The determination of the amount and size distribution of carbonate and noncarbonate materials in a given sediment sample is significant to coastal engineers and geologists. Typically, laboratory procedures physically remove the carbonate portion of the sample from the entire sample, so that the amount and distribution of noncarbonate material remaining can be established. Rather than performing such a procedure, the authors have proposed a Bayesian Carbonate Determination (BCD) technique, which can be used to calculate the percentage of carbonate material in a bimodally distributed sediment sample from data obtained through an initial sieve analysis.
The BCD technique is based on the assumption that the combined probability density function is bimodal and that the carbonate and noncarbonate portions of the sample follow independent distributions. The BCD technique employs the grain-size probability distribution to determine the threshold point that separates the two sample portions with minimum error. For samples with bimodal sediment distributions, the errors associated with use of the BCD technique have been found to be less than the standard deviation between the samples.
The present study extends the original investigation that introduced the BCD technique. This article identifies the limitations of the BCD procedure and introduces a more general Carbonate Determination Filter (CDF) procedure for determination of the percentage of carbonate within a specific size range for a given sample.