By Oded Navon and Mordechai Stein
The measurement of time is no simple matter. Today we have at our disposal many means to measure and determine time. Clocks and calendars, a computer which notes the time and newspapers which note the date. What would we do without all of these? We could estimate the hour by the sun, or the day of the month by the size of the moon, and the months by the stars and the zodiac. But how would we know what the year is? We need a system of notation. Without some way of keeping track of the years we cannot know. In the past they would count the years of kings' reigns: "In the 27th year of Asa, king of Judea, Zimri reigned in Tirzah for seven days" (I Kings 16:15) or "In the first year of Koresh the king of Persia" (II Chronicles 36:22). Even today we count from significant events: the creation of the world, the birth ofJesus, or Mohammed's flight from Mecca to Medina.
Are years also listed in other ways, beyond the historical record? Nature deals with yearly cycles. Storks fly to Africa in the fall and return north in the spring. The almond blossoms every winter. But while the stork does not leave any note, the almond does list its years. The annual cycle of pace and the manner in which the tree grows leads to the formation of annual growth rings. The number of rings are determined by the number of years the tree has been growing, and one can even note particularly bad years based on the width and shape of the rings (illustration 1). If we cut the tree down we could determine its age based on the number of rings. If we compare the set of rings, including thin rings from years with little rainfall and thicker rings from years with abundant rainfall, to a tree of the same type, from the same area, which died while the first tree was still growing, we can determine when the second tree died and what its age was (see, for example, illustration 1). Thus, by comparing rings we can count backwards in time, as long as we can find a basis for comparison between sets of rings. The oldest trees known to us, the Bristlecone pines, grow in the California White Mountains. The oldest of these trees, nicknamed Methuselah, is 4770 years old! By comparing sets of rings from ancient trees which have grown at various times and partially overlap, we can account for all 11,000 years before our time!
Illustration 1. An example of dating via tree rings. The trunk shown schematically on the right was cut down in 2006.Its trunk shows alternating light rings, showing rapid spring and summer growth, and dark rings forming during the slow winter growth. The rings are dated consecutively and show that the tree was planted in 1989. In 2003 the tree grew only slightly and the light ring is narrow; thus, too, in the years 1994-1996. The tree on the left was found on the ground, but comparing the thickness of rings to those of the tree on the right shows the sane three thin rings. The graph of annual ring thickness shows a close correlation between changes of ring growth in both trees (purple squares--right tree, blue squares--left tree). Based on this correlation we can determine when the second tree was planted and when it was cut down (1982-2000) and the continued notation of better and worse years backwards in time. Trees which live for hundreds of years allow precise correlation and determination of time back 11,000 years from the present.
But it is not only trees which note years. Rocks, too, do so. One method of notation is the sediment layers embedded in lakes. The difference in seasonal conditions causes cyclical sedimentation in summer and winter. For example, a lake in the Alps may accumulate shale in the spring when the snows melt, while in summer algae with marl skeletons grow and sink to the ground. Thus, over the years annual layers of shale and marl form. With time, pillars of rock, built of layers, form. One can now count the layers and estimate the number of years it took to form the rock. In many lakes there are dozens and even hundreds of layers. In Israel, too, in today's Dead Sea area and the Jordan valley, there is a lake called Lisan Lake, in which sets of brown shale and white marl accumulated. Each winter streams bring silt and marl dissolved in water. The mud mainly settles during or by the end of the winter and the marl settles in the summer, following a longer process of evaporation and blending. At the foot of Masada or in the area of the Flour Cave one can see the sediment of the Lisan Lake (illustration 2). Tens of thousands of layers of sedimentation are exposed there, testament to a lake which was in the area some 14,000 to 70,000 years before our time.
