Streamlining data integration and sharing by small ecology labs

Background/Question/Methods

Defining, understanding, and modeling regional environmental change are dependent on integrating many data sets collected across a region of interest. Local data sets are often collected by small academic labs that use inexpensive automated sensors to augment or supplant manual field measurements. These labs often struggle with sharing data freely and efficiently, limiting their broader impacts and participation in synthesis efforts. Ideally, their data sets would be well documented, discoverable, and downloadable from computers and phones. In reality, data availability generally lags far behind collection due to a lack of resources for hiring dedicated information managers or to the lack of a free, easy-to-implement information management system that handles data from multiple disciplines.

Our interdisciplinary team of ecologists and computer scientists developed a new information management system that helps small labs with their data management and sharing. We tested this framework on a typical academic lab, the Systems Ecology Lab (SEL) at the University of Texas, El Paso. SEL researchers integrate data from manual field work and >100 automated sensors at a Chihuahuan Desert research site (within the USDA-ARS Jornada Experimental Range, NM) to study controls on land-atmosphere interactions in this ecosystem. The site is also used test new sensors.

Results/Conclusions

The new information management system allows researchers to access and visualize hundreds of data sets quickly and efficiently. It was developed specifically with the needs of ecologists in mind, using mostly free, community-supported and developed software that many ecologists are already familiar with. Our generic and customizable schema supports sensor- and manually-collected data and is implemented in MySQL.  Most data analyses are performed by stored procedures in MySQL or with external R scripts. While maps are generated with commercial ArcGIS software, graphs are dynamically created with free Javascript components in HTML5, which supports the use of all major cell phones and tablet computer operating systems.

Development is ongoing and focused on the transition from more traditional data management activities to our new system, which provides an alternative to commercial packages currently in use. We are confident that this conceptual framework, code, and components can be used, modified, or extended by other labs that wish to have an information management system for their labs.  Ultimately, we envision that this type of system will help researchers better adapt to the dynamic needs of the environmental sciences. Future development will focus on the semantic annotation of data and processes.