second important requirement of a seismograph pen- dulum is that it must be damped to prevent it from swinging mostly in its own free period and obscuring other shorter and longer periods that are present in the ground motion. Some seismographs are damped with oil; that is, a vane or paddle fixed rigidly to the pendulum mass moves to and fro in a container full of oil whenever the pendulum moves, relative to the earth. If such a pendulum is pulled to one side one inch and then released, and if it overshoots its original position of rest by 0.1 inch, the damping ratio is said to be 10 to 1 which is about proper. Suet) a pendulum will give a good record of the ground motion. The great majority of modern seismographs, however, are not damped with oil but by a copper plate that passes closely between the poles of a powerful horseshoe magnet fixed to the instru- ment frame. 24 COAST AND GEODETIC SURVEY Figure 6. — Modern seismograph as used in the World-Wide Network of Standardized Stations. The third requirement of a seismograph pendulum is that its motion, relative to the ground, must usually be magnified many times by a mechanical lever system, by optical levers, or by electrical methods. Mechanical lever systems that magnify the pendulum motion from 10 to several hundred times are used on most older type seismographs. Such a lever in principle rotates horizontally on a pivoted vertical spindle set in jewel bearings that are fixed to the instrument frame. The short end is connected to the pendulum mass, while the long end that rotates through a much greater dis- tance carries a delicately mounted steel pin or stylus that scratches a fine line on a sheet of smoked paper that moves beneath it. In making records of this kind a strip of paper about a yard long and 6 inches wide is wound around a drum, then covered with a fine smoke film over a "sooty" flame and the drum set back on the seismograph recording mechanism. The drum usually turns through one revolution an hour and at the same time moves side wise in a spiral so that 24 lines EARTHQUAKE INVESTIGATION IN THE UNITED STATES 25 Figure 7. — An assembly for visible-recording seismograph. can be scratched on the paper without the lines overlapping. When an earthquake vibrates the ground beneath the stationary pendulum, a wavy line is scratched on the paper. This is the seismograph record and on it can be seen the principal seismic wave groups explained in the preceding section. Greater magnification is obtained by dispensing with cum- bersome mechanical levers and smoked paper and using beams of light that write fine lines on a sheet of photographic paper. This requires operating the seismographs in dark rooms that are characteristic of so many seismological sta- tions. Such instruments are designed so that a mirror in the pendulum system rotates with the pendulum; conse- quently, when a fixed beam of light is reflected from the mirror onto the photographic recording paper it traces a wavy line during an earthquake just as in the case of the smoked paper recorder. Most seismographs using such sys- tems magnify the pendulum motion one or two thousand times. 26 COAST AND GEODETIC SURVEY The greatest magnifications are ordinarily obtained with so-called electrical recording seismographs. In most of these a coil made of many turns of fine wire is fixed to the seismo- graph pendulum and forced by the earthquake motion to oscillate between the poles of powerful magnets fixed to the instrument frame. The current thus generated is sent through a sensitive galvanometer that makes a continuous record on photographic paper. With such instruments the ground motion can be magnified several hundred thousand times or more. Special seismographs made to record destructive earth- quake motions usually have very short-period pendulums. In fact, they are more like weighted springs. They record the acceleration or force of the motion on photographic paper. Instead of operating continuously like regular seismographs they remain inoperative until a strong earthquake starts them. A starting pendulum closes an electrical circuit that causes the entire recording mechanism to operate for about 1.5 minutes. Seismographs do not record the motion of the ground directly. The relationship between the actual ground move- ment and the movements shown on a seismograph record is very complex and will not be discussed here. It will suffice to say that