Go to the next chapter.
This provides a non-technical overview of the Electronic Rothamsted Archive, looking at why we need it, and what it does.
Based on an article in the December 1992 Edition of Agrenet News. Copyright © 1992 Andy Caiger, Rothamsted Experimental Station.
Rothamsted Research (also known as Rothamsted Experimental Station) celebrated its 150th birthday in 1993. Not only have agricultural experiments been conducted on the site since 1843, but some of these experiments have continued right up to the present day. The world-famous Classical Experiments conducted in the fields at Rothamsted are a unique inheritance from the earliest days of agricultural science, continuing from the dawn of the industrial revolution to the present day. Such long runs of data records are increasingly in demand to help answer questions concerning issues like global climatic change and nitrate leaching.
Unfortunately, only a small part of the data from the Classical Experiments are available in an electronically readable form. Most data exist in published or unpublished papers, archives, notebooks or on data sheets. In addition, information that is vital for interpreting the data (for example, concerning changes of treatments) is often known only to a handful of experts, some of whom have already retired.
As we enter increasingly into the information age, the Electronic Rothamsted Archive (ERA) project aims to provide a permanent managed database to securely hold all this important data and associated information. Users from Rothamsted and elsewhere will then not only have easy access to the data, but have specialist information available to them as well.
Datasets currently available in ERA include long-term daily meteorological records; yields and species composition data from the Park Grass Experiment; and yields from the Broadbalk continuous wheat experiment. ERA can be used to archive contemporary data as well.
The Electronic Rothamsted Archive is based around an ORACLE system running on Unix. ORACLE provides all the database management facilities that an archive needs - indexing, concurrency control, integrity and security - whilst programs specific to ERA, written in Tcl-Tk, C, SQL*Plus, and PRO*C (ORACLE's own C/SQL mix) perform tasks like data entry, description and extraction.
Because of the nature of 150-year old experimental data, and ever-changing experimental practice, the archiving of these long-term experiments is a far from trivial problem. To illustrate the ERA solution, let's take a non-technical look at the `life-cycle' of some data, from an old tatty page to the disk space of a user.