Big Data and Ecological Forecasting: Integrating NEON Observational and Sensor Data from Reach to Continent

Abstract:

The ability to forecast the response of varied ecosystems to changes in climate and land use will be crucial for the management of resources and ecosystem services. Ecological forecasting presents many significant challenges within each of the aspects of data capture, assimilation, and modeling. High space-time resolution sampling is required to address the challenges of scaling from the site level to the continent. Determining the uncertainty of data used for model input and parameterization is critical for constraining the model for accurate representation. The National Ecological Observatory Network (NEON) is poised to greatly expand the scale and availability of biogeochemical and aquatic ecological data. NEON is a continental-scale facility designed to collect and disseminate data that addresses the impacts of climate change, land-use, and invasive species on ecosystem structure and function. Using a combination of standardized observational sampling and sensor measurements, NEON will provide a rich source of biogeochemical and biophysical data from 34 aquatic and 47 terrestrial sites spatially distributed across the US, including Alaska, Hawaii and Puerto Rico for 30 years. Sites were selected to be representative of major ecosystems and maximize scalability. In addition to standardizing measurements, NEON is determining the quantitative uncertainty of each data product making them well suited to constrain models. NEON aquatic data will not only serve to baseline aquatic ecology in major ecosystems but also presents opportunities to bolster Hydrologic Models as well as incorporate aquatic biogeochemical cycling into Land Surface Models. Here we present examples of published and provisional data currently available from deployed aquatic sites, as well as an overview of the full scope and release schedule of the open source ecological data to be published on the NEON web portal. Several use cases, such as whole stream metabolism, groundwater exchange, high-resolution geomorphology, and modeling applications will be presented as well as the methodologies used to determine measurement uncertainty and automated quality control algorithms.