Illustration 2. Layers and ages in rock from the Lisan Lake in the Perazim gorge. A. A cliff on the all of the Perazimgorge. The cliff shows a cross-section some 40 meters tall of Lisan origin, across-section which is composed of tens of thousand of layer sets. B. A close-up of the intercalation of dark slate layers, rich in minerals, and the light marl layers, rich in the mineral aragonite (thickness of each layer is 1-2 millimeters). Sets of layers show sedimentation throughout the year. Stream water brings mud and dissolved marl all winter long. The mud sinks by the end of the winter and the aragonite sinks later on as a result of evaporation of the top layer of lake water in the summer and the blending of the top layer of water and the deeper water. C. Age of sample compared to their height in the cross-section. The samples were taken from a cross-section which passes through the height of the cliff, approximately where the Y-axis of the graph is. The age of aragonite sedimentation is determined through a system of radioactive decay of uranium-thorium. The age of the samples is less as one ascends the cross-section. The layer at the base of the cliff settled some 67,000 years ago. The large jump in age to the next sample shows a gap in the cross-section. (Either no material settled or the area was exposed at the end of the missing period and the cross-section of that sedimentation is missing.) Between years 60,000 and 50,000 the sedimentation rate is steady (uniform slant on the graph). Between years 50,000 and 41,000 the cross-section is missing, as it is for years 35,000-31,000. For the rest of the time the layers settled in generally uniform sets. The yellow diamonds represent the age of samples as compared to their height in the cross-section. Yellow lines represent deviation in the ages of samples. The blue line represents similar sedimentation rates in different time periods. A manual count of layer sets shows a strong correlation between the count of layers and the age of samples which have been dated.
Cave deposits, such as stalagmites and stalactites, also note the passing of the years. The rate of dripping and of evaporation changes between summer and winter and the accumulation of marl on the stalagmite forms annual rings. Cave deposits and lake sedimentation accumulate data on different climate eras, for example, rainy periods or dry periods. One can identify climactic changes which parallel the cave deposits, lake sedimentation, and tree growth rings. These independent methods show similar eras for these events and exemplify the (albeit imperfect) precision of the three methods.
Nature allows us another way to measure time--radiometric dating. Radiometric dating is based on the existence of unstable atomic nuclei. Most atoms which make up our world have stable nuclei. But there are nuclei which are unstable, like those of uranium and radon, which are the best known. From time to time this sort of nucleus destabilizes and forms a different nucleus. This destabilization is called radioactive decay because the amount of the element decays in a process accompanied by emission of energy (radio=rays, active=active; radioactive element=element which emits rays of energy).
The process of decay is random, with each type of unstable nucleus having a given likelihood of decay. For example, when we check the rate of decay in radon nuclei, we find that it depends only on the number of nuclei in the sample.
Number of radioactive nuclei in sample
rate of decay
number of nuclei which decay in a specific period of time
Despite the random nature of the process, because there are many atoms (some one thousand billion billion, or some 1021 atoms in a gram) the statistics are amazing and the rate of decay is known with astonishing precision. Let us note that the number of nuclei which decay in any given period of time is not uniform. It depends on the number of nuclei in the sample. With the passing of time and the decline in the number of nuclei, the rate of decay decreases. Determining the rate of decay involves determining the relative share of nuclei which decay in any given period of time.
The process is like what happens to money deposited in a Swiss bank, where the interest rate is negative. Each year bank charges cost 1% of the deposited amount. If we deposit 1,000 francs, the bank will charge our account for 10 francs the first year, but only 9.90 francs the second (1% of the remaining 990 francs). Because the rate of decay decreases with time, we are left with half our initial deposit not after 50 years, but after 69 years. In the 70th year the bank will charge us five of the 500 francs left, so after 69 more years we will still have 250 francs in our account, and so on: in the 217st year, after an additional 69 years, we will have 125 francs, and in the 276th year we will have only three francs (illustration 3). Each 69 years we will be left with half of our money.This time is called by scientists "half-life." (From this example it should be clear that we do not mean the money will be finished after twice the half period. After a half-life we will be left with half the amount we started out with. A better phrase would be "half life-time.") Our account will remain active for a great number of years, but after 10 half-lives (some 690 years) we will be left with less than one franc, less than 1/1000 of the amount we deposited. Another 10 half-lives are required to leave the account with only 1/1000th of a franc.
Illustration 3. The balance on a deposit account with negative interest. Each year bank charges amount to 1% of the balance in the account. The balance falls to half each 69 years. A similar graph describes the number of radioactive atoms in a rock sample. The number decays to half in a specific period of time called a "half-life." The relationship between the amount left in the account (the current number of radioactive atoms) and the original deposit (the beginning number of radioactive atoms at the time the rock was formed) determines how much time has passed (for example, in half-life units) and permits a determination of the age of the rock.