when a seismograph is properly calibrated the seismologist knows just how much his instrument will mag- nify ground waves of all periods. This depends primarily on the pendulum period, the ground period, and a number of other factors, and explains why records of the same ground motion made with various type instruments can look so utterly different. Most important to the seismo- graph station director is the matter of keeping his time controls accurate to the nearest second. Earthquake waves pass the station at speeds as high as 5, 10, and 15 miles per second so that if he wants to determine earthquake distances from the arrival times of various wave groups he must be diligent with regard to time accuracy. In the United States about 150 seismograph stations are in continuous operation. (See Table 1, page 37.) Through- out the world there are approximately 900 stations. The World-Wide Network of Standardized Seismographs, which EARTHQUAKE INVESTIGATION IN THE UNITED STATES 27 28 COAST AND GEODETIC SURVEY was initiated in 1961, now has 108 operating stations. The records from this network are submitted to the Data Center in Washington where permanent file copies are made. At present, this Dat£ Center is reproducing over 200,000 copies of seismograms per month to fill requests of seismologists throughout the world. Descriptive reports on earthquakes One of the important functions of the Coast and Geodetic Survey is to collect descriptive information on earthquakes. This is done mostly from the San Francisco office of the Sur- vey because most earthquakes occur in the Pacific coast and western mountain regions. The information is obtained from thousands of cooperators who keep earthquake ques- tionnaire cards on hand, from postmasters and others in shaken areas who are canvassed immediately after an earth- quake, and from field surveys made by seismologists who hurry to the scenes of destructive earthquakes immediately after they occur. This information is valuable not only for statistical purposes but it furnishes important data on the earthquake intensity distribution in a shaken area. This is a complex phenomenon that is just beginning to be under- stood. The violence of an earthquake at a given distance from the epicenter depends not only on the strength of the earthquake at the focus but, as previously stated, largely upon local geological factors. In this country the Modified Mercalli Intensity Scale of 1931 is used to evaluate earthquake intensity. It separates the violence of earthquake motions into 12 different grades. Grade I is such a feeble vibration that persons observe it only under very favorable circumstances such as lying quiet- ly in bed with no distracting influences around. Grade VI represents earthquake motion that may cause superficial damage such as cracked windows, plaster, chimneys, etc. Grades XI and XII represent the most violent and destruc- tive kinds of earthquake motions — vibrations that will flat- ten whole villages, destroy or damage most buildings in a large city, and cause disastrous mountain slides. Strong earthquakes frequently force water from the ground and flood local areas. EARTHQUAKE INVESTIGATION IN THE UNITED STATES 29 In recent years there has been much confusion in the press and in the public mind concerning the definitions of earth- quake intensity and earthquake magnitude. Intensity is used to describe the severity or violence of an earthquake disturbance in any part of a shaken area and is based en- tirely on the earthquake eftects reported on people and on inanimate objects, including damage. Magnitude, on the other hand, depends on the amount of energy released at the focus of an earthquake and is determined from the am- plitudes of the ground vibrations registered at various seis- mological stations. Magnitudes are not reported in terms of energy but in terms of a mathematical function of the energy. The magnitude scale was developed about 27 years ago by seismologists of the Pasadena Seismological Laboratory of the California Institute of Technology. Engineering seismology The Survey's seismological program includes obtaining geophysical data needed by the structural engineer in designing safe structures in earthquake areas. The Figure 9. — Coast and Geodetic Survey type accelerograph . 