Workshop to define quality management standards for data completeness in derived data products

Integration of data sets from multiple sources is an important prerequisite for network level and cross-site synthesis studies in LTER, and scaling or re-sampling of primary data is often performed as part of this process. In the EcoTrends project, for example, fine-scale (e.g. hourly) measurements collected by LTER sites are rescaled to monthly or yearly time steps for analysis and display of long-term trends. However, discussions within the LTER IM community and in ASM workshops elucidated problems with both handling and reporting of missing and flagged measurements in source data when derived data sets are produced by and for the EcoTrends project and also by the LTER ClimDB/HydroDB database. We believe it is essential that LTER establish standards for re-sampling and temporally scaling incomplete and flagged measurement data used in synthesis projects, and for documenting this process to provide quality control information for the derived data products. In addition to benefiting LTER synthesis projects, these standards will inform development of data integration tools by LTER sites, LNO, and other projects and organizations such as NCEAS, SEEK, the Canopy Databank and CUAHSI. In order to address these issues, we conducted an initial 2-day workshop on Jan. 31-Feb. 1, 2007, in Las Cruces New Mexico. This workshop was attended by 13 people with broad expertise in environmental monitoring, data synthesis, and information management. Participants represented 9 LTER sites, the LTER network office, CUAHSI, NCEAS, the Canopy DataBank Project, as well as various standing committees and working groups in LTER (NISAC, IM-Exec, EcoTrends). Additional information about the workshop, major findings and products is presented in the attached report.