NTRODUCTION

The role of data management with regard to acquisition, evaluation, assessment, publication and documentation of project data in scientific projects and monitoring programs is very often underestimated. Because of different objectives, scales, time periods and other project relevant factors, it is not possible to offer a detailed generic template for a data management plan for scientific projects. But, in fact, most projects are liable to similar data collection, processing, assessment and storage.

This overview intended to emphasize the importance of an integrated data management in scientific projects and monitoring programs.

AIMS AND BENEFITS FOR PROJECT DATA MANAGEMENT

The main task of project data management is to ensure an end-to-end processing of data which starts with raw data collection and ends up with the scientific results of the project. The main steps and the advantages of an integrated project data management during the project are:

Data acquisition and quality control

Raw data from primary data sources like sensors, surveys, etc. has to be calibrated and corrected and sets of basic usable data have to be compiled. Quality, validity and consistency checks, including attended documentation, guarantee a definite basis of all data sets. An early and close contact of scientists and data managers allows an immediate and well defined data acquisition from the beginning of the project. Only close cooperation between data originators and data managers guarantees high data quality.

Data processing

For further temporal and spatial analysis, the project data must be formatted and processed. Data management requires standardized and documented data processing which allows for an optimal project data flow with more efficient cooperation of the different project branches.

Data publication and data storage

Data assessment and the publication of results are the final steps of the project.

An integrated data management will decide processing, presentation and publishing software in an early stage of the project. This allows a project product with long-term storage and public access to project data with additional presentation tools. The product will increase the reuse of data and, therefore, the value of scientific projects.

The integration of data management in scientific projects means an economic increase of value, more acceptance and help in environmental management and an enrichment of science in the future.

DATA MANAGEMENT - AN INTEGRATED PART OF SCIENTIFIC PROJECTS

After the definition of the project objectives by scientists, the development of a data management plan must already have been initiated in the implementation phase of the project. At this point, data managers have to participate in project meetings to integrate data handling experiences and establish personal relationships with scientists, which is of utmost importance. The relationship between data originators and data managers is the key to successful data management.

Project data management in practice

In most projects, the data is of a very different nature and often a large amount of coexisting data exists. After publishing the results of the project, scientists are often not interested, or cannot increase effort in delivering the used data to the project data management for long-term storage. Often, the data which would be valuable for the future is lost forever. Therefore, a sophisticated data handling system is needed with a well defined data collection plan which must be highly acceptable by the scientists and easy to use. The delivery of data to project data handling for a reuse of the data should have the same praiseworthy status as publishing the results.

Normally, the incoming project data is incomplete, hardly even documented, and the necessary originator quality check is missing. But, project data acquisition requires the most input for data management. That is the reason why techniques of how to get data (data tracking) are more important than the correctness of data.

The outcome of a project must be a reusable information product with standardized data sets, description of used formats, protocols of statistical and processing procedures, the complete meta data, documentation of data tracking, possibly the used software tools, etc.

The publication and presentation of data in a final product (CD-ROM, WWW, interactive viewing tools, multimedia applications, etc.) should, just like the data management plan, have already been defined in the implementation phase of the project. If the project or monitoring program was paid for by public funds, the same public should have the possibility of getting access to a final product. To the target audience of scientists, managers and politicians, this fact increases the possibilities for data reuse and the general acceptance of scientific projects. The higher value of not isolated data and the higher level of trust of well documented data speaks for itself.

Financial and temporal requirements of data management

It is estimated that projects require three times of the time and cost for data management as for practice. Data delivery should be part of the contract of the project and be finished 1/2 year before the end of a project. In most cases, data delivery stops at the end of the project and no money is available for further and necessary data management, as mentioned above. Therefore, the funds for data management should be fixed beyond the project end.

Authors address:

Gerold Lüerßen
CWSS
Virchowstr. 1
D-26382 Wilhelmshaven
luerssen@cwss.whv.net