30 COAST AND GEODETIC SURVEY compilation of descriptive information and statistics on damage as just described is essentially a part of this pro- gram but the nucleus of it is a network of special seismo- graph stations operating on the Pacific coast, at various points in the western mountain region, and at a few places in Central and South America. The accelerograph used to obtain records of destructive earthquake motions at these stations was briefly described in the section on Seismographs. Ninety-one such instruments now stand ready to record any strong or destructive earthquake motion in the areas named. An accelerograph record shows the high frequency tremors of an earthquake best. They are the waves that have the highest acceleration, that are most often felt, and that do the most damage. In terms of actual ground motion (displace- ment) they are the waves of smallest amplitude. All of the longer period waves have much larger amplitudes. In order to obtain a complete picture of the motion, the acceleration curve is put through a mathematical process called integra- tion which produces another curve that shows the velocity of the ground vibration at any instant (not the speed of the wave through the ground). When this new curve is inte- grated the result shows the true movement of the ground in inches. On this so-called displacement curve the high fre- quency waves are so small they are sometimes difficult to see, but the long-period waves are very clearly defined. Through this analytical process all of the periods and amplitudes in- volved in a complex earthquake motion cam be determined and studied in detail. Only recently has it been possible to associate the periods and amplitudes of destructive earthquake motions with the various grades of earthquake intensity according to the Modi- fied Mercalli Intensity Scale of 1931. This has important significance in scientific and engineering investigations. One of the engineers' difficult problems is to predict the motion, and consequently the stresses, in a building when it is subjected to a known earthquake motion. One approach is to assume that a building will respond to earthquake mo- tions in "somewhat" the same way as a simple pendulum or vibrator that has the same vibration period as the building under study. Experience has shown, however, that it is a EARTHQUAKE INVESTIGATION IN THE UNITED STATES 31 difficult problem to compute the response of even a simple vibrator to a complex earthquake motion unless costly com- puting machines are used. An interesting and simple de- vice used for this purpose is the torsion-pendulum analyzer. It has been found that when the torsion head of a simple torsion pendulum is turned through (positive and negative) angles that are proportional to the changing values of the linear ground acceleration (as registered on accelerographs), the mass of the torsion pendulum will turn through angles that are proportional to the linear displacements of the vibrator under study. In practice, the entire torsion pendu- lum operation is slowed down one or two hundred times so that the effect of the earthquake motion on the vibrator is reproduced in slow motion. Since it is clear that the natural periods of buildings and other structures play an important part in determining prob- able earthquake stresses it is necessary to measure tne nat- ural periods of many structures. To obtain this informa- tion the Coast and Geodetic Survey has made vibration measurements on more that 700 buildings and other struc- tures in the Pacific coast and western mountain regions. After enough earthquakes have occurred it will be possible to tell how building damage may be related to the period of a building and to use this information in designing new buildings. Since buildings are always in a state of vibration, mostly from wind, their periods can be measured by setting up portable seismographs on the top floors. A study of building damage reveals certain factual infor- mation that should be of interest to everyone concerned with the earthquake menace. To mitigate damage and possibly prevent injury and loss of life the following factors must be considered. First is the geological character of the ground itself. Soft ground, and especially fill, is definitely sus- ceptible to greater intensity of motion than rock outcrops in the same area. Obviously, special caution must be exercised in building structures on soft ground. More dangerous than this, however, is to build a structure partly on rock or hard ground and partly on soft ground, for then the differential motion between the two types of ground during an earth- quake could wrench and damage a structure as much as if it were built on an active fault. 