Skip to main content
ARS Home » Plains Area » Las Cruces, New Mexico » Range Management Research » Research » Publications at this Location » Publication #337114

Research Project: MANAGEMENT TECHNOLOGIES FOR CONSERVATION OF WESTERN RANGELANDS

Location: Range Management Research

Title: Analytical approaches to quality assurance and quality control in rangeland monitoring data

Author
item Mccord, Sarah - New Mexico State University
item Karl, Jason
item Van Zee, Justin
item Courtright, Ericha

Submitted to: Society for Range Management Meeting Abstracts
Publication Type: Abstract Only
Publication Acceptance Date: 1/9/2017
Publication Date: 1/29/2017
Citation: Mccord, S., Karl, J.W., Van Zee, J.W., Courtright, E.M. 2017. Analytical approaches to quality assurance and quality control in rangeland monitoring data [abstract]. Society for Range Management, Jan 29-Feb 2, 2017, St, George, Utah.

Interpretive Summary:

Technical Abstract: Producing quality data to support land management decisions is the goal of every rangeland monitoring program. However, the results of quality assurance (QA) and quality control (QC) efforts to improve data quality are rarely reported. The purpose of QA and QC is to prevent and describe non-sampling errors that introduce noise into monitoring datasets, thereby increasing the repeatability, defensibility, and usability of the data collected. Quality assurance is a proactive process designed to prevent errors from occurring, while QC is a reactive process whereby the number, nature, and implications of errors are identified. Common QA practices include careful design and documentation of the monitoring programs and protocols; training and calibration of data collectors; and structured management of resulting data. Quality control describes errors via data checks for incomplete or invalid values, variance decomposition, and evaluation of signal-to-noise ratios. We analyzed the calibration results and field data of two national monitoring datasets collected between 2011 and 2016. Field and lab calibration data demonstrate that the iterative learning by data collectors which occurs in QA improves field sampling results and can point to areas of future training. We also explored quantitative QC approaches for identifying and visualizing erroneous observations by comparing results to other data sources; and analyzing between-observer variability, patterns of missing data, and seasonal trends in method implementation - all of which help to describe the errors in these national datasets. We conclude that the results of the QA and QC processes can and should be evaluated to improve and document the quality of monitoring data. These evaluations can be used to improve monitoring data collection efforts and to support the role of rangeland monitoring in decision making.