Quality Assurance Plan

The University of Washington, Department of Earth and Space Sciences, Stable Isotope Laboratory (IsoLab), strives to generate quality data from stable isotope analyses. Our quality assurance plan summarizes our practices from sample receipt to data delivery.

Sample Set Organization – Samples are entered into our samples database by the researcher interested in the data, or otherwise responsible party, via IsoLab’s website and a unique sample set identifier is created. The samples database records set creation date, contact person, analysis, number of samples, sample list filename, and any comments from the submitter. Once samples arrive in the lab, the submitted sample list is checked against the physical samples, and the set is entered into the analysis queue. A directory is created with the unique identifier where all metadata, raw instrumentation data, and data reduction products are stored.

Sample Analysis – All samples are analyzed along with three reference materials. Two are used to generate a linear two-point calibration curve to appropriate international scale (e.g. VSMOW for water). The third is treated as an unknown and serves as our quality assurance reference material. This third reference material provides estimates of precision and accuracy. Prior to starting an analysis session, a log is completed by the user to ensure an appropriate and consistent instrument state. In addition, a longer-term log is used to document instrument maintenance, concerns, and problems.

Data Reduction – The raw data file(s) is transferred from the instrument computer to the sample set directory mentioned above where it is archived. From here MatLab is used to view the run, make appropriate corrections, generate final data and a reduced data file. Data from reference materials included in the run are also stored into a collated log file for the purposes of long-term monitoring of the method.

Delivered Data – The reduced tab-delimited text file referred to above is formatted as a report and includes run information such as reference materials, as well as precision and accuracy. The sample data contained within this file are checked against the original sample list to ensure completeness. Final data are emailed to the appropriate researcher.

Data Storage and Archiving – The raw file is a comma separated values file (.csv) and the reduced data file is a tab-delimited text (.txt) file. Two copies of all data are stored: one on our server and one in a separate-building backup server. Hand-written lab notebooks for sample prep / loading indicating technician, date, and sample IDs are kept and stored in the lab office.