32 COAST AND GEODETIC SURVEY EARTHQUAKE INVESTIGATION IN THE UNITED STATES 33 A great percentage of earthquake damage is undoubtedly caused by poor materials, especially mortar and plaster, that fail to hold up under earthquake forces. In many earth- quakes old school buildings have illustrated some of the worst examples of poor construction materials. Equally bad is failure in many cases to tie the walls of brick structures to the interior framework, a step that frequently results in the collapse of entire walls while the rest of the building re- mains virtually intact. This holds true also for cornices, parapet walls, and other parts of a building. Frame houses should be tied well to their foundations or they may slide off. The geometric pattern of a building is also important. Simple structures of rectangular cross section that sway as a unit without interference from other structures hold up best. When a building consists of two or more distinct parts that are not likely to vibrate in unison under seismic forces, the two parts tend to "hammer" each other, thus caus- ing excessive damage or even collapse. In 1946 during one of the great earthquakes of the century off the coast of the Dominican Republic, a priest on the island actually observed the battering action of the steeple on the auditorium section of his church and the eventual collapse of both. These are simple but basic rules about building in earth- quake areas that can be verified after almost any destructive shock. The truth is that after the strongest earthquakes ex- perienced in this country, one could find hundreds of build- ings showing little or no damage mainly because they were well-designed and well-constructed in the first place. What to do when an earthquake strikes A very practical question about earthquakes is, "What should one do in an earthquake to escape injury?" First, remember that during an earthquake there is probably less chance of being injured than when taking a weekend auto- mobile trip. About 1,500 persons have been killed in the United States by earthquakes and resulting tsunamis since the country was settled and 700 of these were lost in the California earthquake of 1906. Presumably, very few of these were killed by the actual collapse of the structures they were in as pictures taken immediate^ after that shock 34 COAST AND GEODETIC SURVEY show practically all buildings still standing, although dam- age was widespread. Most people were killed by rushing out-of-doors just as cornices, bricks, and other parts of build- ings were crashing to the sidewalks and streets. Others remaining inside were injured or killed by falling chunks of plaster. The lesson to be learned is to get as quickly as possible under doorways, arches, tables, desks, or anything that will be a protection from a falling ceiling; but above all, do not rush into the street, at least without first taking a quick glance overhead. If you are on a sidewalk get off it as quickly as possible, either into an open doorway or the middle of the street if it is a wide one. In many tropical countries the earthquake menace is made more serious by the widespread use of a weak lime mortar in building construction. Residents should reduce the haz- ard by observing building code provisions that have proved efficacious in other seismic areas. Tall, poorly braced frame structures with stone and lime mortar panels and walls are a particular menace, especially when built on water-soaked alluvial formations such as are so often found at the heads of bays and inlets. Officials in seismic areas bordering on the open ocean should also be aware of the possible danger from destructive seismic sea waves that may sweep over low-lying coastal areas fol- lowing a strong submarine shock. Residents of shore areas should rush to high ground following strong earthquakes in their area since no sea wave warnings can be issued through normal channels in time to protect residents near the area of generation of such waves. The most disconcerting feature of a destructive earth- quake is not necessarily the temporary panic and destruction that follow the principal shock but the repeated occurrence of aftershocks that fill the populace with a hopeless sense of insecurity. The victims should