The value of consistent methodology in long-term environmental monitoring

Long-term monitoring has a substantial history in both the biological and physical sciences. Over time the procedures and analytical methods involved in long-term monitoring have changed to improve the quality of data, but even over short time spans, differences occur that can make direct comparison of measurements either difficult or impossible. In many instances the lack of strictly defined methods or practices means that data from one project cannot be used to enhance other projects with any degree of statistical rigour. This is amply demonstrated in the field of soil classification where improvement in soil definitions, refinement of cut-off points and changes in descriptive techniques between soils is such that in many cases direct comparison of old with new data is impossible. The causes of, and safeguards against, such measurement inconsistency are examined here in the context of the United Kingdom Environmental Change Network (ECN) project. Examples of incompatible data arising from environmental studies are given and the efforts used to standardise methods and practices in the ECN programme are described in detail. The need for standard practices is demonstrated and considered in the light of the limitations of operating what are relatively rigid procedures.