Repository overview: This repository houses data downloaded from NOAA's National Estuarine Research Reserve database (http://cdmo.baruch.sc.edu/), and scripts for importing, cleaning, processing, and analyzing data for nutrients, water quality, and climate collected at Old Woman Creek (a tributary to Lake Erie) and the York River estuary (a tributary to Chesapeake Bay)
Project overview: Nutrients are essential drivers and indicators of aquatic ecosystem function and health. Both water quality and climate factors influence nutrient sources, sinks, and processing, but these relationships can be complex (e.g., non-linear, or additive). Further, nutrient data are currently collected via grab-sampling and laboratory analysis, while water quality and climate data are available at high temporal resolution via in-situ sensors, resulting in a disconnect in frequency of available data streams. Here, we seek to utilize machine learning methods that offer flexible, robust alternatives to traditional methods for modeling driver-response relationships (e.g. multiple linear regression). We have selected two NERR sites proximal to COMPASS study locations with contrasting ecosystem characteristics (small stream vs large estuary, freshwater vs saltwater, seasonal ice cover vs ice-free, etc.) to develop our statistical approach.