first realize that such aftershocks follow all major earthquakes and may in some cases continue for months, but with decreasing frequency and violence. Some of them will be quite sharp but it is very seldom that one ever reaches the intensity of the main shock. Aftershocks generally center at various points in the epi- central area, sometimes 15 miles or more from the source of EARTHQUAKE INVESTIGATION IN THE UNITED STATES 35 the main shock. No seriously damaged building or home, however, should be reoccupied without the approval of local authorities. A building damaged by the main shock is made weaker by each aftershock. It is a false philosophy which argues that a damaged building will withstand strong after- shocks just because it withstood the main shock. Experience shows otherwise. Microseisms The land surfaces of the world are no more in a state of absolute quiet than the water surfaces. Minute waves called microseisms are continuously moving through the rocks over the entire surface of the earth as can be seen by examining a sensitive seismograph record obtained in any part of the world. The most prominent and important waves have periods from about 4 to 7 seconds. In amplitude they range from approximately Ksooo inch to Kooo inch. There are others of much shorter period and smaller amplitude and of more localized character. These waves are all called microseisms. , % k— 2 Min. —I lv Figure 11. — Seismogram showing microseismic storm of De- cember 1 and 2, 1956, at Sitka, Alaska. 36 COAST AND GEODETIC SURVEY Microseismic waves are of meteorological origin just the same as ocean waves; in fact, ocean waves may play an im- portant part in their generation. Regardless of the mech- anism of their origin it is a fact that storms and low-pressure areas at sea are always accompanied by great increases in the amplitudes of microseismic waves recorded at seismo- graphic stations in the surrounding coastal areas. In the West Indies microseisms have been recorded from hurri- canes as far as 2,000 miles from the recording stations. Cold-wave fronts and other forms of meteorological changes cause various unique types of microseismic disturb- ances that have been the subject of much interest and investi- gation among seismologists. The basic concepts that will satisfactorily explain the generation and propagation of microseismic waves are yet to be advanced. EARTHQUAKE INVESTIGATION IN THE UNITED STATES 37 Appendix Table 1. — Active teleseismic stations of the United States COAST AND GEODETIC SURVEY AND COOPERATING STATIONS Albuquerque, N. Mex_ Boulder City, Nev Bozeman, Mont Butte, Mont Byrd, Antarctica Chicago, 111 College, Alaska Do Columbia, S.C Eureka, Nev Fayetteville, Ark. 1 Flaming Gorge, Utah_ Glen Canyon, Ariz Guam, Mariana Islands Honolulu, Hawaii Hungry Horse, Mont_ _ Kipapa, Hawaii Rapid City, S. Dak Salt Lake City, Utah__ San Juan, P.R Sitka, Alaska South Pole, Antarctica Tucson, Ariz Ukiah, Calif Washington, D.C Coast and Geodetic Survey Seismological Laboratory. Bureau of Reclamation. Montana State College. Montana School of Mines. Coast and Geodetic Survey Seismological Station. University of Chicago and U.S. Weather Bureau. Coast and Geodetic Survey Observatory. Coast and Geodetic Survey Seismological Outpost Station. University of South Carolina. Coast and Geodetic Survey Seismological Station. University of Arkansas. Bureau of Reclamation. Do. Coast and Geodetic Survey Observatory. Do. Bureau of Reclamation. Coast and Geodetic Survey Seismological Station. South Dakota State School of Mines and Technology. University of Utah. Coast and Geodetic Survey Observatory. Do. Coast and Geodetic Survey Seismological Station. Coast and Geodetic Survey Observatory. Coast and Geodetic Survey Latitude Observ- atory. Coast and Geodetic Survey Seismological Station. See footnotes at end of table . 