We have a good way to count the years: Tell me how much money you deposited, the interest rate, and how much is left in the account and I can tell you how long ago you deposited the money. That's the same principle as radiometric dating. We have to know how many radioactive nuclei were in the sample when it was formed, how many are now left, and by means of half-lives or determining the rate of decay (the negative interest in the bank account) we can determine the age of the sample. First we will demonstrate this principle using a system based on decay of the radioactive nucleus of carbon--carbon 14. In nature, the element carbon has three isotopes (three types of carbon atoms): carbon-12, carbon-13, and carbon-14. The three isotopes have an identical number of electrons, six, and they all have very similar chemical characteristics. But the nucleus of each of the three is different. Each has six protons, but carbon-12 has six neutrons, carbon-13 has seven, and carbon-14 has eight neutrons. The first two isotopesare stable and do not decay. The nucleus of carbon-14 is unstable. There is a known probability that one of the neutrons in this atom will emit an electron and become a proton. With seven protons and seven neutrons, the atom becomes an atom of nitrogen-14, similar to the nitrogen in the atmosphere.
We can measure the decay rate from carbon-14 to nitrogen-14 in a laboratory. Each such disintegration emits a typical radiation, which can be measured. If we measure the decay rate in a fresh tree sample, we will find that in a sample consisting of one gram of carbon we will find some 15 disintegrations per minute. Using a different instrument, a mass spectrometer which separates carbon atoms based on their mass, we can also measure how many atoms of each isotope we have, and we will find that one of every 780 billion carbon atoms is an atom of carbon-14. After we ascertain their number by multiple measurements, for example from different plants or the carbon found in carbon dioxide in the air, we can calculate the rate of decay. The value we get is one disintegration per minute for all 4 billion carbon-14 atoms (3.89394X10-12 decays per second). We can also translate that rate to a half-life of 5730 years. That means that the number of carbon-14 nuclei in a sample of charcoal is reduced by half each 5730 years.
If so, why doesn't carbon-14 decay and disappear? What causes its appearance in carbon in the atmosphere or in flora and fauna? Carbon-14 is created in the atmosphere from the collision of powerful particles which come from space and which thoroughly mix with carbon already in the atmosphere. Since it is simultaneously being created and decaying, and since the rate of decay depends on the number of nuclei existing, a balance is created (rate of creation=rate of decay), so the ratio between the number of carbon-14 atoms and all carbon atoms in the atmosphere (1 of each 780 billion). All carbon atoms are absorbed by plants during photosynthesis, so the relationship between isotopes in plants now growing is set and nearly identical to the relationship in the atmosphere. When the plant dies there is no more absorption of carbon-14 from the atmosphere, only radioactive decay. The concentration (relative to the stable isotope of carbon) decreases, with a half life of 5730 years. We know how much carbon-14 there was at the beginning (the initial deposit into the Swiss bank account) and the rate of decay (the negative interest charged by the bank). If we measure how much carbon-14 is in the sample now, we can determine the date on which photosynthesis ceased (for example, charcoal from a tree killed in a blaze, or a scroll made from harvested papyrus). After several half-lives, very few carbon-14 atoms will be left, and errors in determining age will increase. A significant error will also be found for a tree which died only a few dozen years ago, because only a very small amount has had time to decay. Therefore this method is best for dating remains whose age is between hundreds of years and some 50,000 years.
Carbon-14 dating is verified through various and sundry means.
- The dating of documents whose age is known gives an accurate age. For example, an age of 2,000 yearsfor a document from the Roman era. (The time between the harvesting of the tree and the production of paper is generally negligible.)
- Separate dating ofremains of trees felled and burned in volcanic eruptions (all were covered in the same layer of volcanic ash) gives uniform dates in a narrow range of uncertain measurement. (For example, all fell within a range of years 12,900 to 12,500).
- Dating remains of trees which were felled in a later incident (for example, a volcanic eruption; we know this because volcanic ash from this incident [including the remains of trees in the ash] are in a layer above those of an earlier incident) give us a younger date (for example, 9700±300 years).
- The difference between carbon-14 dating of organic remains which sunk in lakes in which one can countlayers and the number of layers which divide them. For example, in the sedimentof Lisan Lake near the Perazim gorge of the Dead Sea one can find remains of trees which were swept into the lake and in the end, saturated with water, sunk to the lake floor. Dating of carbon-14 matches the dating of the sedimentation layers. Thus we can move from relative dating: "This layer is found some 7500±20 layers above another layer" to absolute dating: "This tree died 22,000±500 years ago and the tree we found is 7500 layers below it died 29,100±300 years ago." Note that the difference in ages determined by the two methods is comparable, considering the standard deviation of measurement and that there is a particular uncertainty in counting layers.
