Quality Control and the Research Question

December 9, 2022

Some minor quality control issues were evident in our analyses. These issues, and their potential effects on our data, are discussed below.

Dripwater:

Due to a user error during analysis, our B-check quality control replicate was closed and not run as often as we would have hoped (every 22 samples). Instead, it was only analyzed twice, before any of our unknowns were analyzed. Thankfully, we also ran the NIST 1643f quality control standard every 22 samples, so we were able to evaluate the quality of our data using that standard. Because of this backup, we are able to remain confident in the quality of our data, and therefore in the plausibility of our hypotheses regarding the dripwater unknowns.

Some trace metal concentrations fell within the lower half of our calibration curve. In the future, we would like to add a bit more of these trace analytes to our calibration standards. However, since all of our unknowns still fell within the calibration curve range, we are confident in the quality of our data even at low concentrations.

Plate Calcite:

Due to a dilution error in which we diluted our Si and P single-element stocks an extra 10x in the creation of our calibration standards, our Si/Ca and P/Ca ratios were outside of our calibration range. We are not able to be confident in the accuracy of those two analytes for our plate calcite method. In future analyses, we will need 10x more Si and P in our calibration standards.

Some of our quality control recoveries for plate calcite standards were shaky, especially for the B-check and NIST 1643f standards. Our MACS-3 and FEBS-1 recoveries were relatively accurate, with average recoveries of 1.04 and 1.01 (excluding nickel), respectively. However, due to the accurate recoveries for our key analytes (Sr, Ba, Mg), which have an average recovery of 1.05, we are confident in the quality of those plate results. 

In the future, we would like to run the plate calcite using the double-dilution dripwater method (to measure TE in one dilution and Ca in the other) to check the accuracy of our new method. This would allow us to directly compare the old method and the new method to determine the relative accuracy. We will also consider purchasing or producing new, well-quantified speleothem standards to better optimize our quality standards for calcite plate analysis.