38 COAST AND GEODETIC SURVEY Table 1. — Active teleseismic stations of the United States- Continued JESUIT SEISMOLOGICaL ASSOCIATION STATIONS Berlin, N.H Weston Observatory. Bloomington, Ind Indiana University and Saint versity. Louis Uni- Buffalo, N.Y Canisius College. Cape Girardeau, Mo___ Southeast Missouri State Teachers College and Saint Louis University. Caribou, Maine Weston Observatorv. Chicago, 111 Loyola University. Cincinnati, Ohio Xavier University. Cleveland, Ohio John Carroll University. Denver, Colo Regis College. Dubuque, Iowa Loras College and Saint Louis Universitv. Florissant, Mo Saint Louis University. Greenville, Mo Do. Little Rock, Ark Saint John's Seminary and University. Saint Louis Machias, Maine Weston Observatory. Manhattan, Kans Kansas State University and University. Saint Louis Milford, Ohio Xavier University. Milo, Maine Weston Observatory. New Orleans, La Loyola University. New York, N.Y Fordham University. Rochester, N.Y McQuaid Jesuit High School. Rolla, Mo. Missouri School of Mines and Metallurgy and Saint Louis University. Saint Louis, Mo Saint Louis University. Santa Clara, Calif University of Santa Clara. Spokane, Wash Mount Saint Michaels College University. of Gonzaga Spring Hill (Mobile Co.), Ala Spring Hill College. Washington, D.C Georgetown University. Weston, Mass Weston Observatory. See footnotes at end of table. EARTHQUAKE INVESTIGATION IN THE UNITED STATES 39 Table 1. — Active teleseismic stations of the United States — Continued PASADENA SEISMOLOGICAL LABORATORY STATIONS (CALIFORNIA INSTITUTE OF TECHNOLOGY) 3 Barrett, Calif China Lake, Calif El Centro, Calif Fort Tejon, Calif Goldstone, Calif Haiwee, Calif Hayfield, Calif Isabella, Calif King Ranch, Calif. _ Mount Wilson, Calif Palomar, Calif Pasadena, Calif Riverside, Calif San Nicolas Island, Calif. Santa Barbara, Calif Tinemaha, Calif Woody, Calif Barrett Reservoir, City of San Diego. Naval Ordnance Test Station, Inyokern. Imperial Irrigation District. Division of Beaches and Parks, State of California. Jet Propulsion Laboratory. Haiwee Reservoir, Bureau of Water Works and Supply, City of Los Angeles. Metropolitan Water District, Los Angeles. U.S. Army Corps of Engineers, Sacramento. E. R. King Ranch, McKittrick. Mount Wilson Observatory. Palomar Observatory. Kresge and Donnelley Laboratories. City of Riverside. United States Navy. Santa Barbara Museum of Natural History. Bureau of Water Works and Supply, City of Los Angeles. Kern County Forestry and Fire Department. See footnotes at end of table. 40 COAST AND GEODETIC SURVEY Table 1. — Active teleseismic stations of the United States- Continued UNIVERSITY OF CALIFORNIA STATIONS 3 Areata, Calif Humboldt State College. Berkeley (Haviland), University of California. Calif. Byerly (Berkeley Straw- Do. berry), Calif. Concord, Calif. _ Diablo Valley College. Ferndale, Calif __ City of Ferndale Fresno, Calif Fresno City College. Jamestown, Calif California State Department of Water Resources. Llanada, Calif University of California. Manzanita Lake, Calif __ Lassen Volcanic National Park. Mineral, Calif ___ Do. Mount Hamilton, Calif _ Lick Observatory. Oroville, Calif. __ California State Department of Water Resources. Palo Alto, Calif _ Stanford University. Paraiso, Calif University of California. Priest, Calif Do. Reno, Nev University of Nevada. San Francisco, Calif University of San Francisco. Santa Cruz, Calif University of California. Vineyard (Local), Calif _. W. A. Taylor Winery. Vineyard (Telemeter), University of California. Calif. INDEPENDENT STATIONS Ann Arbor, Mich Atlanta, Ga Baker, Oreg Balboa Heights, C.Z. 2 Bellingham, Wash Bermuda Blacksburg, Va Bowling Green, Ohio. University of Michigan. Georgia Institute of Technology. Blue Mountains Observatory, Vela-Uniform Observatory ; Technical Direction- A FT AC. The Panama Canal Company. Western Washington State College. Lamont Geological Observatory of Columbia University. Virginia Polytechnic Institute. Bowling Green State University. See footnotes at end of table. EARTHQUAKE INVESTIGATION IN THE UNITED STATES 41 Table 1. — Active teleseismic stations of the United States — Continued independent stations — continued Canyon Junction, Wyo_ U.S. Geological Survey. Chapel Hill, N.C Univeraii/y of North Carolina. Corvallis, Oreg Oregon State University. Dallas, Tex Southern Methodist University. Dugway, Utah University of Utah. El Paso, Tex Texas Western College of the University of Texas. Fort Sill, Okla Wichita Mountains Observatary, Vela-Uniform Observatory ; Technical Direction — AFTAC. Glen Cove, N.Y Private Station. Victor Aiello Golden, Colo Colorado School of Mines. Hanover, N.H Dartmouth College. Hawaiian Volcano Ob- servatory U.S. Geological Survey. Honolulu, Hawaii University of Hawaii, Institute of Geophysics. Do University of Hawaii, Manoa Valley. Houston, Tex Mandrel Industries, Inc., Electro-Technical Labs Division. Do William Marsh Rice University. Junction, Tex Texas A&M University. Klamath Falls, Oreg Oregon Technical Institute. Laramie, Wyo University of Wyoming. Lawrence, Kans University of Kansas. Longmire, Wash University of Washington. Lubbock, Tex Texas Technological College. Madison, Wis University of Wisconsin. Mammoth Hot Springs, Wyo Do. McMinnville, Tenn Cumberland Plateau Observatory, Vela-Uniform Observatory ; Technical Direction — AFTAC. Minneapolis, Minn University of Minnesota. Morgantown, W. Va West Virginia Universitv. New York, N.Y City College of New York. North Reno, Nev University of Nevada. Ogdensburg, N.J Lamout Geological Observatory of Columbia University. Oxford, Miss Universitv of Mississippi. See footnotes at end of table. 