Project Specific Innovations

In GLEON, analytical models often are innovated to suit a particular science question and the available data. These models and their use are growing in their sophistication and demands on computer resources. For example, determining uncertainty estimates in daily lake metabolism parameters can be accomplished through a bootstrapping technique that refits the model 1,000x per analysis day (Science Applications – Lake Metabolism Across the GLEON Network). The compute demands are not high for one day. However, when 25 lakes are analyzed over months there are thousands of parameter days to bootstrap. This requires distributed computing to complete in a reasonable amount of time. We expect GLEON scientists to increase the use of these kinds of analytical approaches.