A streamlined, array based, alternative framework for processing FVCOM data that runs faster than PyFVCOM (but lacks high-level features) for those conducting big data analyses who do not have access to a computing cluster. I have also included R code (built from a great framework provided by Mark Rowe (NOAA GLERL)) for node based processing, spatial analyses, etc.. but R is prone to run into vector memory and speed issues on most PCs with Big Data.. even with the parallel, garbage collection, chunking, and full array based operations. ** See the framework, dims/operations used and note that python and R read .nc files differently before using/modifying code to save yourself a headache of playing with (XXXX) dimensions in your head between variables and programs.**
The FVCOM data was provided by Mark Rowe (NOAA). The FVCOM model used to generate the output is described in (Rowe et al., 2019).
A volumetric analysis of Cisco habitat using a GRP (growth rate potential) modeling framework (Arend et al., 2011, Hanson et al., 1997; Rudstam et al., 1994; Kitchell et al., 1977). Analyses of Lake Erie water quality and cold-water fish habitat for 17 historical years. Includes two Dissolved Oxygen Penalty frameworks (Arend et al., 2011; & a framework I may publish from work in my '.._IN_lakes..' repository- inspired by Jacobson et al., 2008 & Arend et al., 2011).
The regex file pattern for reading in files is due to variable file size, dates and structure where the FVCOM can generate output files that contain 23 hours of one day and 1 hour of the next day (..lehf dirs...).
Assess your file structure, timescales of summary, etc., etc., etc., and modify appropriately for your task.
If you use this software or derivatives in your research, please cite:
Kumar, R. (2025-2026). FVCOM output array-based framework: GitHub. https://github.com/RajeevKPurdue/FVCOM_ArrayBased_Processing_CiscoGRP
This project is licensed under the MIT - see the LICENSE file for details.
If you use this code in published work, please acknowledge its use and consider citing it for any/all relevant publications.