- Relevant eras can be dated using another method of dating, based on the series of isotopes formed during the radioactive decay of uranium. The ages calculated using this method are similar to those found by counting layers or by carbon-14 dating.
The main problem in determining exact dates using this method lies in the assumption that the concentration of carbon-14 in the atmosphere is steady. Changes in solar activity, in the magnetic field of the earth, and in the cycle of carbon in nature lead to changes in the atmospheric concentration and thus in the base concentration in a living tree. These changes are too small to disturb the correlation between different dating methods. Comparisons made in recent years between the methods has increased the precision in translating ratios of isotopes in organic remains and years before the present date.
To date more ancient stones, we make use of isotopes whose half-life is longer. We have access to a number of isotope systems and in this essay we will present two of them. The chemical element potassium has a stable isotope, potassium-39, and an unstable isotope, potassium-40, which decays into an isotope of another element, argon-40. Potassium is found in the earth's crust and is part of many common minerals. Argon, on the other hand, is a noble gas found in the atmosphere; it is found in only minute quantities within the planet Earth but is not part of minerals as formed. As magma cools, crystal grow in it, including crystals containing potassium. In newly-formed crystals there is no argon, but as time passes, we are left with all the potassium-39 atoms (the stable isotope) which were originally in the sample, but some of the potassium-40 atoms have decayed into atoms of argon-40. We know how many atoms of potassium-40 we now have in the sample (how much money is left in the account), but we do not know how much was at the start (the initial deposit). But all atoms of argon-40 were created from decay and have remained trapped in the matrix where the atoms of potassium-40 once were, before the decay. We can free them, collect them, and count them. Using the bank example, the atoms of argon which have been created are the money which the bank took from our account. We know how much money the bank transferred from our account to its own. Thus we can calculate the starting number of potassium-40 atoms in the sample (how much money is left in our account + the money taken by the bank) and by setting the decay rate, known to us through laboratory experimentation with samples of potassium (the negative interest), we can calculate the age of the sample.
Potassium decays quite slowly. Its half-life is 1.3 billion years. Determining the concentration of potassium-40 and argon-40 in samples of basalt from the Golan Heights tells us they were formed several million years ago. The basalt which flowed did so atop those which preceded them. If we date the different flows, we will see that the lower basalt is older. In the cliffs of the Zavitan River in the Golan a number of basalt flows are exposed, one above the other. Samples from four flows were dated using isotopes of potassium and argon. The age of a sample from the lowest flow was 3.3 million years, a sample from a higher flow was some 3 million years, a sample from the middle of the cliff is 2.3 million years old, and the last, from one of the highest flows, is 1.8 million years old. A flow built up on an area outside the Zavitan Mall is only 700,000 years old. The correlation between results of the dating and the height of the flows of the cliff testifies to the veracity of the results. This method is quite accepted for the dating of basalt and other rocks formed from the crystallization of magma, and it has given reliable results in hundreds and thousands of places in the world. We can use it to date rocks which are millions and even billions of years old. Young samples (100,000 years or less) of potassium-rich rocks, in which an amount of argon sufficient for measurement are formed in a relatively short time, can be dated.
The element uranium has two common isotopes, uranium-235 and uranium-238. The half-life of the first is 700 million years and of the second 4.5 billion years. Uranium decays into lead in a series of radioactive disintegrations during which alpha particles (containing two protons and two neutrons) or beta particles (electrons) are emitted. Each disintegration forms a different isotope which is also radioactive and continues to decay with a relatively short half-life (between a thousandth of a second and some 250,000 years) until tellurium-207 decays into a stable isotope: lead-207. (In a series which begins with uranium-238, plutonium-210 decays into lead-206.) We can also measure, in a laboratory, the concentration of suitable isotopes of uranium and lead and use them to date rocks. But lead presents a new problem: as opposed to argon, it is likely to become fused in the matrix of minerals as it is formed. We cannot distinguish between lead-207 formed from the decay of uranium in a mineral and that which was part of the mineral from its formation (the bank put money from our account into a general account with moneys from other people). Since uranium and lead have different chemical properties, we can use minerals which prefer uranium over lead so absolutely that we can ignore any pre-existing lead in the face of the lead formed from uranium decay. Such a mineral is zircon, which is formed during the formation of granite stones. In measurement of uranium and lead in zircons from granite mountain boulders west of the city Eilat, ages of some 620 million years were obtained. Note that in this sort of dating one may use both uranium-235/lead-207 and uranium-238/lead-206 combinations. If one obtains the same date (within standard deviations), one has testimony to the veracity of both methods.