42 COAST AND GEODETIC SURVEY Table 1. — Active teleseismic stations of the United States Continued independent stations — continued Palisades, N.Y Payson, Ariz Philadelphia, Pa. 2 ___ Price, Utah San Diego, Calif Seattle, Wash Seattle, Wash Socorro, N. Mex State College, Pa Sterling Forest, N.Y Sunnyside, Utah Terre Haute, Ind Tonopah, Nev Troy, N.Y Tulsa, Okla Tumwater, Wash Unionville, Nev Vernal, Utah Waterville, Maine. _ Lamont Geological Observatory of Columbia University. Ton to Forest Observatory, Vela-Uniform Observatory; Technical Direction — AFT AC. The Franklin Institute. Carbon College of the University of Utah. San Diego State College. Private Station. Gerald Marshall. University of Washington. New Mexico Institute of Mining and Tech- nology. Pennsylvania State College. Lamont Geological Observatory of Columbia University. U.S. Geological Survey. Private Station. Gerald Shea. University of Nevada. Rensselaer Polytechnic Institute. Jersey Production Research Co. University of Washington. University of Nevada. Unita Basin Observatory, Vela- Uniform Observatory; Technical Direction — AFT AC. Colby College. 1 Publishes own station bulletin. 2 Results published by the Coast and Geodetic Survey. 3 Correspondence with reference to the Pasadena and Berkeley network of stations should be addressed to: Seismological Laboratory, 220 North San Rafael Ave., Pasadena, Calif, or to Seismographic Stations, University of California, Berkeley, Calif. EARTHQUAKE INVESTIGATION IN THE UNITED STATES 43 Table 2. — Lives lost in major United States earthquakes 1811 1812 1868 1872 1886 1899 1906 1915 1918 1925 1933 1934 1935 1940 1946 1949 1952 1954 1955 1958 1959 1960 1964 New Madrid, Mo San Juan Capistrano, Calif Hayward, Calif Owens Valley, Calif Charleston, S.C San Jacinto, Calif San Francisco, Calif Imperial Valley, Calif Puerto Rico (killed by sea wave from earthquake in Mona Passage) Santa Barbara, Calif Long Beach, Calif Kosmo, Utah Helena, Mont Imperial Valley, Calif Hawaii (killed by sea wave from earthquake in Aleutian Islands) Puget Sound, Wash Kern County, Calif Eureka- Areata, Calif Oakland, Calif Khantaak Island and Lituya Bay, Alaska Hebgen Lake, Mont Hilo, Hawaii (killed by sea wave from earthquake off the coast of Chile) Prince William Sound, Alaska (sea wave claimed nearly all of those killed along Gulf of Alaska and west coast of the United States) Several 40 30 27 60 6 700 6 116 13 115 2 4 9 173 8 14 1 1 5 28 61 115 44 COAST AND GEODETIC SURVEY Table 3. — Damage caused by major United States earthquakes 1865 1868 1872 1886 1892 1898 1906 1915 1918 1925 1933 1935 1940 1941 1944 1946 1949 1949 1951 1952 1954 1955 1955 1957 1957 1959 1960 1961 1964 San Francisco, Calif San Francisco, Calif Owens Valley, Calif Charleston, S.C Vacaville, Calif Mare Island, Calif San Francisco, Calif Fire loss Imperial Valley, Calif Puerto Rico (sea wave damage from earthquake in Mona Passage) Santa Barbara, Calif Long Beach, Calif Helena, Mont Imperial Valley, Calif Torrance-Gardena area, Calif Cornwall, Ontario; Massena, N.Y Hawaii (sea wave damage from earth- quake in Aleutian Islands) Puget Sound, Wash Terminal Island (oil wells only) __.Do Kern County, Calif Eureka- Areata, Calif Terminal Island (oil wells only) Oakland- Walnut Creek, Calif Hawaii (sea wave damage from earth- quake in the Aleutian Islands) San Francisco, Calif Hebgen Lake, Mont, (damage to timber and roads) Hawaii and the west coast of the United States (sea wave damage from earth- quake off the coast of Chile) Terminal Island (oil wells only) Prince William Sound, Alaska (sea wave damage in Crescent City, Calif, not included) $500,000 350,000 250,000 23,000,000 225,000 1,400,000 24,000,000 500,000,000 900,000 4,000,000 8,000,000 40,000,000 4,000,000 6,000,000 1,000,000 