Home Page General Articles Genesis Science Mission
Scripture Categories Genesis Science Mission Online Store

Creation Links

Anomalies Creation Science Talk Blog

Concordant Dates

Concordant Radiometric "ages" from
Isua Series rock of Greenland

Method "Date" in Billions 
 of years
U-Pb 3.60+/-0.05
Pb-Pb 3.56+/-0.10 
Lu-Hf 3.55+/-0.22 
SM-Nd 3.56+/-0.20 
Rb-Dr 3.62+/-0.06 

From: Radiometric Dating II

Charts such as this are often used to illustrate the consistency of  radiometric dating. The claim is that the agreement among such dates shows the legitimacy of the dates. Now concordance ( agreement ) does occur, sometimes vary closely, but this proves nothing about the accuracy of the dates. The first  assumption is that the odds of coincidental concordance are extremely small. The second is that the only process that would produce concordant date is radioactive decay over time.

Curiously the only place on the internet that  I could find charts of radiometric dates such as the one above are on web sites that are attacking Young Earth Creationists. The only exception is the source sited above but they relied heavily on two web sites that do. What such sites don't show you is that vast disagreements in dates can be demonstrated as well.

When I found a chart of dating data from The Alaska Division of Geological & Geophysical Surveys, I found a different story. The data was collected from various locations scattered over Alaska. The report containing the charts called  Radiometric dates from Alaska - a 1975 compilation is simply a chart of data with no interpretation beyond the date calculations.

Now 62% of the locations had only one date and so there was nothing to compare the dates to. Another 35% were remeasured by the same method and   therefore they only show the isotopic consistency of the rock. Only 3% of the locations had dates by different methods.

Percent difference 
between dates
occurrences %
Identical 14
<5% 22
5-10%   7
10-25% 29
25-50% 14
>50% 14

The result is that after evaluating the radiometric dates from 509 locations scattered over Alaska, is that there is no statistical indication of trend favoring concordance. Yes, this is based on only 14 samples, but that shows how seldom more than one dating method is used ( or reported ) in radiometric dating.

As stated before vast disagreements in radiometric dates can be demonstrated as well. The following is data on three difference rock formations in the Grand Canyon. The isochron ages are from the K-Ar and Rb-Sr model ages next to them.

Uinkaret Plateau
Ages of Rocks in Millions of Years

K-Ar Rb-Sr Rb-Sr Isochron Pb-Pb Isochron
0.01 1230 - 1310 1300 - 1380 2390 - 2810 
1.0 - 1.4 1260 - 1380    
2.63 1310 - 1370    
3.6 1320 - 1440    
3.67 1360 - 1420    

Data from: Radioactive Dating Explained - Part 2

The Uinkaret Plateau samples are from lava flows on the plateau, at the top of the Grand Canyon and some flows down into the canyon so that the eruption must have occurred after canyon was formed. These are indisputably some of the youngest rock in the canyon, but yet it's radiometric ages range from a 10,000 years K-Ar date to a 2.81 billion years Pb-Pb Isochron date.

Even still the Rubidium strontium dates place the rocks at about 1.3 billion years. This would make them pre-Cambrian rocks, rather than the modern Quaternary rocks; which they obviously are.  There is overlapping of the Rb-Sr dates, meaning that they are concordant, as is the Rb-Sr Isochron date but yet the dates are clearly wrong. Also the bottom two K-Ar dates are concordant but disagree dramatically from the concordant Rb-S dated. None of the dates from different methods overlap, other than the Rb-Sr and Rb-Sr Isochron.

Cardenas Basalt
Ages of Rocks in Millions of Years

K-Ar K-Ar Isochron Rb-Sr  Rb-Sr Isochron
771 - 811 682 - 748    920 -1040 1000 - 1140 
809 - 877     990 -1130   
838 - 868     990 -1190   
780 - 820   1010 -1170  
800 - 840   1030 -1110  
    1050 -1150  

Data from: Radioactive Dating Explained - Part 2

The Cardenas Basalts are classified as late pre-Cambrian rocks and the radiometric dates reflect this but there such disagreement among the dates that we need to ask which age should you believe? The only overlap in method is between the Rb-Sr and Rb-Sr Isochron. Once again we have concordance among the dates of a given method but not between methods, so which is correct?

Diabase Sills
Ages of Rocks in Millions of Years

K-Ar  K-Ar Isochron Rb-Sr  Rb-Sr Isochron
874 - 954 926    680 - 1020 1000 - 1140 
924 - 984     740 - 1100  
    1030 - 1070  
    1030 - 1090  
    1100 - 1280  
    1300 - 1440  

Data from: Radioactive Dating Explained - Part 2

The Diabase Sills are some of the deepest rocks in Grand Canyon and so they should be the oldest, but yet they yield some of the youngest Rb-Sr dates. Now it did yield the oldest K-Ar Isochron dates, but they were not old enough, since they should be several billion years. Curiously the Rb-Sr Isochrons of both Cardenas Basalts and Diabase Sills are the same. This one is better on the concordance of dates between different methods but only because Rb-Sr dates spans about 800 million years, but internally the Rb-Sr dates are hardly concordant at all.

The real question is how likely is coincidental concordance of dates this question was first dealt with in a study using completely random data, by John Woodmorappe in his book Mythology of Modern Dating Methods. The study show a high likely hood of fortunate concordance.

Each trial consisted of strings of 100 random numbers. The range for concordance defined as +- 2.5%, so that overlapping ranges were considered concordant.

 When I tried his little exercise I got the following results. First try concordant pairs which are at the same point in their respective lists occurred about  4.75% of the time. When the position on the lists were not important the odds of getting concordant pairs in random data were vary good. In fact one needs to go to more than 6 concordant pairs before the odds of  concordance drops below 50%. 

# of concordant pairs 
on the other list 
% of time concordant 
pairs occur
2 90%
3 80%
4 70%
5 60%
6 52%
7 41%

The fact is that the odds of fortunate concordance in random data are quite good. This is particularly true given the large number of available samples of radiometric dates.

Now real radiometric dates are definitely not this random, at the very least the physical processes that affect isotope ratios limit date ranges increasing the odds of fortunate concordance. So the odds are sufficiently in favor of fortunate concordance that such agreement of dates can not be used to claim accuracy.


Custom Search