We can also date samples in which there was lead from the very beginning. Measuring the number of minerals (or rocks) with a range of uranium/lead ratios formed at the same time gives us a straight-line diagram, and we can calculate the age of rocks from the angle of the line. We have many other combinations available to us, such as rubidium-87 decaying to srontium-87, samarium-147 to neodymium-143, hafnium-176 to lutetium-176, or rhenium-187 to osmium-187. All of these combinations allow dating using similar methods. There are rocks which can be dated through a number of different combinations and the ages obtained match each other (illustration 4). This result strongly supports the underlying theory of all radiometric dating methods--the assumption that rates of decay do not change with time. If the rate of decay of the different elements changed over time, the correlation between different methods would be broken. Considerations stemming from the model of the nucleus currently accepted by physicists also suggest a lack of change of radioactive decay rates over time.
Illustration 4. Dating of a single stone (sample of stone 10072, lunar basalt) using different combinations of isotopes. The points are the ages obtained and the lines show the estimated deviation in this determination. (There is a 95% chance that the age determined will fall in the range defined by these lines.) The light yellow rectangle shows the deviation of results from their average. You can see that aside from the first result, all results show approximately the same age, 3.55 billion years. Red--age based on decay of samarium to neodymium, black--rubidium to strontium, dark blue--potassium to argon in various minerals, light blue--potassium to argon in a sample of dust which represents the entire rock. The lower ages were calculated using this method and apparently stem from a loss of argon in the sample. (The least argon is now found in the sample which had the least time for potassium to decay into argon, and therefore the age seems younger than is the case.) The ages are taken from G.B. Dalrymple (1986) USGS Open-file report 86-110.
An excellent example of this is the processes which take place in a nuclear reactor. In a nuclear reactor atoms of uranium-235 are smashed in chain reactions which release a great deal of energy. The process of fission in the reactor is governed by the same constants that regulate uranium's half-life. The question of whether these constants have changed over geological time has an amazing answer: in the distant past natural atomic reactors were formed in the Gabonese Republic, Africa. These are uranium deposits in which a natural process of uranium-235 fission takes place. Fission reactions effected other elements in the rocks and caused changes in the ratios of the various isotopes of elements like samarium and europium. The ratios measured in these rocks are similar to those obtained in nuclear reactors and testifies to the lack of change in the constants which govern the processes of fission and decay. The great age of the rocks (half the age of the earth itself) testifies that for at least half the history of the earth these constants have not changed.
Using radioactive combinations, geologists succeed in dating rocks and important events in the history of the solar system and of Earth. To date, Earth rocks whose age approaches 4 billion years have been dated. Zircon crystals in these ancient rocks have shown even greater ages, up to 4.1 billion years. (The zircon was formed from magma some 4.1 billion years ago and after the original rock was worn they were transported, sunk, and became part of young sedimentation.) Older rocks were found amongst the moon rocks, and the earliest specimens reach back some 4.5 billion years.The oldest radiometric dating measured is of meteorites. Most meteorites were formed over a period of tens of millions of years and the oldest substantiated dates measured are 4.56 billion years before our time.
It is common to see that period as the start of the formation of the entire solar system. The sun was formed from a gaseous cloud which, because of its gravity, was almost completely gathered in the formation of the sun. Part of the gas formed the disk which surrounded the young sun. As the disk cooled, dust was created and coalesced into small solid bodies--meteoroids. Most of the meteoroids coalesced to form planets, including the planet Earth. A few remained in the asteroid belt between Mars and Jupiter, which from time to time ejects a meteoroid that is drawn and falls to Earth. Because of their small size they froze and retained their ancient age. It is now accepted that the moon was formed a little after the formation of the meteoroids, as a result of the collision of a large object (approximately the current size of Mars) with the planet Earth. The collision scattered a great deal of material, some of which returned and was drawn to Earth, and some of which formed the moon. The order of formation, then, was the sun, then the meteoroids, most of which coalesced into planets (one of which is Earth), and, finally, the formation of the moon. Earth, like other planets, has remained active and dynamic to this day. The smaller moon froze after a shorter period of activity (up to 3.1 billion years ago) and also retains its older rocks.