2,000,000 25,000,000 25,000,000 9,000,000 3,000,000 60,000,000 2,100,000 3,000,000 1,000,000 3,000,000 1,000,000 11,000,000 25,500,000 4,500,000 400,000,000 EARTHQUAKE INVESTIGATION IN THE UNITED STATES 45 SEISMOLOGICAL PUBLICATIONS OF THE COAST AND GEODETIC SURVEY The following publications are available from the Super- intendent of Documents, Government Printing Office, Wash- ington D.C., 20402. Earthquake History of the United States, Part I — Continental United States (Exclusive of California and Western Nevada) and Alaska, No. 41-1, Revised (1956) Edition. This is a descriptive catalog of the stronger shocks of historical record through 1956. Price 60 cents. Earthquake History of the United States, Part II — Stronger Earth- quakes of California and Western Nevada, No. 41-1, Revised (1960) Edition. This is a 55-page summary of the stronger earthquakes of the region. New data on many of the earlier shocks are included. Price 40 cents. Earthquake Investigations in California, 1934-35, Special Publi- cation 201. This is primarily of engineering interest and is rather technical as it concerns the precise measurement of ground and build- ing motions resulting from natural and artificial causes, including earthquakes. A large quantity of observational data is given in- cluding data on normal and forced vibrations of buildings, elevated water tanks, bridges, and other structures; also data on ground vibra- tions. (Out of print.) Principles Underlying the Interpretation of Seismograms, Special Publication 254. This is designed to meet the needs of the student and to explain seismic waves and earth structure, the response of seismographs to seismic waves, travel-time tables and charts, inter- pretation of seismograms, and other items such as intensity, magni- tude, and nomenclature. Price $1.50. TSUNAMI! The Story of the Seismic Sea Wave Warning System. This 46-page brochure essentially describes how the Warning System operates. It includes the methods employed to warn people of a possible tsunami and outlines cautionary steps to be taken when the tsunami alarm is given. Price 35 cents. The following processed reports are available from the Director, Coast and Geodetic Survey, Rockville, Md., 20852. United States Earthquakes series, the annual seismological report of the Bureau. These are essentially statistical catalogs (available from 1946) of earthquakes felt over the entire country and outlying territories. A summary of in stru mentally determined earthquake locations is given in the later editions. Beginning with the 1933 report a section has been added on results of the strong-motion earth- quake program of the Bureau. (Prices furnished on request.) The Seismograph and the Seismograph Station. — This is written pri- marily for use in the selection of seismographs and station sites for the recording of earthquakes. It briefly treats on factors governing site selection; choice of instruments, including seismometers, galva- 46 COAST AND GEODETIC SURVEY nometers and recorders; time control; telemetering seismic signals; and amateur seismology. Price 30 cents. Seismological Bulletin. — This pamphlet contains the seismogram interpretations for all Coast and Geodetic Survey and cooperating stations and gives resulting epicenter locations. Issued monthly on mailing list CGS-7. Preliminary Determination of Epicenter Cards. — These postal cards give epicenter locations of strong shocks occurring throughout the world, and are issued twice weekly and within two weeks after occur- rence of shocks. Abstracts of Earthquake Reports for the Pacific Coast and the Western Mountain Region. — These quarterly reports contain the unabridged information that has been collected from thousands of volunteers and other observers who furnish information on felt earthquakes. Issued on mailing list CGS-3. Quarterly Engineering Seismology Bulletin. — This leaflet contains brief summaries of work connected with the engineering phase of seismology and preliminary analyses of all accelerograph records ob- tained during the period. Issued on mailing list CGS-5. Earthquake Data Report. — This report contains data used in the computation of the "Preliminary Determination of Epicenters" and is issued twice weekly to research institutions and individuals having need for the station arrival times, individual distances, azimuth, and travel time residuals. U.S. GOVERNMENT PRINTING OFFICE: 1965 O— 768-393