Can we verify this theory? A strong correlation between the oldest ages of rocks from Earth, the moon, and meteorites (all above 4 billion years old) gives us a general framework for the formation of the solar system. The increase in age correlates with the chain of events suggested. Earth rocks have seen a lot of history since the formation of the planet and give us younger ages. The moon went through fewer changes and maintained the older ages. Meteorites maintain the original make-up of dust in the ancient sun (the chemical composition of the most primitive meteoroids is very close to that of the sun) and gives us the oldest ages, the age of the formation of the sun. Another support for the theory comes from discovery of the disks surrounding the young stars common today, from theoretical calculations of the process of collapse of gas in space to the formation of the sun and the disk of dust and gas around it, of the development of nuclear reactions in stars like the sun, and from the dynamic of the collection of dust to form meteoroids and planets orbiting the sun.
Can we date Earth itself? Can we determine how old are the materials of which it is made? The matter is not simple. Since Earth has gone through many changes, no rocks from the time of its formation remain. To date the planet Earth we must find a sample to represent the whole and to which we can compare meteorites. We cannot complete the task, but we can come close. In the Fifties, Clair Patterson examined the ratios of lead isotopes (created as a result of uranium decay) in meteoroids and saw that on the appropriate diagram they fell in a straight line denoting an age of 4.55 billion years. To compare that to the planet Earth, Patterson examined sediment which had gathered in the center of the ocean. He hypothesized that this sedimentation came from many tunnels, which combined well-represent the continents. The ratios measured from the ocean sample fell along the same line described by the meteoroid samples, a line whose angle correlates with an age of 4.55 billion years. This observation greatly supported the hypothesis that the planet Earth was formed from an accumulation of meteoroids in the early stages of the development of the solar system. Since then the theory has been tested in different ways and has been verified again and again. Illustration 5 shows the ages of meteoroids and moon stones. We can see the precision of measurement (length of the vertical lines) and the strong correlation with the various methods of isotope measurement.
Illustration 5. Determining the age of the solar system by dating samples of meteoroids and moon rocks. The ages of the older moon rocks match those of the older meteoroids. Ages shown here (from G.B. Dalrymple (1986) US Department of the Interior, Open-file report 86-110) and additional ages suggest that the solar system was formed some 4.56 billion years ago.
The careful reader will have certainly noticed that this essay presents dating as a supposition and not as verified fact. The reason for this is not a problem in the method of dating, it is part and parcel of the scientific philosophy. In contrast to the religious mind-set which places the truth in the past and relies on more recent information less than on the words of earlier people, the scientific mind-set is more modest. The natural sciences do not provide absolute logical proofs. We observe the world, gather our observations, develop theories, and test them. A reasonable theory which survives repeated testing we call a hypothesis, and with the passing of time and repeated verification we tend to treat it as a law (like Newton's law of gravity). But laws, too, are not absolute truth; truth is part of the unattainable future. We work to bring the truth closer and improve our knowledge base all the time. It is not certain that we can ever reach the absolute truth, but the knowledge which we have accumulated allows us to understand nature and determine ways in which we can improve this understanding. For every practical purpose, our present understanding serves us well.
For example, we now have a good model of the build of the atom. The model works well and permits us to understand the connections and reactions between atoms and molecules, to develop new materials and medicines, to understand the source of nuclear energy and to take advantage of it. There is a chance that one day we will develop a different model which will explain everything we already know in an even better way. This doesn't effect the value of the present model. In our day-to-day life we treat this model as the truth, current as of today, and we work excellently with it. The age of the planet Earth or the age of ancient human bones found in caves is known to us with a high level of certainty. Of course, the determination that the age of trees, bones, and rocks is greater than 5,800 years has been verified in many different ways, only a few of which have been presented here. The veracity of this determination is much stronger than any proof we know that the age of Earth is only a few thousand years.  We determine the age of the planet Earth or the age of the solar system with the same level of reliability as the distance to the moon, to which we have already traveled. The basic approach of the scientific method obligates us to see everything as a theory, no matter how well based. But we must always remember that this theory, like many other scientific theories, is based on multiple observations, verified by many experiments, and, most importantly--it can be examined by more observations and experiments; it has been examined, found to be correct, and gives significant results time after time after time.
Oded Navon is a professor at the Hebrew University of Jerusalem's Institute of Earth Sciences. Dr. Mordechai Stein is a researcher at the Geological Survey of Israel in Jerusalem.
 For more on this issue, see the excellent article by Dr. Roger C. Weins at http://www.asa3.org/ASA/resources/Weins.html